I ran a Game Jam all about randomness! Check out the results of the 2023 GMTK Game Jam, where the theme was "Roll of the Dice" - th-cam.com/video/XNCGdi2A6fQ/w-d-xo.html
In players' defence, it's not just the disconnect between probability and our broken understanding of probability that causes anger (although that's part of it); it's also the disconnect between our intuitive understanding of *the situation* and the outcome. Yes turn-based systems are always abstract but I'm still seeing a special forces soldier point a rifle at a stationary target 3ft away and somehow miss. It completely shatters the fantasy of controlling this awesome elite squad.
I agree with this and to add on with my experience. I used to play a lot of fortnight and apex the former being discontinued and the latter being pretty close to it. Both suffer from the problem that your likely to die in the first 5 or ten minutes because you got a pistol and sniper while someone else gets one of the broken guns and just slaughters you in a few milliseconds, for apex you might not get armor or anything to give you health while people get both in surplus. Some randomness I understand but if I continuously die like this I'm going to get angry
Add to that the bizarre way in which things act in sequence despite them supposedly being simultaneous - hence the miss chances and the like. Goes all the way back to DnD for this particular issue.
@@SolaceCafe that is more a problem on how the designer's design the level by making certain things too artsy like hologram's for example people know there is nothing there so why can't my guy shoot better missing 5 80% chances in x-com is unlikely but not impossible the problem is in X-Com the numbers are fudged in your favour as in a 80% is actually like a 85% i think missing is still bad but it is like rolling 1 5 times in a row on a d6 it is unlikely but you have to plan for it if you want to be good also in X-com something to remember is the RNG generated is the same when you save scum if you do so just doo attacks in a different order then the 5 misses you will have inone go will not be wasted on 80% shots
@@deathtoll2001 yes but D&D for example is actually more controlable average level 1 character will have a +5 to hit most things they will fight will have a AC lower then 15 so you will always be above 50% and rounds tend to not last that long in D&D epic battles tend to take less then 1 minute of actual time where as in like x-com rounds can be longer then ten and you still aren't finished with the mission even the best laid plans fail
I like luck in games bc i need to think even for if im unlucky so for example: I have a plan with 95% chance of succes But due to how probability works i should have backup plans for my backup plans
It's been talked about ad nauseum at this point, and still I want to draw your attention to 14:42 - something about X-Com's animation system is such a f***** slap in the face. 99% chance to hit, gun is deadlocked on the target. > Soldier moronically pivots to the floor to miss a full burst of machine gun fire at 2 yards. It'd be a legit way better experience, with the *exact* same number stats, if the animation somehow made sense. Like bullets flying underneath arms, or hitting cover really close by. At least that way the feedback would be "shit happens" and not "your soldier who you have spent a lot of skill points to upgrade and has seen many victories in many different parts of our alien-infested world is an absolute mongrel, but only every now and then."
The original XCOM (UFO: Enemy Unknown or XCOM: UFO Defense depending which version you had) was worse. You had "time units" (action points) which you spent for each step of movement, as well as when you were shooting, and even for turning on the spot. If you whiffed your shot, you'd fire in a random direction, but if that was more than like 20° off from the target, you had a risk of spending TUs turning and not having enough to fire any more. If you did have the TUs for it, you could move up your rocket dude, line up for a shot into an alien spaceship, and "accidentally" turn 180° and shoot your squad in the face, blowing up everyone with a single shot.
Agreed. I always wished the animation was one where the enemy would just dodge behind cover or just hit the deck at the last second. Or at least give enemies a DC to hit like they calculate already and subdivide lower rolls as reduced damage with one chunk reserved as a complete miss.
@@a-blivvy-yus Actually, I liked the old games way better because of time units. They gave you the flexibility of spending all actions on moving or solely on shooting, not forcing you to make 1 move and 1 shot every turn. I know there's actually an option for longer move without shooting, but still I find time units more flexible and allowing you to adapt to situation. E.g. you could be moving a step at a time consuming just small portion of time units and assessing your surrounding with every move. In newer games you would just make one step and deplete all your move capabilities. Moreover, time units were growing with your soldier getting experience, so after some time in the game your troops were quite a commandos. I suppose in the newer series it's the same pace through all the game (one "time unit" for move in a given proximity plus one "time unit" interchangable between shot or extended move).
@@michaniewiadomski7911 There are good and bad things to the new system. It's more streamlined and easier to manage without losing too much of the depth of the gameplay, but it is very distinctly different from the originals in a way that doesn't feel necessarily better. I like both types of gameplay for different reasons, so I won't claim one to be better than the other - which I play is dependent more on my mood at a given time than one being a better game.
@@michaniewiadomski7911 And the problem I'm mentioning was a problem partly caused by the game's use of TUs... which isn't a problem in the new game, and which was (eventually) patched in the originals.
I think you should have mentionned the games where the player is in charge of the luck. In Armello, you can burn cards to choose the result of the dice, so the player decides how much luck they're willing to rely on, and what price they'd pay for it. That way, when the result isn't favorable, the payer blames themselves for not paying more, instead of the randomness. Danganronpa also has a similar system with the MonoMono Machine, where you can pay more tokens to decrease the chance of obtaining an item you already own.
I played a game called void tyrant where you're basically playing dice blackjack against your enemies to attack, but you can play cards to change your rolls or attack directly, and you get to chose between playing more cards or doing more damage.
The old Fighting Fantasy gamebooks had a variant on this. You could spend 'luck points' to influence dice rolls in combat, decreasing damage done to you or increasing damage done to opponents, but your total luck points also acted as a buffer against unfortunate events where the player is asked to roll against their total luck score so there's a balancing act there. Luck points could be increased throughout the adventure, typically as a reward for making smart decisions or being curious.
Melee does that same thing with the trophy lottery, where you spend more coins to decrease the chance of a duplicate. It doesn't affect gameplay, but it's still an interesting system that likely influenced those games you mentioned.
@@bacchus9389 displayed hit chance in XCOM lies, and lies a lot - there is multitude of hidden factors (and mods to show them), including things like character/global miss streak modifier that up your chance of landing a hit. You can have guaranteed hits, the game just won't show them to you - and when you memorize how it works, it's possible to play around/abuse the system for your own advantage (like wasting low percentage shots to guarantee a takedown before end of turn).
@@asmonull Minor correction: XCOM2 doesn’t use these mechanics on the highest difficulty level (others aren’t interesting anyway). XCOM EU and EW don’t cheat in your favour at all.
@@asmonull actually it doesn't lie, u can intentionally go for low chance shots to boost the hit chance u need to kill a big bad enemy and the displayed hit chance will go up every time. Well... i guess if ur talking about ur soldiers hit chance then yeah the game doesn't tell u where that extra hit chance is coming from but the percent chance to hit an enemy is accurate.
The main problem is that people remember negative events more easily than positive ones. Yeah you'll sometimes miss in XCOM with a 99%, and you will distinctly remember just a couple rounds ago when you also missed with 99%. But you will forget the 100 shots you took in between that hit, they were exactly what you were expecting so you just move on and forget about them. This is also why no one complains about when they have strings of good luck, it's not just that the good luck makes them happy it's also that they are less likely to even realize that they are having a string of good luck at all.
I think the opposite is what makes dark souls so good, well at least for me. Instead of good things happening to you all the time, bad things do, this makes you remember the good moments because they are a lot rarer. For example, I don't remember all the times I lost to a boss, but instead, I remember how good it felt to finally beat that boss.
you never saw me play DnD 5e whenever i get a streak of good rolls, i get nervous that my luck will turn soon and i roll very bad. that makes you consider to think about your options, maybe instead of attacking the enemy with a direct attack that can miss, you instead cast a spell that weakens them and improves your teams chances, for example the 2nd level cleric Spell "Bane" cause that up to 3 enemies have to perform a wisdom check every time they make a check, if they fail, they suffer a 1d4 penalty on that throw. also this spell cannot miss, which makes it rather neat, but its a 1 minute concentration spell, so it limits what you can do. you could also use "Sacred Flame" (yes, im playing as a Cleric) which will never miss, and deals rather low damage but the enemy has to perform a saving throw to take half damage. this cantrip will guarantee some damage and i love to use it to finish off enemies. otherwise its rather weak. but i still go for the risky throws, surprisingly i dont fail my dex checks that often, even though i have -3 on all my dex checks. also critical hits with a weapon just feel awesome.
@@firebladeentertainment5739 "whenever i get a streak of good rolls, i get nervous that my luck will turn soon and i roll very bad." This is a perfect example of how stupid the human mind is when it comes to randomness. We have zero clue how likely a long string of a certain outcome is and we tend to couple events together when they are independent. It is so common to think of luck as a finite resource, so when we have great luck in succession, we are bound to some bad luck, but that simply isn't the case. If you ask people to throw a coin thousand times and note the heads and tales, you would most likely be able to see who actually did it and who phoned it in. The ones that feature long streaks of either heads and tails are the true notations, but we humans doesn't perceive this as random and will likely never make streaks as long as they appear in real life. Funny side note. Most shuffle system for music playlist are actually not random, because the truly random doesn't feel like it to us. Instead they are made less random to feel more random.
I'm reminded of when Team Fortress 2 first released weapon drops. Basically, every time you died, you had a chance of getting a random weapon. The percentage chance was set so that you'd get something every dozen deaths or so, and that's how it worked for the vast majority of people, yet some people were going hundreds of deaths without a single drop. Valve's mistake was that the game "rolled the die" on every death. Across millions of players, this meant that some unlucky folks ended up on the far far far end of the probability distribution. Put another way, the odds of rolling 10 six-sided dice and getting all 1's is about 1 in 150 million. However, if you have 20 million people each rolling 10 dice several times a day, someone will eventually draw that shortest of all short straws. Valve's solution was to change their game so that, after each drop, it picked a random number from a range that determined how many times you needed to die before you got your next drop. This created a hard limit on how long you could go without a drop. Similarly, I don't doubt that some (maybe many) of the people who take to forums to complain about things like the hit chances in X-COM are suffering from a psychological misunderstanding of the odds, but I also suspect that, in a game that sold millions of copies, it is not impossible that a few people did, in fact, miss a dozen or more 95-99% chance shots in a row. It's this sort of thing that makes me think developers using RNG should always include some sort of backstop to prevent the game from being ruined for whoever ends up on the highly improbable end of their probability curve.
Valve uses a system called "pseudorandom distribution" for many games. Most notably, DOTA 2. Whereas a "X%" chance to trigger was a true "X%" chance, they changed it so it starts as X% on the first try, and it becomes more and more likely to trigger the more it doesn't, and less and less likely to trigger immediately after a successful trigger. Basically, it "evenly" tries to space trigger events (based on the "base" % to trigger), and thus plays with our expectations (biases) of randomness.
@@francoiscoupal7057 It's a really cool system, originally created by Blizzard for WarCraft 3. And because Dota was using the WarCraft 3 engine, it ended up using it as well. So when Valve created Dota 2, they copied the system over.
@@SchemingGoldberg Honestly, it's such a basic system (applied statistical events manipulation) that claiming a a game company "invented" it is a very shaky claim. That they systematically used it in their games before another dev did, fair enough. Remember though that's it's not a "revolution", it's not "better" than using a true random chance element. It's deliberate fudging with randomness to level a playing field. (Think Mario Kart randomixation to give a chance to the least performing player).
One of my friends was one of the "shortest of short straws." I did the math, based on the then-known probabilities and time scales (it was something like 0.05% rolled every 5 minutes) and how long he'd gone without seeing something and it worked out to 1 in a million for that two week period (with an active player base of 2 million).
and yet, I still seem to just get drop notifications *at the time where I'm waiting for a respawn and I want to take advantage of the wait time to see how the game flow is going* . I die, I want to watch the kill cam to see where I go next, but I'm prevented of doing that because the drop system kicks in with a big "HEY! YOU SEEM TO BE IN A HURRY TO GET BACK INTO THE GAME! HERE! LOOK AT THESE NEW COOL ITEMS WE HAVE FOR YOU!" and it's a weapon of which I probably have three copies of. I mean, how "random" is "right in the middle of the wait time between two critical moments of play"?
I feel like the issue with % is that we generally don’t understand the idea of that % being rolled every time something happens. For example, just because flipping a coin has a 50/50 chance of landing heads, getting a tails on the first flip does not increase the odds for heads on the next one, the probability is still 50/50 and will always be that. It’s just that mathematically we *expect* it to land on each side equally, even though in practice that rarely ever happens
True, but all of that depends on the total number of "possible throws" that the chance is based on. Let's say that two archers try to hit their mark 100 times. Archer A hits the mark 70 times Archer B hits the mark 30 times. If the exercise is repeated, then A has a 70% chance of hitting, and B a 30% chance. It's an indication that A is more skilled, and therefore has a higher chance of hitting. This also means it's strictly possible that for the next 100 shots, Archer A misses the first 30 shots. What I'm trying to say is: in many games, we have no idea on what exactly the "chance to hit" stats are based on. Are they based on 100 simulations of the action? On 10? On a million? That naturally isn't really something the developers can make clear to the player while they are playing, but it's a factor of frustration nonetheless. For your example, while it is true that there is a 50/50 abstract chance for a coin to land on heads, in reality, that chance will be more or less divergent. If the coin keeps falling on tails, and after 100 throws, the coin in fact landed on tails 93% of the time (literally 93 out of a hundred) then we can no longer say the chance is 50/50. Obviously there would be something wrong with the coin (think of loaded dice, if that makes more sense.) Because the possible simulated throws of the coin is potentially infinite, the 50% chance to land on heads is really not an accurate representation of what will likely happen. All in all, it's about the information the player receives. If what I'm reading in the comments section is correct, then those "chance to hit" stats for many games are bullshit anyway, as they are either higher or lower in reality. Misinformation leads to frustration just as much as ignorance does.
well, as he talked about in the video some game developers tend to modify luck so that it matches people's expectations, like if you expect to land 1 of 2 shots cuz of 50% chance, they will guarantee you one shot of them so you can actually feel satisfied
@@RedFloyd469 I'm sorry but I think you have a wrong understanding of probability which is called the gambler's fallacy. It is a common misconception. Just as the original comment pointed out, there is no connection whatsoever between past and future outcomes. Let's say you flip a coin: Even if you hit heads 1 million times in a row, that doesn't influence the outcome of the next flip. To phrase it differently, the sequence HHH(...)HH is exactly equally probable as HHH(...)HT. The higher the number of tries, the closer you expect the results to get to their probabilities, but there is no guarantee that will happen for ANY number of tries. This seems counterintuitive at first, but it makes sense.
10:36 - "These cards are terrifying, game changing events that can completely demolish your team." As someone who has played Pandemic several times, that is so true it's not even funny. It's always really dramatic when one of those cards rears its ugly head. If it were more dramatic, the lights would dim and the card would glow green upon reveal, illuminating the horrified face of the unlucky person who drew it. The other players would gasp a little, and one player would ask, "Is it an Epidemic Card?" even though they already know the answer. The poor lad who drew the card in the first place would look up at the rest of their team, drop the accursed card on the table for all to see, and answer solemnly, "Yes."
My favorite moment in XCOM was playing with my friend. We were into the late game, only one death early on, main squad pretty beefed up. None were showing pyscikik promise, so we swapped out one with a promising member. Nickolas Cage was his name. He was getting up in skill as well. We entered the downed alien ship, fighting a brutal battle. We make it to a long hall way, three Cyber Disks show up. All of our shots miss. They fire, kiliing two. We fire again, not much effect. They fire, missing only on Nickolas Cage. The team is dead. We say F it, and charge with Mr. Cage. He kills one. All of their shots miss. He gets a reaction shot. He kills another. He takes a blast to the face, still lives. He kills the last. Mission completes, huge haul of resources making research possible that lets us crush enemies with (relative) ease till the end of the game.
Had a similar experience, at least in the backseat watching a good friend play XCOM (Enemy WITHIN, I believe). Turned a decently-lucky soldier into a giant mech guy, and from that point on there wasn't a damn roll he couldn't make. Never missed a shot, period.
The big thing about output randomness is it feels much better when the player makes a choice to take that risk, rather than just being forced into it. If you have to choose between option A that has a 50% chance of success or option B that has a 50% chance of success, there's only an illusion of choice. But if you have to choose between option A which has 100% chance of 1 success, or option B which has a 45% chance of 2 successes, and 5% chance of 3 successes then it becomes more of a decision. Not only is it better to have actual choices, but it puts more emphasis on when you choose each one. You might only gamble when it is hopeless anyway and that 5% chance sometimes saves the day. Or you might gamble when you're so far ahead that a loss doesn't matter as much and success allows you to feel even more powerful. It gives the players better control over how much they are willing to be hurt/helped by RNG.
Reminds me of a D&D session we had. I was helping my friend to make the best thief ever, helping him choosing the right perks, the right equipments and powers and magic weapons. We gave the character the name Chad. With all that combined, his thief became so OP he could hit anything on a roll of 5 or above. Turns out that, that whole night, he couldn't roll anything above 4! It has been years but to this day we still remember and laugh at the tale of Bad Luck Chad.
Feels like one of my newer sessions. Going through a dungeon with 20 AC and still being hit into submission way more than that mage with 12 AC... welp, that's what you get for trying to be a tank, I guess. And still other party members complained about them having bad luck!
Playing War with some friends, one of them used 30 armies to try to take a place occupied by 3 armies of another friend. The defender lost two armies, but kept the territory.
That reminds me of a battle in Axis & Allies many years ago. My opponent should have crushed me with his overwhelming force, but he kept rolling high numbers--one after another after another. It was the most remarkable string of bad luck that I have ever seen in my life. I won the battle with my tiny defensive force, and I awarded them medals for their bravery and valor in the face of overwhelming odds against them.
Quoting someone from one podcast I listen to: “I get angry whenever I watch Mark Brown. You can be good at editing, good with words or being right, but not those three at the same time; yet he does all of them every time”.
I'm surprised there wasn't a bit more discussion of Fire Emblem. As you said, they famously use 2RN (i.e., they roll two random numbers and take the average) to make up for how bad our brains actually are at understanding what "75% odds" means. But the most recent entry in the series introduced the Divine Pulse mechanic, which gives you a limited number of chances to go back in time, but the RNG stays the same. It stops you from resetting the level entirely when something goes wrong, and the RNG being fixed means you don't just try the same thing repeatedly until it works in your favor. But my favorite part about it is that it transforms the output randomness into a matrix of input *and* output randomnesses. The entire level's RNG sequence is fixed from the beginning, sort of like a level in Spelunkey or Gungeon. And if you understand how the RN system works, you can incorporate this randomness into your strategy. I never felt better at FE3H than when I was crit by the Death Knight, but used Divine Pulse to roll back the clock and changed my actions just enough so his critical hit became my own, which required: 1) knowing that his crit was the result of a low RN, 2) knowing exactly how many RNs were used before it, 3) coming up with Plan B to use up that many RNs so that the Critical RN is in the right place for me to use it, and 4) making sure that burning those RNs in Plan B doesn't put any of my other units in jeopardy who were safe in Plan A. It's a bit like counting cards, I guess, but pulling it off really made me feel like I had mastery over the game, and its number of uses are so limited that it doesn't really make up for playing the game poorly.
This is also interesting to me because before the double RNG check was introduced, the most common way the games were balanced was that most playable characters attacked twice during combat. It was supposed to just be based on the speed stat, but the vast majority of enemies were so slow that they were easily doubled, and in the last game before the new system, there were a whole bunch of new weapons with the Brave effect, which hit twice per attack, and thus mimic the potential "half hits" in Phoenix Point.
@@jeffd.683 Yes, but in there the RN changed so reseting until a positive result occured happened a lot. Also i believe Echoes didn't use the 2RN system because it was a remake.
What's interesting too is I found FE3H actually uses multiple strings of RNs, unlike in most of the previous games which only used one (some of the earliest games in the series used a clock RNG). First there's one set of RNs for combat actions (hit/miss and critical chances), and then each playable character has their own string for level ups. This means that the stats a character gains on a particular map are always set, and cannot be affected by rewinding time. If, say, Edelgard gains only a paltry +1 Magic on her first level-up in a map, you cannot rewind time and get something different, even if you take different actions leading up to the level-up. I think this was done to prevent the player from abusing Divine Pulse to get super-strong characters, though on the Normal difficulty setting you can use Divine Pulse with the Retreat option to get the same effect.
Players when they get the item that has a 1 in 50,000 drop rate: FINALLY I GOT IT I OPENED SO MANY CRATES Also players when they have a 1% chance of missing and they miss: WHAT IT IS SO RARE ITS NOT SUPPOSED TO HAPPEN!!!!!!!!!!
@@scottrauch1261 Except that most players will take 200 shots with 99% hit chance, and complain both times they missed that "it was 99% how did I miss?" then because those moments were pivotal in their strategy, blame "bad RNG" for their loss instead of a plan that didn't account for the fact that sometimes you're going to miss. Psychological studies have suggested that people feel bad results twice as strongly as good ones, so an accurate experience with 70% hit chance misses often enough for most players to feel like it's a 60% miss rate. People have deconsructed, reverse engineered and tested XCOM's randomisation algorithm and confirmed it to be fair, but there's still complaints and claims that it's horribly broken in spite of evidence to the contrary. Because "it feels like I never hit" when you're taking 60 - 70% shots and hitting 60 - 70% of the time but feeling the impact of those misses more than the hits so you think you're hitting 20 to 30% of your shots instead.
@@scottrauch1261 In which game? Because for XCOM, people have accumulated insane amounts of stats from multiple players who have hundreds of hours of tracked and verifiable results that have been compiled and analysed. Individuals have had results which are off from the normal "expected" values to varying degrees, but the overall stats are consistent with what should be expected when you're checking the data on a large scale. Of course, there's still the bug that makes it possible to miss with a 100% chance to hit and hit with a 0% chance (I've done the latter in multiplayer, it was hilarious). And it got worse in XCOM 2, where they claim to have "fixed" the problem, so now on the easiest difficulty, an attack with 94% accuracy has a lower probability of missing than one with 100% accuracy... I could be wrong, but I feel like they might be confused about what "fixed" means.
“Scientists have calculated that the chances of something so patently absurd actually existing are millions to one. But magicians have calculated that million-to-one chances crop up nine times out of ten.” ― Terry Pratchett, Mort
Xentillus IIRC they didn’t actually get the shot a million to one. I think the odds of that shot not being a million to one was a million to one. I could be wrong though, it’s been ages since I read the books
As I think about this, I think for me it comes down to the following: Good luck-based gameplay says "X happened, now how will you deal with it?" and opens up options. Bad luck-based gameplay says "X happened, too bad, so sad" and destroys options or even ends your game. It's interesting to be provided a set of new options, whether those options are you thinking about how the game works or are explicitly provided. And while occasionally removing some options to encourage players to not get into the same pattern forever might work sometimes, it generally just means you go from like 3 possibilities to 2, or worse, being railroaded into having to do a single thing, in which case your input feels redundant.
I think that's how Dysco Elysium does it and it works great. If you fail a roll you are not put directly in a fail state, but the dialogue continues so that you can try to recover (if you have heals) or even dig a bigger hole for yourself if you are not careful.
What you described is part of why I prefer playing XCOM2 in Ironman now. Its taught me to manage and calculate risk and makes my decisions more impactful. When I didn't play on Ironman, missing a shot meant I reloaded. Now it means I'm thinking in advance "if I miss this shot, how will I manage things to try and make sure no one dies." Part of that risk management can even involve choosing who I put into a risky situation. A Psi-Operative who has the cheat death ability? Take the risk with them basically every time. A rookie I've invested basically nothing into? You're going into the line of fire baby. A max-rank dude who I've dumped tons of bonus AP into? I protecc, stay safe friendo. It still sucks to miss a 97% (curse you 97, I hate you so much), but I feel like I've grown as a player when I'm able to assess where I'm at every step of the way and sometimes even abandon a mission that is going poorly rather than try and get the win after losing a character or two. Those setbacks can make the game even more enjoyable in the long run, because I have to manage the consequences of failure.
I don't think it gets redundant if at that point your reaction time or ability to notice what needs to be done get tested. But you present a good way of thinking about it
This is generally why, as a massive roguelite nut, The Binding of Isaac is my favourite roguelite ever made, and I think FTL is overrated. Isaac has ludicrous room for player growth and potential and (contrary to popular belief) essentially any run can become an incredibly OP gamebreaker in the right hands. FTL is mostly a game of math where the numbers you get for your equation are random. (this is not to say FTL is a bad game, it really succeeds at providing a Star Trek-esque strategy game and is still the best game I can think of for that specific experience. As a roguelite though IMO it kinda fails)
@@coffeedude I was more thinking of turn-based games where it's more about what set of more tightly-defined choices you make (i.e. selecting an attack via menu vs. having to manually perform the attack) since they tend to not offer as much wiggle room in terms of gameplay actions and allowing high skill to overcome a bad circumstance, and also because they tend to be the games that employ luck-based gameplay far more than games that require twitch reflexes, aiming skills, timing, etc. But you're absolutely right.
I believe the rule is that if the hero's plan is explained or otherwise known in advance, it will always fail. It only succeeds if the audience finds out during or after the execution, which is how heist films work
I mean, hitches in the planning are a staple of the franchise. For example from the first movie, they have to steal the emp because the casino discovered the vulnerability in their power system the thieves were originally planning to exploit. The theft of the emp goes wrong and leads to Yen needing a cast. That cast gets caught in the vault door during the heist. The fun of heist movies is that things never go exactly according to plan, but they find a way to work around the surprises.
15:00 Explaining to people that it's just as likely to get heads 500x in a row as heads and tails alternating for 500 throws is one of the funniest things
@Luke13139 if its specifically alternating, then the chance is the same, because its following a specific pattern, so the statistical probability is the same
Uh no it isnt. Unless you mean alternating as in HTHTHTHTHTHTH... exactly (which literally nobody cares about because literally nobody thinks HTHTHTH exact is likely at all) then the probably of getting alternating throws is much much higher than only HHHHHHH... For example with 5 tosses: HHHHH HTHTH HHTTH HHTHT HTTHH TTTHH THTHT HTHHT ... You see how there's gonna be a bajillion more sequences that alternate Heads and Tails? If I flipped a coin 500 times and i got all heads, that coin is rigged as fuck.
This is perfectly put honestly! Output randomness does tend to be more frustrating, and percentages in games tend to be lies to make you feel better (like the example you put where it says 90% but is more so 99%, it makes nailing that 90% feel better and when you loose it's not as bad as loosing a 99% chance to get something good). Output randomness is often better when in the favor of the player, a random chance to get something great always feels better than a random chance to get absolutely decimated by RNG and loose. Input randomness is by far preferred as well, as you mentioned! I believe a small balance of the 2 makes a truly satisfying game, input randomness with a little output randomness sprinkled in to spice it up (both good and bad, like the chance for a few shots to miss whilst also having the random chance for the ones that do hit to crit doing more damage or having the chance for the enemies shots to miss to which makes you feel great for them missing).
An important thing about statistics is that they converge towards the odds, so frequency is important. For example, if you'll only ever fire one shot that has a 70% chance of hitting it's an all or nothing bet. If you'd keep firing forever you would over time hit exactly 70% of the shots to the point where you could just as well see it as having 100% hit rate but 30% damage reduction. This maths holds up also for one-off events, but it's only really relevant when you can make enough statistical choices for the statistics to even out. That's why for example the solution in Phoenix point where you fire multiple shots with the same hit percentage becomes much more predictable and might feel more "fair" but less risky. So if you design a game with random events happening seldomly they will likely feel more unfair than random events happening often.
Kinda true. We are still very bad at randomness. If we fired 5 shots every time with 50%, and we fired that burst like 20 times we would be surprised when we miss all five shots at some point when we shouldn’t be, so I think it can also help to skew the odds towards the center a bit. It’s better to make that zero shot hitting much rarer, but also making all 5 shots hitting much rarer than to actually keep it at 50% for each shot because we don’t realize how after doing it so many times it’s not actually surprising that we missed.
For all-or-nothing scenarios there are solutions for that too, by modifying chances behind the scenes. One nice solution for critical hits I've seen was tracking recent critical hit ratio, comparing it to critical hit chance and adjusting the roll on next attack so outcome of crit ratio gets as close to crit chance as possible, while still keeping it random. This kind of mechanic could work well for likes of pvp games, where you want to keep system fair for both sides, but also keep the feeling of it being fair and random at the same time.
I think there actually was something faulty in iTunes, where it actually was built to have a greater chance to play a song from the same artist than any other song. This is not something that has been tested, so we can't tell if it actually was a thing or just placebo. But people haven't complained about this with other music players like Spotify or Windows Media Player, who hasn't stated that they have modified the randomness to avoid this kind of situation. So my theory, after experiencing this effect myself with iTunes before, that there actually was a built-in feature to increase the odds to play a song from the same artist. Some developer who probably built that in intentionally or acidentally, and people forgot about it. But with the new system that combats this, the old randomiser is basicallt overritten.
@@dinoactual yeah they _most certainly_ have some shady stuff going on under the table. How easy is it for an artist to say "hey i'll pay you $$ if you play our songs more often" and then they alter the algorithm? I'd imagine a points system would automate so much of that, (of course now i'm just scheming in my head)
SpaceMissile Hold on, hold on. I never agreed to these conspiracy theories. Spotify has probably done a similar thing in that they altered it so songs by the same artist play less often, to give people the feeling of more randomness when it isn’t actually random.
I heard at a conference once that its also important *when* in a game randomness has impact. Dying purely because of bad luck at the start of a game, not 10 minutes in, can even be fun. But dying because of bad luck one hour into a game and losing all that progress can make players quit, and as game developers we dont want that. So it could be a good idea to make it so luck has less and less impact the more you progress into a game, and make skill and strategy be more important in those later stages Assuming its not a game where you can quicksave at any moment and never lose any progress, of course
I'm not so sure about this. I played xcom with a friend through parsec and my whole team died because i missed almost every shot. At this point I'm thinking that the rest of the game will be exactly like this. Needless to say, i hate xcom and will never play any title in that series again.
@@prinzeszelda3650 Yeah, like if they land a critical hit, or just have a higher base speed and beat you to tackling, or they get lucky and use a bit of strategy, yeah it's pretty common to lose those.
"dying because of bad luck one hour into a game and losing all that progress can make players quit, and as game developers we dont want that" Yup, sounds absolutely right.
I think Ocean's isn't really an exception, because they don't fully _explain_ the plan to the audience beforehand. In many cases, it _seems_ like things are going sideways, only to be revealed further on that those setbacks were _all part of the plan_ .
Fire Emblem tends to have "2 RNs" in most of their games. It means it rolls the number twice and avarages them out. So let's say your hit chance is 90%, that means you'd have to roll two numbers which their avarage is lower(or I suppose equal to) than 10. You could roll a 1 first and then let's say 22, for the same 90% chance check, and it'll hit even though you rolled a 1 first hand.
@@ryanoutram7059 Xcom teaches us that a 70% chance to hit is terrible odds. You play any mmo and you'll often get a weapon or item that has 1% chance for extra fire damage or something. Which is usually pretty good. That 1% extra means over the course of your session you'll see it pop quite often because we roll the dice a lot in those games. Xcom is in the perilous position where you are destined to fail at some point because we roll the dice very few times, so the few hits on a 30% have a greater impact on the gameplay than they should while 90% chance to hit always feel like you're playing with fire. A whole shootout can just straight up end or be won on 2-3 dice rolls. I've never gone back to xcom because of that. I don't want to spend the rest of my life making backup plans and alternative strategies for a game that's supposed to be fun. It just became work after a certain point. Replaying any mission feels like you wasted time, losing men feels like you're weakening yourself for the end game. There's no benefit to playing the game ironman. There is no incentive to work through your bad luck. In which case, why not just cheat the whole way through? Why bother playing at all?
@@Madhattersinjeans That's precisely the reason I play Xcom - yes it's chance based, but you make your own luck. You have to react and adjust your plans, and outside your battle you build up your team and base to mitigate that risk.
also, those events that negatively surprise you (missing a 95% chance shot) tend to linger in your head far longer than the opposite (the opponent missing a clear shot or you getting a 10% chance shot). When that happens you get happy for a while but move on, but when you're unlucky, you just have that anger stuck for much longer
Humans tend to remember negative events for longer, most likely to try to avoid repeating the mistakes that lead to them. That’s another reason being unlucky is frustrating, you can’t avoid it.
I remember the fact in XCOM I have two missed 95% and (I think) 99% I hit. Can't tell you the number hit, nor 90+ hit, or missed shots from the aliens.
Actually, no. No one will ever try to make a 10% chance shot. Maybe in some VERY dire situations when playing ironman (no save run) - then, if it hits and saves the day, it will linger for much longer and stronger than any 99% miss.
I feel like this is one of the reasons Hades has felt so good to play. There’s a lot of random elements for each run, but they usually consist of a choice of two or three (room rewards, boons, etc) from a random assortment you become familiar with. And because there’s such a large assortment of different upgrades and boons, you can keep finding different ways to succeed no matter what your “luck” is (similar to what you mentioned with Slay the Spire) so that you’ll keep sticking with that run to see how it goes.
It's the difference between playing Ludo and backgammon. My favorite games always have aspects of luck in them, they just include a lot of methods to manipulate the odds as well.
A part of that is because you're almost guaranteed to get stronger some way or another, albeit more damage, more synergies, more support items/boons, etc... (Which is a good thing, because otherwise the RNG just becomes frustrating at times) And this is why I find Binding of Isaac: Afterbirth to be so frustrating. There is no guarantee to get stronger in any way and it feels like you can miss way to many opportunities to get items which might not even be good to begin with. Even though there are a lot of DPS increasing, synergising and good support items, there is simply no guarantee that you will get any of those and then you just end up with a frustrating underpowered build.
Merkules Even more so because once you’ve collected a boon you are more likely to receive booms from the same god, which will usually improve your previous one if you do choose.
About randomness in role-playing games: tabletop RPGs inherently live and die on output randomness. The effectiveness of every action a player character takes--and most actions from the opposition--are determined primarily by the dice they roll and the numbers that come up. A 2d6 weapon in D&D is often considered better than a 1d12 weapon, because you're more likely than not to hit a seven. Players lean towards strategies that have consistent results, and many systems gain and lose fans based on the odds that an unlikely success or an unfair failure will come up during play. Why is this important? A good dungeon master can make or break a game, even in the most or least popular systems, and many of the best DMs have learned to preserve player engagement with a system often referred to as "failing forward." Failing Forward means that the player always has a chance to succeed, but the worse their roll the more severe the consequences of that success. A good example is trying to bypass a locked door: a perfect roll allows the party to slip by undetected, but a failure means that rather than failing to open the door entirely, the act of opening the door triggers an alarm. Like you mentioned in this video, the output randomness from a roll of the dice becomes input randomness for the following turns. I like to think this same principle can be applied to game design. By merely complicating the player's current plan rather than bringing it to a halt, we can retain player investment while still using chance to diversify the types of strategies our players have to use.
I think Streets of Rogue does this quite well. For example, failing the Threaten check causes an NPC to attack you, but you have the option of either running away or attacking them.
Well, 2d6 also have better minimal damage - the range is 2-12 and not 1-12. Overall it is more complicated than this because you have various feats that work with damage dices but that's a different tale. Also, DM can mitigate a lot of randomness. Lets go with the simplest example - players are searching room for a secret passage - that passage is under a rug as a trap door. DM can ask a question "Describe what your character is doing during this action" and if they described "I look under the rug" DM can say "You found it, you don't have to roll". This way it is IMHO more rewarding as a player - you succeeded just because you are clever.
This is false actually. The D&D-like (and more precisely "modern D&D"-like) TTRPGs are indeed worship output randomness like the holy cow. But it definitely not true for all TTRPGS. Old-school D&D-like (nowdays called OSR and retro-clones) TTRPG preach the mantra of "if it comes to diceroll you botched the planning part" and require a DM to always provide a roll-free solution for any situation (even if this solution is Joestar-special running away). The narrative-TTRPG (Fate, PBTA, etc) often provide you with a way to sent dice packing: if you got yourself in good enough spot, dice can only force you to spend some resources to overrule them. (They also almost always implement the Failing Forward mechanic) Finally, there is a Polaris "something about North and Tragedy" which doesn't use RNG at all (and uses complex bargaining-like system between players instead). I'm actually is in agreement with you about the appliance of this principle to game design. Just wanted to correct you opening paragraph.
Yep great points... at some point someone just needs to program an RPG where you just roll the dice all the way through the story to emphasize that RNG is making too much of the decisions... if all of your choices are Output RNG then are you really making choices? Might as well just Roll for the very decisions you make Attack or Take a Nap is how effective some of these rolls can be? Yes I know that this is offset by rule manipulation, saving throws, and stat/skill modifiers but there is a reason that you don't start dungeons off fighting a Dracolich and a Goblin as a final boss. It is fundamentally backwards to the idea of progression. My Level 1 Druid took down a Dracolich the first time I played... people are going to be... whatever...
Another nice thing about Dicey Dungeons (other than subverting dice from the usual output into input rng) is that it features numerous abilities that manipulate dice whether it's splitting, copying, or even flipping them. This give the feel of 'cheating luck', that goes along with the narrative theme
The beauty of that is that the plan isn't revealed until the end. It gets into the hidden information alternative to randomness. Suspense is kept because events seem random until the end when it's revealed to be according to plan.
Didn't an Asian guy get hurt by another accomplice and then the bandage got stuck somewhere and said asian guy got almost blown up to smithereens hadn't the bomb remote control not have dead batteries, But he was lucky enough that it did causing mere seconds delay while one of the guys with the remote control change the batteries with fresh ones?
@@highchair208 ^Cause their plans totally work out exactly as they were made... Not. Or are you saying Finn and Rose being caught and imprisoned cause they got betrayed by the wrong master hacker was part of the plan?
With both Ludology and Deck Building (edit: and Pandemic!) references in one video, it almost makes me wonder if GMTK should consider a video on "What can video games learn from board games?" Bonus points if you coordinate with Geoff Engelstein.
I’d like them to learn from an actually-punitive RNG game like Blood Bowl, where failure is common and punished to the point of hilarity sometimes, and there are no RNG skews you do not pay for. You can’t play it for long without either ragequitting or reaching a level of RNG Enlightenment where either you understand risk and roll no dice that are not material to your success, or you do not care about risk and laugh as hard at your six vampire TV1000 team literally taking itself off the pitch in turn two as you do the other guy’s legendary Wardancer tripping and instantly dying on a sprint to the endzone. You don’t need to coddle a player when you teach them instead. And sometimes the lesson is that the dice may hate you, but it’s nothing personal.
That's what I was thinking as well. I'd be interested in Mark's take on board game design and what is its relation with video game design. I'd also be curious to see his take on adapting board games to digital.
Reminds me of a very old rule of thumb about stuff like hitboxes. While it seems like you'd want your hitboxes to be 'perfect', in practice (especially when computing power was limited) hitboxes always end up being vague approximations of the real situation. The rule of thumb then? Always bias hitbox inaccuracies in the player's favour. Enemy hitboxes should be larger than the enemy appears to be. Conversely player hitboxes should be smaller than the player. (obviously this line of reasoning breaks down in multiplayer contexts, but you get the idea). The thing is, you can make the same argument about 'randomness', and the same overall logic applies; Players are far less resentful of things that work in their favour than things which work against them.
and gacha games (AKA mobile cancer) has similar systems too. they bring better drops when you come back to the game after a hiatus. output RNG is just the worst.
That's the problem with trading in general, you need more loot that's irrelevant to you but good for someone else. The smart loot was added to fix the problem of not being able to trade.
On "there's never been a good movie where the heroes explain their plan and then it goes off without a hitch", the entire heist genre actually did exactly that for quite a long time. The idea that the plan should be disrupted is a surprisingly recent innovation (which is probably why the Ocean's Eleven remake seems to be the example everyone is jumping to).
I haven’t watched many movies in a good while, but even in those old heist movies, wasn’t there usually still a small margin of error where they had to at least switch from plan A to plan B? Maybe not full disruption, but just a small hiccup that they had a second plan for?
@@aqwkingchampion13 My memory of the original Oceans 11 and original Italian Job is that things went very wrong very late. But they certainly didn't all go right. I think I could find dozens of movies that prove that not all heist movies used to go right though.
Even if the plan goes well, we usually see that happening WHILE the plan is explained to us, sometimes as a montage. The idea is that any plan explained in advance is going to go wrong
15:18 To elaborate on that the Fire Emblem series has actually had many different RNG systems over the course of the series. +In the early games, mostly before they were released in the west and the creator Shouzou Kaga was still involved the games operated on a "single roll system". Meaning the game only calculates a percentage roll once and takes that as a result. The number the game gives you is the actual chance you have of missing, landing a critical, ect. +After Kaga had left and the games start moving to the west the games moved to a "double roll system", where instead of the game rolling the percentile dice once and taking that as the result it rolls the percentage dice twice and calculates the average of the two rolls as the result it uses to calculate. For example if the games says you have a 75% chance to hit it will roll two numbers between 1-100 and lets say they're 68, and 82, which added together and divided gives you a 70 which the game uses as the result for your attack. The result of this was the number the game told you was actually incorrect and was always more exaggerated than it looked. Chances that where closer to a 50% chance were more accurate while high chances were effectively higher than they were and the lower rolls were even lower. A 90% is more around 99% and a 10% chance is more around 1%. This was done to help create a sense of stability to the game so that if the player was told their chance to hit was high there would be less of a chance RNG would give them those random rolls that screw them over despite that they did everything right. This system is often heavily criticized as hiding information from the player and tilting the games systems even more in their favor. Edit: It's comes to my attention that the double roll system primarily applied to hit rates rather than secondary rolls such as crit rates which function off of single roll. +In "Fire Emblem: Fates" released in 2015 and "Fire Emblem: Shadows of Valentia" in 2017 tried to marry the two systems into a "hybrid roll system". Its formula is much more complicated than the past two systems but what it effectively did was for chances that were 50% or below the game functioned on the single roll system, this giving lower odds to have a greater chance of surprising you. Odds 51% or above functioned on less exaggerated version of double roll. The number the game would tell you wouldn't be your actual odds and would actually be higher than the listed number but less so than in the double roll system. Less 90% is actually 99% more 90% is 95%. This was done so more likely events were more consistent and could be accounted for and players knowing and engaging in the games systems would be rewarded but without as big of a player advantage as the double roll system game. The hybrid roll system would only stick around for "Fates" as the next main title in the series "Fire Emblem: Three Houses" released in 2019 would return to the double roll system. Edit: Grammar Edit #2: New information added
Why don't they just put a "90% - 99%" on screen, and roll a dice that's actually between those numbers. So when it rolls 90 they can't complain and when it rolls 99 and still misses, well they might think it's the "10%" that missed, even though it was the "1%" chance of missing.
The "double roll system" doesn't make sense as you described it. If you averaged two dice rolls, a 90% or greater would be less likely, because out of all the possible numbers you can roll there are more that are outside the success range (90-99). If you roll two numbers and average them together, even if one of them is within the success range the other is probably not, which would drag down the result. Likewise, extremely low rolls would be more rare for the same reason. It would generally be pulling all results towards the middle. Like how a single 1-6 die roll can result in any number equally, but TWO 1-6 die rolls more often result in numbers around 7. This is like saying you roll two dice, so that you're more likely to get a 12 (compared to a single 1-12). It can't work that way. I haven't played the game, and the only description I know is in this video and this comment, so that's the information I'm working from. But it sounds more like an "advantage" system like in D&D 5, where there are two rolls and you just take the better roll... or if low extremes are also more likely, you'd have to take the better roll if it's above 50% or so, or take the worse roll if it's below.
@@All4Tanuki Chess is one of the most popular games, and is growing in popularity, exactly because people are starting to lose the mentality you display here: a 12 year old thinking it's edgy to call board games boring because it's not in a videogame.
"optimizing the fun out of games" Basically my problem with every multiplayer game at high level. I don't like it because it becomes a contest of being the most efficient at executing the same optimal strategy.
This is why I quit League. I'm just keeping my account so I can rip champion and skin models. I don't see what there is to actually achieve in games that are heavily multiplayer dependent.
Thats why I typically like to try and find a strategy that specifically beats that one. Even if it isn't "Good" if it beats the #1 its my favorite way to play.
@@battleland7263 and watch as pros finally showed everyone how to abuse it, followed by everyone using it in all situations(even if they lack what made the pros successfully run it)
I feel the pain, but the fact is, you can pull a lot of victories using very wild strategies, because your opponent has no plan for it, as they are expecting you to use the "optimal" one.
Randomness is fun when you just passed a save point and engaging an enemy/event that presents just enough of a challenge that if randomness makes it harder, you just adjust your strategy and make up the difference. It's not fun when you're on the third stage of a final boss that is already immensely overpowered and they use a multi-attack that lands with critical damage and all of a sudden you've been put in the tailspin of having to heal every turn. That can get real annoying real quick.
The talk on probabilities in civilizations was really interesting, glad you mentioned it. It's funny how people think that with a 1/3 chance, if they lose the first two, then obviously they should win on the third try. Although, I remember thinking that way as a kid myself.
Incredible insight, it's reminding me of how screenwriting is often taught, i.e. something must be unexpected. Whether the unexpected elements come in at the front end of a story or at the back end, it's always a reliable way to give your story gravity. Additionally, the part about information flow is very reminiscient of dividing the narrative into scenes or acts. Making sure new information is drip-fed is a fundamental element of successful storytelling. Surprises are necessary to keep audiences engaged, and the same can be said of the randomness you discuss.
An immensely good video, one that I find myself returning to for.. probably the 4th time now. I went from making games in my spare time to actually studying game design. Much thanks to your immensely informative and engaging videos. Thank you
Output randomness: Everyday there's a tiny chance a new GMTK video is published. If there's no video, no matter, I can wait. If there *is* a video, YEAH!
It's input randomness. You act *after* the fact whether there's a new video. Also, releasing a new video is anything but random. There's no chance he can make 1000 videos a day. It's also unlikely that he'll wait years before releasing a video, unless he quit this job.
I've got to say, the game that taught me about randomness and risk mitigation was Blood Bowl. The core mechanic of the game being that you and your opponent alternate turns, but your turn ends abruptly if anything you do goes wrong. So you've got to perform actions in a sequence to maximize upside and minimize downside.
The turnover mechanic is brilliant in BB. I also think it's worth noting that BB allows players to buy a small number of precious re-rolls which can be used to counter the worst rolls, and all this combined with a full roster of units on your team to control -- and all the chaos that can occur each turn for both your team and the opposing one -- gives players a boatload to think about each turn. Plus it can be hilarious even when we're really unlucky. I think that's often helpful with games that chaotic from the RNG is that they're sort of funny and generate lots of amusing stories to tell.
0:58 "Do we just like luck when it lands in our favour, and hate it when we lose? No." 17:40 "We found that if there's randomness where you're expecting something bad and then you get something good, no one ever ever complains."
8:53 "carefully tuned output randomness" the importance of this tuning, especially in competitive multiplayer games, cannot be stressed enough. One example of bad output randomness that comes to mind is the random crit system of Team Fortress 2. In theory, it was created to break stalemates and give players of any skill level a fighting chance, but in practice there are better mechanics that solve stalemates, and it ends up helping the players that are already doing well (since your crit chance is increased based on damage done recently). Uncle Dane's video "Remove Random Crits From TF2" goes a lot more in depth on it, and even touches on the fact that even the player dealing the crit can feel unsatisfied about it (I understand, though, that 17:26 - 17:45 is about single player games). Also, it's worth noting that competitive games already have an irremovable random factor: the other players. The way a player acts, reacts and how they expects other players to act has enough randomness. Using poorly implemented output randomness here can severely undermine the players' skills and mind games.
"In theory, it was created to break stalemates and give players of any skill level a fighting chance" - on this note, you're actually incorrect- a developer message on one of the maps states outright the ACTUAL intent was to reward those doing well. "This creates those rare, high moments where one player goes on a rampage." "In short, the better you do, the more likely you'll continue to do well." This was the developer intent, and sadly, it works out exactly that way. And it's really not fun. For either side.
@@NimonoSolenze You're right, I was misremembering it. Maybe I read somewhere that it had that function (on top of creating "'rare' high moments"), but the dev commentary says otherwise. Which is a shame, because then random crits have one less good reason to exist. It's a mystery why they weren't removed after all these years.
League of Legends is rumored (I couldn't find devs confirmation) to have a biased crit chance that helps keeping random system fair - adjusting each subsequent roll based on your no-crit/crit streak so overall it adds up to your actual crit chance as fast as possible: if your character has 25% crit chance and first attack doesn't crit, it'll up the crit chance with each subsequent non-crit until crit happens, then reset; doing similar crit in reverse. It still keeps crit system random, but at the same time pushes damage output much closer to theoretical calculated value over much shorter period of time.
I was enjoying this video so much the end snuck up on me and now I'm kind of bummed it's over. I'm 4 lines away from 2 pages of notes from this video, there's so much value here thank you so much for sharing it! I can't wait to dive into more videos :D.
rain world absilutely deserves an honorable mention here! the world itself is static and meticulously crafted, and the mechanics are very minimalistc. but all other creatures, most of which concider you a tasty morsel, and some of which you can eat, are all free roaming AI with vastly different behaviors, temperaments and intellects, and they fight over territory prey and surviva. .. or your body. things never play out the same twice, and you are always on your toes. that one really opened me up to the mindset of hardcore gaming
(7:22) "pre-luck" and "post-luck" feels like better terms. "input-randomness" sounds like you're inputting something into a randomiser, and "output-randomness" is the randomness you get from the randomiser. Both "pre-luck" and "post-luck" would be the result of said "output-randomness".
@@Liggliluff Yes. from a programmer's perspective, we (are supposed to, in computer language) treat 'functions' and 'methods' i.e. getting a random number as machines, so input and output are bad analogies in this case.
@@jr.p.8429 So because this is about game design, where you have developers; avoiding input/output is a good idea in this case. Plus people are more familiar with pre/post than input/output. We say prehistoric, pre-launch, pre-workout and postpone, post-sync, post-event for example, not input-historic, input-launch, input-workout, and output-pone, output-sync, output-event. Pre/post are prefixes ("input-fixes") well known by the people.
@@Liggliluff well yeah.. because "pre" doesn't mean the same as "input" and "post" doesn't mean the same as "output". your input your credit card into the ATM, and the ATM output the money. you don't pre-put your card into the machine and post take the money. Same goes for the luck thing, luck doesn't equal randomness. pre-luck and input-randomness doesn't mean the same at all. pre-luck/post-luck doesn't even make sense in my head. luck is a result, it's not something you can have before (pre), or after (post). It's by seeing the result that you know if you were lucky or not. randomness is much better than the word luck in that aspect. Randomness is not the result but what lead to the result. Throwing a dice give you a "random" number and getting a 5 would be one possible output of the randomness. and THEN, whether you wanted a 5 or not, you know if it was a lucky throw or not. Input-randomness and output randomness make more sense to me. Although "input" sound a bit weird because like you said, it almost sound like the player had something to do with it. But It is still better than pre luck which sound like the user "own" luck in a tangible way (a luck stat maybe ? But even then, it would only be a variable in the randomness, not the randomness itself)
Sometimes, developers can make something seem random when it isn't, or only make it partially random but have clear trends and tendencies. In Freedom Planet, almost none of the enemies and bosses behave randomly - when it seems random, it's usually the enemy doing something based on your own movement. Spade, a miniboss in the Sky Battalion stage, appears random at first, but with practice, he can be completely controlled. The final boss, Brevon, has a final phase that is only random on Hard difficulty. He seems random on lower difficulties, and still kicks your butt, but you slowly pick up on his predictable pattern and learn to use it against him. Final Fantasy Tactics has plenty of randomness, including zodiac signs, hit chances, enemy secondary skills and equipment, and random encounters. However, the AI has some very clear tendencies that you can learn with experience, and a seasoned player can often exploit these tendencies, setting up situations to strongly motivate the enemy to walk into traps. Chrono Cross has your typical JRPG steal system, with a few key differences. You can only steal once per fight, and you aren't guaranteed to steal something, but every battle - even bosses, even the final boss - can be fled in order to start fresh from the beginning. This means you can start fights with a steal attempt, and flee until you get the item you want.
Reminds me of some enemies in Slay The Spire which just make a sequence of predetermined moves, and repeat, in a non-random cycle. I wasn't ever good enough to take this much advantage of it, but I presume it's the way the game intends you to survive when you use the relic that hides enemy intents from you.
No games are purely random. In theory, it's possible to manipulate all RNG in all games, because of how computer works and how a "random" value is generated (Which is actually a bad term, the correct is Pseudo-random). It's possible to manipulate what stats your Pokemon will get and stuff like that, if you know how to.
The chrono cross thing seems tedious more than anything. Why even make it possible to fail if you can just retry infinitely until you either get it our give up from boredom.
When playing RPGs, I like dice rollers generally, but when I'm making a roll I care about, I do find I can accept failure better from a dice than from an app. So, I use dice rollers when DMing, and dice when playing.
@@takatamiyagawa5688 Most cheap dice are badly balanced. They're manufactured by the injection mold process, which often leaves air pockets in the final product. The only dice you can count on to be as unbiased as possible are dice which are carved down from a solid material, but those are a lot more expensive.
Another awesome video! I'm so glad you mentioned Hearthstone and esports in there, because that is a perfect example of a game that I LOVED the randomness - both input (card draw) and output (cards with varied effects). Some of my favorite memories in that game involved Yogg-Saron, which would cast a completely random spell with random targets for every spell you'd cast in the game so far - with often hilarious effects that could instantly win or lose the game, no matter the current state otherwise. ...until I got highly skilled and competitive at it. Then even minor sources of randomness became pure frustration, to the point of screaming at the screen and slamming doors when it "always" gave me the "worst" draws or outcomes and "stole" the game from me. Sometimes this was actually the case, but usually it just felt like it and left me constantly wondering if I made the right play and got screwed, or really wasn't as good as I thought.
I've always looked at randomness in general like a spice in games. I think you have to use it sparingly, it can't be the meal. A good chef knows the right way to season something. When you get it just right you get something special. I think games like Diablo 2 used it in a way that stimulated a lot of people just right. When you get it wrong you get an inedible mess. Like a F2P mobile game that just says "make it random" at every step in place of actual considered game design and satisfying progression.
You're missing a *massive* part of randomness here: the minimum possible number of bad rolls it takes to have irrecoverable consequences. The problem in XCOM isn't the output randomness vs input randomness; it's that one bad roll can have game-killing consequences. If an enemy had 100 starting heath and the output randomness on a shot is 10-20 damage, that's an adaptability test -- especially if sampled without replacement (or with delayed replacement, or any one of many such systems). But instead you end up with enemies with a base health of 6 and a sniper with a 99% chance of 12 damage but a 1% chance of 4 damage, against an enemy that has a plausible chance of a 1-hit crit on your units. So one or two bad rolls -- with no chance to adapt between them -- can cause huge problems, which is why it feels so bad. This is the same reason that FPS players are fine with randomness in the firing of automatic weapons but not sniper weapons -- over 20 bullets from an automatic the expectation is still narrow, whereas a sniper that can just randomly jam or completely whiff feels terrible. Pandemic is an interesting example because the epidemic cards have a 2-bad-rolls-loss possibility, if you get unlucky enough with what's on the bottom of the deck. But the investment in a single pandemic game isn't that high, so it feels fine to lose to that randomness occasionally -- you can just play another round. Whereas in XCom that 2-bad-rolls can come after playing a save for weeks. (If mission consequences weren't so devastating on your team in XCom then it'd be fine there too -- you'd lose some unavoidably, but you'd pull together for the next one. But that's not how things go, so they just incentivise loading a save instead.)
The problem with output randomness isn't output randomness. It's fail states. Imagine XCOM, but instead of "70% hit chance" you have "30% suppression chance" so 70% of the time you deal damage (same as now), and 30% of the time you don't deal damage, but apply a debuff to the enemy that makes them weaker on their turn. The fact that missing is no longer a pure fail state changes the dynamic. By changing the fail state to a less beneficial success, you're making players feel more agency to their actions.
@@revimfadli4666 The idea is based on an old tabletop wargame I barely remember playing almost 30 years ago. Don't know the name, but when you attacked enemy units, your shots either inflicted damage or suppression, with both having value in different ways.
An iconic form of Output Randomness: The dreaded/beloved Random Crits in many video games, such as the often hated on ones in Team Fortress 2. One Crocket can wipe a quarter of a 12 man team or more if placed well, but many players will get easily tilted or even enraged over this even if there's only a single kill made out of it. Now I'm gonna hide, cause speaking of said crits often sparks controversy borderlining a war.
Random crits are especially frustrating in TF2 because it's a game built around player skill. Whether you miss or hit a shot depends entirely on your own aiming and the other player's dodging, but you have no control over a lucky crit that randomly ends your life.
@@katerwaul9574 On top of that, contrary to what the video says about randomness in multiplayer games being meant to lessen the value of skill in a game so everyone has similar chances of winning at all skill levels (something I vehemently disagree with being a good design choice), TF2's random crits were designed around rewarding the skilled, as stated in-game by the developers. They wanted to create a sense of specific players being able to go on rampages, so they made crit chance go up the more damage you're doing over a specific period of time. Crits do high damage, especially crit grenades, stickies, and rockets, so crits with those (especially the rockets) tend to snowball into more and more and more crits until there's no more targets to hit. Uncle Dane has a great video on random crits in TF2: th-cam.com/video/WHvwijT2ss8/w-d-xo.html (yes I am passionate about random crits in TF2. yes, i do believe they should be removed BECAUSE of the explicit intent to turn skill or even just random chance into a snowball team wipe. it's not fun for anyone.)
@@katerwaul9574 I mean, Crits are BS but it's not really true you have no control against them. You still need to, you know, get hit by them in order to actually die from them.
This is nothing. In D&D 5e if you play as a crit oriented paladin you can one shot creatures that should rekt a party of four players. You have than one in twenty chance of evaporating almost everyone on your level. It is more powerful than crit rockets in TF2 and can make a deadly encounter a piece of cake.
@@GrimmerPl As someone who is a DM, has played exactly this and made a Paladin/Monk One Punch Smite character, I am proud of you for reminding of this. This is indeed fun. Additionally, the Sneak Attack of Assassin Rouge's also work.
One counter to the dice argument I’ve seen: when watching a stream of Dicey Dungeons I noticed a lot of people putting extra stock in rolling a six and getting annoyed when the streamer didn’t respect a six roll even though the function of the first two characters are to make you value high rolls and then tell you a six is just as rare as a one
I'm pretty sure the second character values *low* rolls (assuming we're talking about Thief here).. though yeah, having a build which can manage both low and high rolls is very good. (Though Warrior's Combat Roll basically does that for you, but to be fair - he's the easiest character.)
15:33 This is where the math "knowledge" of the average player sadly stops. 33% chance to win, I lost 2 in a row, I should win the 3rd, no matter what. They'll throw the mother of all hissyfits if they lose, but we know for a fact they won't bat an eyelash if they win five times in a row with that same 33% chance. Explain that brilliant logic to me...
The problem is probably that real world parlance (how people actually speak and what they mean) and mathematical probabilities (how the numbers are calculated and what they mean) are an entirely different thing. Games express mathematical probabilities of more or less independent events, but players think in terms of success rates. Hence, mathematically speaking 33% chance means that you can fail thousand times, and it is fine, but in real world parlance that is completely wrong. You cannot have 33% success rate and miss thousand times. I think it is games that get this wrong. They should speak the same language as players.
@@janimiettinen74 Your argument is that math is wrong about math. I think it's more accurate to call the players ignorant instead. They do not understand math. Period. If they did, they'd know that a 33% success rate means even after infinite attempts you are NEVER guaranteed to have succeeded. In fact, every time you try, you are more likely to fail than not fail. Period.
@@dontmisunderstand6041 My argument is that "30 % chance" is ambiguous expression. It can be understood in two ways. "30 % chance" is not exactly a mathematical expression because it does not specify in any way what that 30 % chance pertains to. Hence the ambiguity, therefore the misunderstanding.
It's also possible that he was a point that he mostly just has out of hearsay, a fact that everyone knows, but he's not going to put in a few hours of combing in order to find that one patch note that confirmed it.
Wow, phenomenal video. Thank you for all your effort in researching, writing, organizing and presenting this info. The result is both accessible and very understandable, without wasting any of the viewers’ time. Bravo.
Certainly a thing that helps when engaging output randomness is allowing reasonable ways to trade out a chance for a 100% surefire thing. I feel like Xcom would have been less frustrating if more hardflanks became 100% chances to hit more readily. The mario and rabbids figured this out by mostly doing round values and getting 100% chances vs enemies without cover against you. Also holycrap sorting out good success rates that then end up being foiled by ANOTHER chance on a chance you can't readily see, like snakemen in xcom2 dodging, turning a hit into a grazing hit, and throwing your plan out the window. Though Xcom generally is just an awkward case with all of this because it wants you to be so aggressive yet your ability to react to new threats feels generally too low with how fragile your soldiers are.
I played XCOM for a while, it got really frustrating to play. The game constantly pushes you forward but punishes you for moving forward, your soldiers have way too low an accuracy (to the point where you're relying on 40-50% shots regularly), the enemies always get to move as soon as you see them and your soldiers go down fast and hard (which is especially frustrating with how unreliable their accuracy is).
@@Nukestarmaster yeah I agree. The mechanics are much more punishing for playing aggressively because of that. It probably would feel generally better with better accuracy (or just use the mario rabbids model, that one would make flanks godlike and thus encourage having brutal fast flanking teams all up together) or have your soldiers durable enough (or means to delay) so that you have a reasonable chance to deal with multiple pods before deaths roll in. The timer was certainly a poor way to force fast play with how the game was setup.
Heh, i was fortunate enought to be able to play 2 on PC and having read about the timer beforehand the first thing i did before even starting a campaign was find and install a mod that removed it.
The Second Wave option that I think was called Absolutely Critical in XCOM EW changed flank shots to be 100% crit (not hit), which made the risk/reward involved in moving into dangerous positions to flank enemies much more interesting (40% just feels inconsistent).
I was watching my dad play Xcom 1 and the “Last” enemy in the kill all mission he was on was the Stealth squid thing(can’t remember the name) so he put all his units on overwatch and some poor thin man had dropped in the circle of death and instantly triggered all six overwatches at near point blank range, taking a total 20-25 damage. And to help those who don’t play Xcom understand the intensity of this event. Thin Men have only 3 health and Dad had endgame level gear.
That's what I dont like about the new xcom remakes. In the old xcom, they only spend as much time shooting overwatch as they need, and can shoot more than once
@@sol2544 Or someone drops way in the back and exhausts all your overwatch shots, which all miss. Afterwards someone moves at the front. XCom is just so screwed up... a thousand badly thought out mechanics. What I want to point out is that players report their bad experiences, like missing 5x 90% shots. Which then gets excused by "well it can technically happen". But XCom RNG is screwed up when it works in favor of the player, too. I had so many maps with about 60% hit chance, but 95% of my shots hit. Several maps in a row. These are just 2 of XComs thousands of issues. When a game is bad, it's usually bad in several departments. XCom is no exception.
Man The idea of luck only being satisfying when you expect a bad thing to happen, but a good thing happens makes me reevaluate all my rpg design choices Edit: thanks for this video I didn't know I needed
Mutant Year Zero: RtE is a good example of a game ruined by a bad "output luck" design : you have WAY too few % rolls per turn and each of these rolls has WAY too much consequences - worst of all, the kind of consequences you can't repair or at a very high cost which puts at risk the following of the player progression. Thus, perfect fights can be ENTIRELY ruined by only one bad roll, which leads to terrible frustration and having to reset all the time, instead of having to improvise to the situation.
In every single encounter in Mutant you can achieve 100% to hit. I rarely went for a shot that wasn't 100% to hit during my playthrough. And if I remember correctly I had a lot of fights where by a proper setup I had zero RNG elements - just a clean execution.
14:06 This was my main thought throughout the video - unless you’re playing a one-shot game, there’s no meaningful difference between “input” (i.e. anterior) and “output” (i.e. posterior) randomness. I think a better way of thinking about randomness is to separate it into: - “Player-Random” - the game includes actions that can be taken by a playable character which do not yield 100% predictable results. Example: moves with less than 100% accuracy in Pokemon; drawing a Development Card in Settlers of Catan. - “Utility-Random” - where a game distributes utilities (i.e. resources, money, weapons, etc) that a player can’t opt into receiving (i.e. discounting when a player chooses to open a loot box), and some facet of the game’s distribution of utilities is random - e.g. when they’re distributed, who gets them, how many everyone gets, the nature of the utilities that are distributed, etc. Example: resources yielded by dice rolls in Settlers of Catan; the location of “villages” (which give early access to good tech) and resources in the Civilization games. - “Threat-Random” - where a game includes one or more NON-PLAYABLE threats that a player does not opt into receive (i.e. discounting cases like choosing to draw an extra card in Blackjack and going bust as a result), and some facet of the threat is random (e.g. when they happen, or their frequency, or their potency, or their identity, etc.). Example: “Bowser fire” in the Mario side scrollers; summoning the Robber in Settlers of Catan. You could further distinguish between types of “Player-Randomness” by comparing games where the player experiences a chance of error no matter what they do (e.g. passing error in FIFA), and games where the player typically/always has an error-free option but sometimes/always also has an option with a chance of error (e.g. moves in Pokemon, if the player ensures they have a move with 100% accuracy).
Nah, the "other" player would just be the recipient of the results of the "player-random" event. As defined above, utility-random and threat+-random apply only to non-playable facets of the game.
Well TBF the chances of 5 misses are 0.01%. Which seems *_really_* low, but for anyone that crunched the numbers for card games... it's in fact 5 times better than getting a Flush in Five Card Stud. :-B
"Randomness" is one of the things that makes the dynamic ecosystems in Rain World work. Input randomness takes the form of the game's rain timer (which is decided at the beginning of each cycle) and enemy movement (which happens continuously). Players aren't given a birds-eye view of their current region - so the lizard that "randomly" pops out of a pipe to scare you is merely the result of your lack of knowledge. By understanding the specific traits of each ecosystem, players can more easily navigate what are essentially organic rube-goldberg machines.
I've never heard someone else talk about these concepts, so I ended up calling them front-loaded and back-loaded randomness when trying to explain the difference in various tabletop systems to other people, though there's sadly not many tabletop games that use input randomness that only use numerical dice.
I love it so much when you said "that's how we think numbers work" 😆. I work in a casino dealing cards and roulette and I find it mind blowing how many people scream "THAT'S IMPOSSIBLE!" while running back and forth to an ATM when 5 red numbers hit in a row. I used to discuss it with players but they seem to have it in their head that there's some kind of magic entity in the universe that'll pull that ball to a black number simply because of where it landed last and that 7 years of me watching the lack of consistent patterns on that wheel means nothing.
To be fair: 1) Gonzalo Pelayo 2) There is a mathematical strategy to always win, and that's to double your bet every time you lose (it just requires infinite money and a table that allows arbitrarily large bets, all in order to net-gain a dollar).
I always though about randomness in games as possibly being either horizontal, vertical or a mix of the 2, where with horizontal I mean that the different outcomes only add variety and don't change the difficulty of the current challange (or the entire game) in any way, while vertical is the opposite, it directly affects your chances of winning while not giving any variety, most random events are a mix of the 2
I ran a Game Jam all about randomness! Check out the results of the 2023 GMTK Game Jam, where the theme was "Roll of the Dice" - th-cam.com/video/XNCGdi2A6fQ/w-d-xo.html
thanks internet explorer
Cool
: Are you a gamer?
: No, I am a _risk-calculating_ _tactician_
Slap that on your résumé.
Ah yes 420 likes
lol that’s the sex number
Grunkel Stan in a nutshell
Zimmy what?
In players' defence, it's not just the disconnect between probability and our broken understanding of probability that causes anger (although that's part of it); it's also the disconnect between our intuitive understanding of *the situation* and the outcome. Yes turn-based systems are always abstract but I'm still seeing a special forces soldier point a rifle at a stationary target 3ft away and somehow miss. It completely shatters the fantasy of controlling this awesome elite squad.
I agree with this and to add on with my experience. I used to play a lot of fortnight and apex the former being discontinued and the latter being pretty close to it. Both suffer from the problem that your likely to die in the first 5 or ten minutes because you got a pistol and sniper while someone else gets one of the broken guns and just slaughters you in a few milliseconds, for apex you might not get armor or anything to give you health while people get both in surplus. Some randomness I understand but if I continuously die like this I'm going to get angry
Missing 5 80% shots in xcom 2 is so frustrating when it's so hard to even see HOW they would miss
Add to that the bizarre way in which things act in sequence despite them supposedly being simultaneous - hence the miss chances and the like. Goes all the way back to DnD for this particular issue.
@@SolaceCafe that is more a problem on how the designer's design the level by making certain things too artsy like hologram's for example people know there is nothing there so why can't my guy shoot better missing 5 80% chances in x-com is unlikely but not impossible the problem is in X-Com the numbers are fudged in your favour as in a 80% is actually like a 85% i think missing is still bad but it is like rolling 1 5 times in a row on a d6 it is unlikely but you have to plan for it if you want to be good
also in X-com something to remember is the RNG generated is the same when you save scum if you do so just doo attacks in a different order then the 5 misses you will have inone go will not be wasted on 80% shots
@@deathtoll2001 yes but D&D for example is actually more controlable average level 1 character will have a +5 to hit most things they will fight will have a AC lower then 15 so you will always be above 50% and rounds tend to not last that long in D&D epic battles tend to take less then 1 minute of actual time where as in like x-com rounds can be longer then ten and you still aren't finished with the mission
even the best laid plans fail
I passed over this video many times figuring it was about true vs pseudo randomness, but glad I finally clicked. A really cool video. Thanks!
I did this too! Glad I clicked.
They make awesome videos!
me too lol
E
Yeah I thought is was gonna be true randomness vs mathematical randomness
"The risk was calculated, but man am I bad at math."
Houston Paul yup
I calculated the risk, but I’m just terrible at Math.
The risk was calculated but decided to not hit the like button because its at 666.
Sounds like Mangs.
"Do we just like luck when it lands in our favour, and hate it when we lose?"
Yeah.
Yeah.
I'm not going to ruin the number of likes on this comment, but I like this comment
Yup
I like luck in games bc i need to think even for if im unlucky so for example:
I have a plan with 95% chance of succes
But due to how probability works i should have backup plans for my backup plans
I just hate that luck gets involved in my decision
It's been talked about ad nauseum at this point, and still I want to draw your attention to 14:42 - something about X-Com's animation system is such a f***** slap in the face. 99% chance to hit, gun is deadlocked on the target.
> Soldier moronically pivots to the floor to miss a full burst of machine gun fire at 2 yards.
It'd be a legit way better experience, with the *exact* same number stats, if the animation somehow made sense. Like bullets flying underneath arms, or hitting cover really close by. At least that way the feedback would be "shit happens" and not "your soldier who you have spent a lot of skill points to upgrade and has seen many victories in many different parts of our alien-infested world is an absolute mongrel, but only every now and then."
The original XCOM (UFO: Enemy Unknown or XCOM: UFO Defense depending which version you had) was worse. You had "time units" (action points) which you spent for each step of movement, as well as when you were shooting, and even for turning on the spot. If you whiffed your shot, you'd fire in a random direction, but if that was more than like 20° off from the target, you had a risk of spending TUs turning and not having enough to fire any more. If you did have the TUs for it, you could move up your rocket dude, line up for a shot into an alien spaceship, and "accidentally" turn 180° and shoot your squad in the face, blowing up everyone with a single shot.
Agreed. I always wished the animation was one where the enemy would just dodge behind cover or just hit the deck at the last second. Or at least give enemies a DC to hit like they calculate already and subdivide lower rolls as reduced damage with one chunk reserved as a complete miss.
@@a-blivvy-yus Actually, I liked the old games way better because of time units. They gave you the flexibility of spending all actions on moving or solely on shooting, not forcing you to make 1 move and 1 shot every turn. I know there's actually an option for longer move without shooting, but still I find time units more flexible and allowing you to adapt to situation. E.g. you could be moving a step at a time consuming just small portion of time units and assessing your surrounding with every move. In newer games you would just make one step and deplete all your move capabilities. Moreover, time units were growing with your soldier getting experience, so after some time in the game your troops were quite a commandos. I suppose in the newer series it's the same pace through all the game (one "time unit" for move in a given proximity plus one "time unit" interchangable between shot or extended move).
@@michaniewiadomski7911 There are good and bad things to the new system. It's more streamlined and easier to manage without losing too much of the depth of the gameplay, but it is very distinctly different from the originals in a way that doesn't feel necessarily better. I like both types of gameplay for different reasons, so I won't claim one to be better than the other - which I play is dependent more on my mood at a given time than one being a better game.
@@michaniewiadomski7911 And the problem I'm mentioning was a problem partly caused by the game's use of TUs... which isn't a problem in the new game, and which was (eventually) patched in the originals.
I think you should have mentionned the games where the player is in charge of the luck. In Armello, you can burn cards to choose the result of the dice, so the player decides how much luck they're willing to rely on, and what price they'd pay for it. That way, when the result isn't favorable, the payer blames themselves for not paying more, instead of the randomness.
Danganronpa also has a similar system with the MonoMono Machine, where you can pay more tokens to decrease the chance of obtaining an item you already own.
I played a game called void tyrant where you're basically playing dice blackjack against your enemies to attack, but you can play cards to change your rolls or attack directly, and you get to chose between playing more cards or doing more damage.
Olivier Dubois danganronpa is great because of that, i can save time when i have a fortune
The old Fighting Fantasy gamebooks had a variant on this. You could spend 'luck points' to influence dice rolls in combat, decreasing damage done to you or increasing damage done to opponents, but your total luck points also acted as a buffer against unfortunate events where the player is asked to roll against their total luck score so there's a balancing act there. Luck points could be increased throughout the adventure, typically as a reward for making smart decisions or being curious.
Mitigation, that's the special word. KoL forums taught me that. Ways around the random at a cost, and mechanics that cut the long tail.
Melee does that same thing with the trophy lottery, where you spend more coins to decrease the chance of a duplicate. It doesn't affect gameplay, but it's still an interesting system that likely influenced those games you mentioned.
"99% hitchance, still miss!!!"
It's not even a mood anymore, it's a lifestyle at this point.
Thracia 776 in a nutshell, also getting hit at 1% (you can't even get 100% hit or 0% hit chance displayed, it caps at 99% hit and lowest is 1% hit)
Hit or miss, i guess they never hit huh
@@bacchus9389 displayed hit chance in XCOM lies, and lies a lot - there is multitude of hidden factors (and mods to show them), including things like character/global miss streak modifier that up your chance of landing a hit. You can have guaranteed hits, the game just won't show them to you - and when you memorize how it works, it's possible to play around/abuse the system for your own advantage (like wasting low percentage shots to guarantee a takedown before end of turn).
@@asmonull Minor correction: XCOM2 doesn’t use these mechanics on the highest difficulty level (others aren’t interesting anyway). XCOM EU and EW don’t cheat in your favour at all.
@@asmonull actually it doesn't lie, u can intentionally go for low chance shots to boost the hit chance u need to kill a big bad enemy and the displayed hit chance will go up every time. Well... i guess if ur talking about ur soldiers hit chance then yeah the game doesn't tell u where that extra hit chance is coming from but the percent chance to hit an enemy is accurate.
"For those who are unfamiliar with the term RNG, RNG stands for Really Not Good."
- Ceave
I like that definition. It’s surprisingly simple
@Hallowizer So I think you've played terraria and if You've played terraria then you know how evil rng can be.
Hoo ray!
Well, the source of that is actually surprisingly *simp* le
minecraft world without randomness is Really Not Good
The main problem is that people remember negative events more easily than positive ones. Yeah you'll sometimes miss in XCOM with a 99%, and you will distinctly remember just a couple rounds ago when you also missed with 99%. But you will forget the 100 shots you took in between that hit, they were exactly what you were expecting so you just move on and forget about them. This is also why no one complains about when they have strings of good luck, it's not just that the good luck makes them happy it's also that they are less likely to even realize that they are having a string of good luck at all.
yeah, although I view XCOM more as reminder just how big and small 1% chance is
I think the opposite is what makes dark souls so good, well at least for me. Instead of good things happening to you all the time, bad things do, this makes you remember the good moments because they are a lot rarer. For example, I don't remember all the times I lost to a boss, but instead, I remember how good it felt to finally beat that boss.
you never saw me play DnD 5e
whenever i get a streak of good rolls, i get nervous that my luck will turn soon and i roll very bad.
that makes you consider to think about your options, maybe instead of attacking the enemy with a direct attack that can miss, you instead cast a spell that weakens them and improves your teams chances, for example the 2nd level cleric Spell "Bane" cause that up to 3 enemies have to perform a wisdom check every time they make a check, if they fail, they suffer a 1d4 penalty on that throw. also this spell cannot miss, which makes it rather neat, but its a 1 minute concentration spell, so it limits what you can do. you could also use "Sacred Flame" (yes, im playing as a Cleric) which will never miss, and deals rather low damage but the enemy has to perform a saving throw to take half damage. this cantrip will guarantee some damage and i love to use it to finish off enemies. otherwise its rather weak. but i still go for the risky throws, surprisingly i dont fail my dex checks that often, even though i have -3 on all my dex checks. also critical hits with a weapon just feel awesome.
Also no one takes 1% shots because its extremely unlikely to happen so that end of the scale is never explored.
@@firebladeentertainment5739 "whenever i get a streak of good rolls, i get nervous that my luck will turn soon and i roll very bad."
This is a perfect example of how stupid the human mind is when it comes to randomness. We have zero clue how likely a long string of a certain outcome is and we tend to couple events together when they are independent. It is so common to think of luck as a finite resource, so when we have great luck in succession, we are bound to some bad luck, but that simply isn't the case. If you ask people to throw a coin thousand times and note the heads and tales, you would most likely be able to see who actually did it and who phoned it in. The ones that feature long streaks of either heads and tails are the true notations, but we humans doesn't perceive this as random and will likely never make streaks as long as they appear in real life.
Funny side note. Most shuffle system for music playlist are actually not random, because the truly random doesn't feel like it to us. Instead they are made less random to feel more random.
I'm reminded of when Team Fortress 2 first released weapon drops. Basically, every time you died, you had a chance of getting a random weapon. The percentage chance was set so that you'd get something every dozen deaths or so, and that's how it worked for the vast majority of people, yet some people were going hundreds of deaths without a single drop. Valve's mistake was that the game "rolled the die" on every death. Across millions of players, this meant that some unlucky folks ended up on the far far far end of the probability distribution. Put another way, the odds of rolling 10 six-sided dice and getting all 1's is about 1 in 150 million. However, if you have 20 million people each rolling 10 dice several times a day, someone will eventually draw that shortest of all short straws. Valve's solution was to change their game so that, after each drop, it picked a random number from a range that determined how many times you needed to die before you got your next drop. This created a hard limit on how long you could go without a drop.
Similarly, I don't doubt that some (maybe many) of the people who take to forums to complain about things like the hit chances in X-COM are suffering from a psychological misunderstanding of the odds, but I also suspect that, in a game that sold millions of copies, it is not impossible that a few people did, in fact, miss a dozen or more 95-99% chance shots in a row. It's this sort of thing that makes me think developers using RNG should always include some sort of backstop to prevent the game from being ruined for whoever ends up on the highly improbable end of their probability curve.
Valve uses a system called "pseudorandom distribution" for many games. Most notably, DOTA 2. Whereas a "X%" chance to trigger was a true "X%" chance, they changed it so it starts as X% on the first try, and it becomes more and more likely to trigger the more it doesn't, and less and less likely to trigger immediately after a successful trigger. Basically, it "evenly" tries to space trigger events (based on the "base" % to trigger), and thus plays with our expectations (biases) of randomness.
@@francoiscoupal7057 It's a really cool system, originally created by Blizzard for WarCraft 3. And because Dota was using the WarCraft 3 engine, it ended up using it as well. So when Valve created Dota 2, they copied the system over.
@@SchemingGoldberg Honestly, it's such a basic system (applied statistical events manipulation) that claiming a a game company "invented" it is a very shaky claim.
That they systematically used it in their games before another dev did, fair enough. Remember though that's it's not a "revolution", it's not "better" than using a true random chance element. It's deliberate fudging with randomness to level a playing field. (Think Mario Kart randomixation to give a chance to the least performing player).
One of my friends was one of the "shortest of short straws." I did the math, based on the then-known probabilities and time scales (it was something like 0.05% rolled every 5 minutes) and how long he'd gone without seeing something and it worked out to 1 in a million for that two week period (with an active player base of 2 million).
and yet, I still seem to just get drop notifications *at the time where I'm waiting for a respawn and I want to take advantage of the wait time to see how the game flow is going* .
I die, I want to watch the kill cam to see where I go next, but I'm prevented of doing that because the drop system kicks in with a big "HEY! YOU SEEM TO BE IN A HURRY TO GET BACK INTO THE GAME! HERE! LOOK AT THESE NEW COOL ITEMS WE HAVE FOR YOU!"
and it's a weapon of which I probably have three copies of.
I mean, how "random" is "right in the middle of the wait time between two critical moments of play"?
I feel like the issue with % is that we generally don’t understand the idea of that % being rolled every time something happens. For example, just because flipping a coin has a 50/50 chance of landing heads, getting a tails on the first flip does not increase the odds for heads on the next one, the probability is still 50/50 and will always be that. It’s just that mathematically we *expect* it to land on each side equally, even though in practice that rarely ever happens
Basically, there’s a difference between the odds of getting HHH and the odds of getting H after HH, but a lot of people struggle to perceive that
True, but all of that depends on the total number of "possible throws" that the chance is based on.
Let's say that two archers try to hit their mark 100 times.
Archer A hits the mark 70 times
Archer B hits the mark 30 times.
If the exercise is repeated, then A has a 70% chance of hitting, and B a 30% chance. It's an indication that A is more skilled, and therefore has a higher chance of hitting.
This also means it's strictly possible that for the next 100 shots, Archer A misses the first 30 shots.
What I'm trying to say is: in many games, we have no idea on what exactly the "chance to hit" stats are based on. Are they based on 100 simulations of the action? On 10? On a million? That naturally isn't really something the developers can make clear to the player while they are playing, but it's a factor of frustration nonetheless.
For your example, while it is true that there is a 50/50 abstract chance for a coin to land on heads, in reality, that chance will be more or less divergent. If the coin keeps falling on tails, and after 100 throws, the coin in fact landed on tails 93% of the time (literally 93 out of a hundred) then we can no longer say the chance is 50/50. Obviously there would be something wrong with the coin (think of loaded dice, if that makes more sense.)
Because the possible simulated throws of the coin is potentially infinite, the 50% chance to land on heads is really not an accurate representation of what will likely happen.
All in all, it's about the information the player receives. If what I'm reading in the comments section is correct, then those "chance to hit" stats for many games are bullshit anyway, as they are either higher or lower in reality. Misinformation leads to frustration just as much as ignorance does.
well, as he talked about in the video some game developers tend to modify luck so that it matches people's expectations, like if you expect to land 1 of 2 shots cuz of 50% chance, they will guarantee you one shot of them so you can actually feel satisfied
@@RedFloyd469 I'm sorry but I think you have a wrong understanding of probability which is called the gambler's fallacy. It is a common misconception. Just as the original comment pointed out, there is no connection whatsoever between past and future outcomes.
Let's say you flip a coin: Even if you hit heads 1 million times in a row, that doesn't influence the outcome of the next flip. To phrase it differently, the sequence HHH(...)HH is exactly equally probable as HHH(...)HT.
The higher the number of tries, the closer you expect the results to get to their probabilities, but there is no guarantee that will happen for ANY number of tries.
This seems counterintuitive at first, but it makes sense.
Saying it wont rarely happen isnt true since it is still 50/50
10:36 - "These cards are terrifying, game changing events that can completely demolish your team."
As someone who has played Pandemic several times, that is so true it's not even funny. It's always really dramatic when one of those cards rears its ugly head.
If it were more dramatic, the lights would dim and the card would glow green upon reveal, illuminating the horrified face of the unlucky person who drew it. The other players would gasp a little, and one player would ask, "Is it an Epidemic Card?" even though they already know the answer. The poor lad who drew the card in the first place would look up at the rest of their team, drop the accursed card on the table for all to see, and answer solemnly, "Yes."
I love this comment
This comment made me feel emotions I didnt know I could feel
Man this comment is really a masterpiece
I feel like we sound like bots
My dad has a copy of that game and listening to him describe it is accurate with your description
There's nothing quite like watching epidemics explode from city to city around the globe as you realize you have failed, everyone will die.
My favorite moment in XCOM was playing with my friend.
We were into the late game, only one death early on, main squad pretty beefed up.
None were showing pyscikik promise, so we swapped out one with a promising member. Nickolas Cage was his name. He was getting up in skill as well.
We entered the downed alien ship, fighting a brutal battle. We make it to a long hall way, three Cyber Disks show up. All of our shots miss. They fire, kiliing two.
We fire again, not much effect.
They fire, missing only on Nickolas Cage. The team is dead.
We say F it, and charge with Mr. Cage. He kills one.
All of their shots miss.
He gets a reaction shot.
He kills another.
He takes a blast to the face, still lives.
He kills the last. Mission completes, huge haul of resources making research possible that lets us crush enemies with (relative) ease till the end of the game.
Had a similar experience, at least in the backseat watching a good friend play XCOM (Enemy WITHIN, I believe). Turned a decently-lucky soldier into a giant mech guy, and from that point on there wasn't a damn roll he couldn't make. Never missed a shot, period.
I would like to see the RL nicholas cage reenact this event.
XCOM actually swings RNG in your favor if you only have 1 soldier alive. That's how people do solo runs.
@@longevitee Wow that really ... spoils the illusion.
@@longevitee Ye, they do it already when 2 soldiers are dead. Their system is pretty screwed up, if you ask me.
The big thing about output randomness is it feels much better when the player makes a choice to take that risk, rather than just being forced into it.
If you have to choose between option A that has a 50% chance of success or option B that has a 50% chance of success, there's only an illusion of choice.
But if you have to choose between option A which has 100% chance of 1 success, or option B which has a 45% chance of 2 successes, and 5% chance of 3 successes then it becomes more of a decision.
Not only is it better to have actual choices, but it puts more emphasis on when you choose each one.
You might only gamble when it is hopeless anyway and that 5% chance sometimes saves the day.
Or you might gamble when you're so far ahead that a loss doesn't matter as much and success allows you to feel even more powerful.
It gives the players better control over how much they are willing to be hurt/helped by RNG.
RNG in online blackjack
Reminds me of a D&D session we had. I was helping my friend to make the best thief ever, helping him choosing the right perks, the right equipments and powers and magic weapons. We gave the character the name Chad. With all that combined, his thief became so OP he could hit anything on a roll of 5 or above. Turns out that, that whole night, he couldn't roll anything above 4!
It has been years but to this day we still remember and laugh at the tale of Bad Luck Chad.
Feels like one of my newer sessions. Going through a dungeon with 20 AC and still being hit into submission way more than that mage with 12 AC... welp, that's what you get for trying to be a tank, I guess. And still other party members complained about them having bad luck!
>Chad
Playing War with some friends, one of them used 30 armies to try to take a place occupied by 3 armies of another friend.
The defender lost two armies, but kept the territory.
Thats what lucky and reliable talent are for
That reminds me of a battle in Axis & Allies many years ago. My opponent should have crushed me with his overwhelming force, but he kept rolling high numbers--one after another after another. It was the most remarkable string of bad luck that I have ever seen in my life. I won the battle with my tiny defensive force, and I awarded them medals for their bravery and valor in the face of overwhelming odds against them.
"I've never said the word Epidemic so many times"... March 2020 says hi
Have a reply
Have another reply
Have yet another, reply
A reply for the cause, good sir!
have another reply but this time the reply tells you its a pandemic, not an epidemic
I like to think that Mark's regular videos like these are like lessons and his yearly game jams are the pop quizzes
Quoting someone from one podcast I listen to: “I get angry whenever I watch Mark Brown. You can be good at editing, good with words or being right, but not those three at the same time; yet he does all of them every time”.
which podcast?
What podcast?
Podcast which ?
What podcast?
To be fair, being good at editing means you have more time to be crap at the other two things without us noticing.
I'm surprised there wasn't a bit more discussion of Fire Emblem. As you said, they famously use 2RN (i.e., they roll two random numbers and take the average) to make up for how bad our brains actually are at understanding what "75% odds" means. But the most recent entry in the series introduced the Divine Pulse mechanic, which gives you a limited number of chances to go back in time, but the RNG stays the same. It stops you from resetting the level entirely when something goes wrong, and the RNG being fixed means you don't just try the same thing repeatedly until it works in your favor. But my favorite part about it is that it transforms the output randomness into a matrix of input *and* output randomnesses. The entire level's RNG sequence is fixed from the beginning, sort of like a level in Spelunkey or Gungeon. And if you understand how the RN system works, you can incorporate this randomness into your strategy. I never felt better at FE3H than when I was crit by the Death Knight, but used Divine Pulse to roll back the clock and changed my actions just enough so his critical hit became my own, which required: 1) knowing that his crit was the result of a low RN, 2) knowing exactly how many RNs were used before it, 3) coming up with Plan B to use up that many RNs so that the Critical RN is in the right place for me to use it, and 4) making sure that burning those RNs in Plan B doesn't put any of my other units in jeopardy who were safe in Plan A. It's a bit like counting cards, I guess, but pulling it off really made me feel like I had mastery over the game, and its number of uses are so limited that it doesn't really make up for playing the game poorly.
Yeah when I realised the rn sequence stayed the same my divine pulse abuse became really satisfying and less braindead
The rewind ability actually first showed up in Echoes, where it was called Mila's Turnwheel.
This is also interesting to me because before the double RNG check was introduced, the most common way the games were balanced was that most playable characters attacked twice during combat. It was supposed to just be based on the speed stat, but the vast majority of enemies were so slow that they were easily doubled, and in the last game before the new system, there were a whole bunch of new weapons with the Brave effect, which hit twice per attack, and thus mimic the potential "half hits" in Phoenix Point.
@@jeffd.683 Yes, but in there the RN changed so reseting until a positive result occured happened a lot. Also i believe Echoes didn't use the 2RN system because it was a remake.
What's interesting too is I found FE3H actually uses multiple strings of RNs, unlike in most of the previous games which only used one (some of the earliest games in the series used a clock RNG).
First there's one set of RNs for combat actions (hit/miss and critical chances), and then each playable character has their own string for level ups. This means that the stats a character gains on a particular map are always set, and cannot be affected by rewinding time. If, say, Edelgard gains only a paltry +1 Magic on her first level-up in a map, you cannot rewind time and get something different, even if you take different actions leading up to the level-up.
I think this was done to prevent the player from abusing Divine Pulse to get super-strong characters, though on the Normal difficulty setting you can use Divine Pulse with the Retreat option to get the same effect.
Players when they get the item that has a 1 in 50,000 drop rate: FINALLY I GOT IT I OPENED SO MANY CRATES
Also players when they have a 1% chance of missing and they miss: WHAT IT IS SO RARE ITS NOT SUPPOSED TO HAPPEN!!!!!!!!!!
That's certainly the best way I've read to tell the issue.
It's less a problem that it happen but the rate at which it happens.
@@scottrauch1261 Except that most players will take 200 shots with 99% hit chance, and complain both times they missed that "it was 99% how did I miss?" then because those moments were pivotal in their strategy, blame "bad RNG" for their loss instead of a plan that didn't account for the fact that sometimes you're going to miss.
Psychological studies have suggested that people feel bad results twice as strongly as good ones, so an accurate experience with 70% hit chance misses often enough for most players to feel like it's a 60% miss rate.
People have deconsructed, reverse engineered and tested XCOM's randomisation algorithm and confirmed it to be fair, but there's still complaints and claims that it's horribly broken in spite of evidence to the contrary. Because "it feels like I never hit" when you're taking 60 - 70% shots and hitting 60 - 70% of the time but feeling the impact of those misses more than the hits so you think you're hitting 20 to 30% of your shots instead.
@@a-blivvy-yus I've figured the statistics on my own attacks missing but I've always had shit luck so not to surprised to be an outlier.
@@scottrauch1261 In which game? Because for XCOM, people have accumulated insane amounts of stats from multiple players who have hundreds of hours of tracked and verifiable results that have been compiled and analysed. Individuals have had results which are off from the normal "expected" values to varying degrees, but the overall stats are consistent with what should be expected when you're checking the data on a large scale.
Of course, there's still the bug that makes it possible to miss with a 100% chance to hit and hit with a 0% chance (I've done the latter in multiplayer, it was hilarious). And it got worse in XCOM 2, where they claim to have "fixed" the problem, so now on the easiest difficulty, an attack with 94% accuracy has a lower probability of missing than one with 100% accuracy... I could be wrong, but I feel like they might be confused about what "fixed" means.
“Scientists have calculated that the chances of something so patently absurd actually existing are millions to one.
But magicians have calculated that million-to-one chances crop up nine times out of ten.”
― Terry Pratchett, Mort
"60% of the time, this machine fails every time"
Killing a dragon by getting the shot exactly to million-to-one odds is some of the best Terry Pratchett I've ever read.
Xentillus IIRC they didn’t actually get the shot a million to one. I think the odds of that shot not being a million to one was a million to one. I could be wrong though, it’s been ages since I read the books
They messed up on the million to one chance for the shot, but surviving their mistake was a million to one.
I've never heard this quote but I love it, and want it on a t-shirt.
As I think about this, I think for me it comes down to the following: Good luck-based gameplay says "X happened, now how will you deal with it?" and opens up options. Bad luck-based gameplay says "X happened, too bad, so sad" and destroys options or even ends your game.
It's interesting to be provided a set of new options, whether those options are you thinking about how the game works or are explicitly provided. And while occasionally removing some options to encourage players to not get into the same pattern forever might work sometimes, it generally just means you go from like 3 possibilities to 2, or worse, being railroaded into having to do a single thing, in which case your input feels redundant.
I think that's how Dysco Elysium does it and it works great. If you fail a roll you are not put directly in a fail state, but the dialogue continues so that you can try to recover (if you have heals) or even dig a bigger hole for yourself if you are not careful.
What you described is part of why I prefer playing XCOM2 in Ironman now. Its taught me to manage and calculate risk and makes my decisions more impactful. When I didn't play on Ironman, missing a shot meant I reloaded. Now it means I'm thinking in advance "if I miss this shot, how will I manage things to try and make sure no one dies."
Part of that risk management can even involve choosing who I put into a risky situation. A Psi-Operative who has the cheat death ability? Take the risk with them basically every time. A rookie I've invested basically nothing into? You're going into the line of fire baby. A max-rank dude who I've dumped tons of bonus AP into? I protecc, stay safe friendo.
It still sucks to miss a 97% (curse you 97, I hate you so much), but I feel like I've grown as a player when I'm able to assess where I'm at every step of the way and sometimes even abandon a mission that is going poorly rather than try and get the win after losing a character or two. Those setbacks can make the game even more enjoyable in the long run, because I have to manage the consequences of failure.
I don't think it gets redundant if at that point your reaction time or ability to notice what needs to be done get tested. But you present a good way of thinking about it
This is generally why, as a massive roguelite nut, The Binding of Isaac is my favourite roguelite ever made, and I think FTL is overrated.
Isaac has ludicrous room for player growth and potential and (contrary to popular belief) essentially any run can become an incredibly OP gamebreaker in the right hands. FTL is mostly a game of math where the numbers you get for your equation are random.
(this is not to say FTL is a bad game, it really succeeds at providing a Star Trek-esque strategy game and is still the best game I can think of for that specific experience. As a roguelite though IMO it kinda fails)
@@coffeedude I was more thinking of turn-based games where it's more about what set of more tightly-defined choices you make (i.e. selecting an attack via menu vs. having to manually perform the attack) since they tend to not offer as much wiggle room in terms of gameplay actions and allowing high skill to overcome a bad circumstance, and also because they tend to be the games that employ luck-based gameplay far more than games that require twitch reflexes, aiming skills, timing, etc. But you're absolutely right.
"Theres never been a good movie where the heroes scheme works entirely as planned"
*The entire Oceans franchise has left the chat*
I believe the rule is that if the hero's plan is explained or otherwise known in advance, it will always fail. It only succeeds if the audience finds out during or after the execution, which is how heist films work
I mean, hitches in the planning are a staple of the franchise. For example from the first movie, they have to steal the emp because the casino discovered the vulnerability in their power system the thieves were originally planning to exploit. The theft of the emp goes wrong and leads to Yen needing a cast. That cast gets caught in the vault door during the heist.
The fun of heist movies is that things never go exactly according to plan, but they find a way to work around the surprises.
@@helplmchoking The Godfather part 1 would like a word...
@@TheWorldBelowDnDdo you mean when al pacino ices those two gabagools
We need more videos like this where you talk about a specific design concept and not just breaking down how a certain game was designed.
15:00 Explaining to people that it's just as likely to get heads 500x in a row as heads and tails alternating for 500 throws is one of the funniest things
Ahhh true, I meant more one or the other though, but I probably should have phrased it as such. Next time I will be more specific I guess
@Luke13139 if its specifically alternating, then the chance is the same, because its following a specific pattern, so the statistical probability is the same
or any other sequence for that matter
Yes, but there's only one way to get 500 of one specific thing
There's a bell curve distribution of ways to get within roughly equal of two though
Uh no it isnt. Unless you mean alternating as in HTHTHTHTHTHTH... exactly (which literally nobody cares about because literally nobody thinks HTHTHTH exact is likely at all) then the probably of getting alternating throws is much much higher than only HHHHHHH...
For example with 5 tosses:
HHHHH
HTHTH
HHTTH
HHTHT
HTTHH
TTTHH
THTHT
HTHHT
...
You see how there's gonna be a bajillion more sequences that alternate Heads and Tails? If I flipped a coin 500 times and i got all heads, that coin is rigged as fuck.
This is perfectly put honestly! Output randomness does tend to be more frustrating, and percentages in games tend to be lies to make you feel better (like the example you put where it says 90% but is more so 99%, it makes nailing that 90% feel better and when you loose it's not as bad as loosing a 99% chance to get something good). Output randomness is often better when in the favor of the player, a random chance to get something great always feels better than a random chance to get absolutely decimated by RNG and loose. Input randomness is by far preferred as well, as you mentioned!
I believe a small balance of the 2 makes a truly satisfying game, input randomness with a little output randomness sprinkled in to spice it up (both good and bad, like the chance for a few shots to miss whilst also having the random chance for the ones that do hit to crit doing more damage or having the chance for the enemies shots to miss to which makes you feel great for them missing).
"Randomness is also a way to balance a multiplayer game."
*screams internally*
*Screams In Hero*
scream in tf2
pwns scrubs in spite of it
2020 IS bullsht NOOOOOO
Screams in random recoil headshots in r6
An important thing about statistics is that they converge towards the odds, so frequency is important. For example, if you'll only ever fire one shot that has a 70% chance of hitting it's an all or nothing bet. If you'd keep firing forever you would over time hit exactly 70% of the shots to the point where you could just as well see it as having 100% hit rate but 30% damage reduction. This maths holds up also for one-off events, but it's only really relevant when you can make enough statistical choices for the statistics to even out. That's why for example the solution in Phoenix point where you fire multiple shots with the same hit percentage becomes much more predictable and might feel more "fair" but less risky. So if you design a game with random events happening seldomly they will likely feel more unfair than random events happening often.
Exactly.
Precisely
🙌🙌🙌
Kinda true. We are still very bad at randomness. If we fired 5 shots every time with 50%, and we fired that burst like 20 times we would be surprised when we miss all five shots at some point when we shouldn’t be, so I think it can also help to skew the odds towards the center a bit. It’s better to make that zero shot hitting much rarer, but also making all 5 shots hitting much rarer than to actually keep it at 50% for each shot because we don’t realize how after doing it so many times it’s not actually surprising that we missed.
For all-or-nothing scenarios there are solutions for that too, by modifying chances behind the scenes. One nice solution for critical hits I've seen was tracking recent critical hit ratio, comparing it to critical hit chance and adjusting the roll on next attack so outcome of crit ratio gets as close to crit chance as possible, while still keeping it random. This kind of mechanic could work well for likes of pvp games, where you want to keep system fair for both sides, but also keep the feeling of it being fair and random at the same time.
"[Randomness] can be a cruel mistress..." *Darkest Dungeon footage*
Yep that tracks.
"Overconfidence is a slow and insidious killer."
"Dazed, reeling, gasping for air"
@@oof5732 no you got it wrong it's "reeling gasping, taken over the edge into madness"
In the words of Steve Jobs: "We're making it less random to make it feel more random."
I think there actually was something faulty in iTunes, where it actually was built to have a greater chance to play a song from the same artist than any other song. This is not something that has been tested, so we can't tell if it actually was a thing or just placebo. But people haven't complained about this with other music players like Spotify or Windows Media Player, who hasn't stated that they have modified the randomness to avoid this kind of situation.
So my theory, after experiencing this effect myself with iTunes before, that there actually was a built-in feature to increase the odds to play a song from the same artist. Some developer who probably built that in intentionally or acidentally, and people forgot about it. But with the new system that combats this, the old randomiser is basicallt overritten.
Liggliluff Just because the Spotify people haven’t said they modified the randomness doesn’t mean they haven’t. They probably did behind the scenes.
@@dinoactual yeah they _most certainly_ have some shady stuff going on under the table. How easy is it for an artist to say "hey i'll pay you $$ if you play our songs more often" and then they alter the algorithm? I'd imagine a points system would automate so much of that, (of course now i'm just scheming in my head)
SpaceMissile Hold on, hold on. I never agreed to these conspiracy theories. Spotify has probably done a similar thing in that they altered it so songs by the same artist play less often, to give people the feeling of more randomness when it isn’t actually random.
>using str**ming services
Yuck just pirate, big artists can't get hurt
I heard at a conference once that its also important *when* in a game randomness has impact. Dying purely because of bad luck at the start of a game, not 10 minutes in, can even be fun. But dying because of bad luck one hour into a game and losing all that progress can make players quit, and as game developers we dont want that. So it could be a good idea to make it so luck has less and less impact the more you progress into a game, and make skill and strategy be more important in those later stages
Assuming its not a game where you can quicksave at any moment and never lose any progress, of course
Like the first rival battle in pokemon, it's purely dependent on luck. But strategy gets more and more important.
I'm not so sure about this. I played xcom with a friend through parsec and my whole team died because i missed almost every shot. At this point I'm thinking that the rest of the game will be exactly like this. Needless to say, i hate xcom and will never play any title in that series again.
@@haz6908 wait you can loose those without trying to loose?
@@prinzeszelda3650 Yeah, like if they land a critical hit, or just have a higher base speed and beat you to tackling, or they get lucky and use a bit of strategy, yeah it's pretty common to lose those.
"dying because of bad luck one hour into a game and losing all that progress can make players quit, and as game developers we dont want that"
Yup, sounds absolutely right.
Wow, that opened my eyes about how I've been slowly but surely going away from games with 'annoying' or for me frustrating output randomness.
5:09 Counterargument: Ocean's Eleven
I think Ocean's isn't really an exception, because they don't fully _explain_ the plan to the audience beforehand. In many cases, it _seems_ like things are going sideways, only to be revealed further on that those setbacks were _all part of the plan_ .
You here?
exactly what I thought of!
@@ReverendTed that can be translated into a videogame tho
To be fair he said "a good movie" :v
I always felt like I was hitting 90% attacks in Fire Emblem more often than I should have. Guess that comes from xcom conditioning me.
XCOM also inflates hit chances behind the scenes
Fire Emblem tends to have "2 RNs" in most of their games. It means it rolls the number twice and avarages them out.
So let's say your hit chance is 90%, that means you'd have to roll two numbers which their avarage is lower(or I suppose equal to) than 10.
You could roll a 1 first and then let's say 22, for the same 90% chance check, and it'll hit even though you rolled a 1 first hand.
@@ryanoutram7059 Xcom teaches us that a 70% chance to hit is terrible odds.
You play any mmo and you'll often get a weapon or item that has 1% chance for extra fire damage or something. Which is usually pretty good. That 1% extra means over the course of your session you'll see it pop quite often because we roll the dice a lot in those games.
Xcom is in the perilous position where you are destined to fail at some point because we roll the dice very few times, so the few hits on a 30% have a greater impact on the gameplay than they should while 90% chance to hit always feel like you're playing with fire. A whole shootout can just straight up end or be won on 2-3 dice rolls.
I've never gone back to xcom because of that. I don't want to spend the rest of my life making backup plans and alternative strategies for a game that's supposed to be fun.
It just became work after a certain point. Replaying any mission feels like you wasted time, losing men feels like you're weakening yourself for the end game.
There's no benefit to playing the game ironman. There is no incentive to work through your bad luck. In which case, why not just cheat the whole way through?
Why bother playing at all?
@@Madhattersinjeans That's precisely the reason I play Xcom - yes it's chance based, but you make your own luck. You have to react and adjust your plans, and outside your battle you build up your team and base to mitigate that risk.
@@Madhattersinjeans As someone looking to get into Xcom, the Pheonix Point individual bullet chance appeals more than a binary hit chance.
"I've never said the word Epidemic so many times"
Had to instanly check from which year this video was.
That is the most random comment to respond to a "haha relatable" comment with.
also, those events that negatively surprise you (missing a 95% chance shot) tend to linger in your head far longer than the opposite (the opponent missing a clear shot or you getting a 10% chance shot). When that happens you get happy for a while but move on, but when you're unlucky, you just have that anger stuck for much longer
as a pokemon noob, i totally agree
Humans tend to remember negative events for longer, most likely to try to avoid repeating the mistakes that lead to them.
That’s another reason being unlucky is frustrating, you can’t avoid it.
I remember the fact in XCOM I have two missed 95% and (I think) 99% I hit.
Can't tell you the number hit, nor 90+ hit, or missed shots from the aliens.
Actually, no. No one will ever try to make a 10% chance shot. Maybe in some VERY dire situations when playing ironman (no save run) - then, if it hits and saves the day, it will linger for much longer and stronger than any 99% miss.
Cognitive bias
I initially clicked on this video reading “two types of fandom” with great curiosity and worry. I’m glad it’s actually about random in video games
Whaddaya mean, *two types of fandom?*
The two types of fandom:
1) toxic
2) dead
@@jtm8514 never before have i been so offended by something i 100% agree with
@@jtm8514 More like the toxic and the rule 34 types
I feel like this is one of the reasons Hades has felt so good to play. There’s a lot of random elements for each run, but they usually consist of a choice of two or three (room rewards, boons, etc) from a random assortment you become familiar with. And because there’s such a large assortment of different upgrades and boons, you can keep finding different ways to succeed no matter what your “luck” is (similar to what you mentioned with Slay the Spire) so that you’ll keep sticking with that run to see how it goes.
It's the difference between playing Ludo and backgammon. My favorite games always have aspects of luck in them, they just include a lot of methods to manipulate the odds as well.
A part of that is because you're almost guaranteed to get stronger some way or another, albeit more damage, more synergies, more support items/boons, etc... (Which is a good thing, because otherwise the RNG just becomes frustrating at times)
And this is why I find Binding of Isaac: Afterbirth to be so frustrating. There is no guarantee to get stronger in any way and it feels like you can miss way to many opportunities to get items which might not even be good to begin with. Even though there are a lot of DPS increasing, synergising and good support items, there is simply no guarantee that you will get any of those and then you just end up with a frustrating underpowered build.
Merkules Even more so because once you’ve collected a boon you are more likely to receive booms from the same god, which will usually improve your previous one if you do choose.
YAY! Yes!! And HADES is awesome!!!!
@@quintinbassett9467 I don't think that's confirmed. You have pendants that can guarantee you one boon from a XYZ god but I think that's it.
About randomness in role-playing games: tabletop RPGs inherently live and die on output randomness. The effectiveness of every action a player character takes--and most actions from the opposition--are determined primarily by the dice they roll and the numbers that come up. A 2d6 weapon in D&D is often considered better than a 1d12 weapon, because you're more likely than not to hit a seven. Players lean towards strategies that have consistent results, and many systems gain and lose fans based on the odds that an unlikely success or an unfair failure will come up during play.
Why is this important?
A good dungeon master can make or break a game, even in the most or least popular systems, and many of the best DMs have learned to preserve player engagement with a system often referred to as "failing forward."
Failing Forward means that the player always has a chance to succeed, but the worse their roll the more severe the consequences of that success. A good example is trying to bypass a locked door: a perfect roll allows the party to slip by undetected, but a failure means that rather than failing to open the door entirely, the act of opening the door triggers an alarm. Like you mentioned in this video, the output randomness from a roll of the dice becomes input randomness for the following turns.
I like to think this same principle can be applied to game design. By merely complicating the player's current plan rather than bringing it to a halt, we can retain player investment while still using chance to diversify the types of strategies our players have to use.
I think Streets of Rogue does this quite well. For example, failing the Threaten check causes an NPC to attack you, but you have the option of either running away or attacking them.
Well, 2d6 also have better minimal damage - the range is 2-12 and not 1-12. Overall it is more complicated than this because you have various feats that work with damage dices but that's a different tale.
Also, DM can mitigate a lot of randomness. Lets go with the simplest example - players are searching room for a secret passage - that passage is under a rug as a trap door. DM can ask a question "Describe what your character is doing during this action" and if they described "I look under the rug" DM can say "You found it, you don't have to roll". This way it is IMHO more rewarding as a player - you succeeded just because you are clever.
Someone ha been watching extra credits hahaha. For real though, totally agree.
This is false actually. The D&D-like (and more precisely "modern D&D"-like) TTRPGs are indeed worship output randomness like the holy cow. But it definitely not true for all TTRPGS. Old-school D&D-like (nowdays called OSR and retro-clones) TTRPG preach the mantra of "if it comes to diceroll you botched the planning part" and require a DM to always provide a roll-free solution for any situation (even if this solution is Joestar-special running away). The narrative-TTRPG (Fate, PBTA, etc) often provide you with a way to sent dice packing: if you got yourself in good enough spot, dice can only force you to spend some resources to overrule them. (They also almost always implement the Failing Forward mechanic) Finally, there is a Polaris "something about North and Tragedy" which doesn't use RNG at all (and uses complex bargaining-like system between players instead).
I'm actually is in agreement with you about the appliance of this principle to game design. Just wanted to correct you opening paragraph.
Yep great points... at some point someone just needs to program an RPG where you just roll the dice all the way through the story to emphasize that RNG is making too much of the decisions... if all of your choices are Output RNG then are you really making choices? Might as well just Roll for the very decisions you make Attack or Take a Nap is how effective some of these rolls can be? Yes I know that this is offset by rule manipulation, saving throws, and stat/skill modifiers but there is a reason that you don't start dungeons off fighting a Dracolich and a Goblin as a final boss. It is fundamentally backwards to the idea of progression. My Level 1 Druid took down a Dracolich the first time I played... people are going to be... whatever...
Another nice thing about Dicey Dungeons (other than subverting dice from the usual output into input rng) is that it features numerous abilities that manipulate dice whether it's splitting, copying, or even flipping them. This give the feel of 'cheating luck', that goes along with the narrative theme
"there's never been a good movie where the characters come up with a plan and it just works" did you mean: the entire oceans series
or The Sting
what about the star wars sequal trilogy
The beauty of that is that the plan isn't revealed until the end. It gets into the hidden information alternative to randomness. Suspense is kept because events seem random until the end when it's revealed to be according to plan.
Didn't an Asian guy get hurt by another accomplice and then the bandage got stuck somewhere and said asian guy got almost blown up to smithereens hadn't the bomb remote control not have dead batteries, But he was lucky enough that it did causing mere seconds delay while one of the guys with the remote control change the batteries with fresh ones?
@@highchair208 ^Cause their plans totally work out exactly as they were made...
Not.
Or are you saying Finn and Rose being caught and imprisoned cause they got betrayed by the wrong master hacker was part of the plan?
With both Ludology and Deck Building (edit: and Pandemic!) references in one video, it almost makes me wonder if GMTK should consider a video on "What can video games learn from board games?" Bonus points if you coordinate with Geoff Engelstein.
su&sd gmtk collab when
I’d like them to learn from an actually-punitive RNG game like Blood Bowl, where failure is common and punished to the point of hilarity sometimes, and there are no RNG skews you do not pay for. You can’t play it for long without either ragequitting or reaching a level of RNG Enlightenment where either you understand risk and roll no dice that are not material to your success, or you do not care about risk and laugh as hard at your six vampire TV1000 team literally taking itself off the pitch in turn two as you do the other guy’s legendary Wardancer tripping and instantly dying on a sprint to the endzone.
You don’t need to coddle a player when you teach them instead. And sometimes the lesson is that the dice may hate you, but it’s nothing personal.
@@3thanguy7You should probably check out Cool Ghosts.
That's what I was thinking as well. I'd be interested in Mark's take on board game design and what is its relation with video game design. I'd also be curious to see his take on adapting board games to digital.
That would be so awesome!
*50% chance misses twice*
Every letsplayer I ever watched: ":O"
As a Fire Emblem player I've learned never to trust a hit rate below 88% even if it's a double attack
@Aaron I've been playing Thracia 776 recently, and stuff like that happens all the time, my plans almost always get ruined by bad RNG
Oh and Thracia has 1 RN so 88% is actually 88%
@Aaron the odds of missing two 88% rolls is 1.44%, not sub-.08%. Kinda low, but not astronomically so.
(12/100 * 12/100 = 144/10,000 = 1.44/100)
Reminds me of a very old rule of thumb about stuff like hitboxes.
While it seems like you'd want your hitboxes to be 'perfect', in practice (especially when computing power was limited) hitboxes always end up being vague approximations of the real situation.
The rule of thumb then? Always bias hitbox inaccuracies in the player's favour.
Enemy hitboxes should be larger than the enemy appears to be.
Conversely player hitboxes should be smaller than the player.
(obviously this line of reasoning breaks down in multiplayer contexts, but you get the idea).
The thing is, you can make the same argument about 'randomness', and the same overall logic applies;
Players are far less resentful of things that work in their favour than things which work against them.
“You’ll never get a pointless hat or sword.”
Yeah, a pointless sword won’t be that helpful.
Whereas a pointed hat would be great for halloween
Pun.exe has failed to load downloading virus to,brain
I mean if it’s edged it might have use
Take my angry like
Isn't that just a club
The editing in this video was next level. Better than any other GMTK video before it
As a Warframe player I pray to RNGesus at least 15 times a play session.
You mean as a Space Farming Simulator player haha
@@jeckrucel0001 yea, but farming is fun, unless its Railjack, Im not touching that thing with a shit covered 10 meter pole
2% reactor drop chance
Isn't this WACKY and UNIQUE and FUN?
Are you having FUN yet?
WHY AREN'T YOU HAVING FUN?! ANSWER ME!
You misspelled 'session'
@@Cheesecannon25 shit ill fix it, Im not a native english speaker so tell me if I mess up
11:30 The funny thing is that Diablo 3's "smart loot system" was designed to be dumber at launch to ensure you'd sell things on the auction house.
and gacha games (AKA mobile cancer) has similar systems too. they bring better drops when you come back to the game after a hiatus.
output RNG is just the worst.
That's the problem with trading in general, you need more loot that's irrelevant to you but good for someone else.
The smart loot was added to fix the problem of not being able to trade.
On "there's never been a good movie where the heroes explain their plan and then it goes off without a hitch", the entire heist genre actually did exactly that for quite a long time. The idea that the plan should be disrupted is a surprisingly recent innovation (which is probably why the Ocean's Eleven remake seems to be the example everyone is jumping to).
I haven’t watched many movies in a good while, but even in those old heist movies, wasn’t there usually still a small margin of error where they had to at least switch from plan A to plan B? Maybe not full disruption, but just a small hiccup that they had a second plan for?
@@aqwkingchampion13 My memory of the original Oceans 11 and original Italian Job is that things went very wrong very late. But they certainly didn't all go right. I think I could find dozens of movies that prove that not all heist movies used to go right though.
Even if the plan goes well, we usually see that happening WHILE the plan is explained to us, sometimes as a montage. The idea is that any plan explained in advance is going to go wrong
15:18 To elaborate on that the Fire Emblem series has actually had many different RNG systems over the course of the series.
+In the early games, mostly before they were released in the west and the creator Shouzou Kaga was still involved the games operated on a "single roll system". Meaning the game only calculates a percentage roll once and takes that as a result. The number the game gives you is the actual chance you have of missing, landing a critical, ect.
+After Kaga had left and the games start moving to the west the games moved to a "double roll system", where instead of the game rolling the percentile dice once and taking that as the result it rolls the percentage dice twice and calculates the average of the two rolls as the result it uses to calculate. For example if the games says you have a 75% chance to hit it will roll two numbers between 1-100 and lets say they're 68, and 82, which added together and divided gives you a 70 which the game uses as the result for your attack. The result of this was the number the game told you was actually incorrect and was always more exaggerated than it looked. Chances that where closer to a 50% chance were more accurate while high chances were effectively higher than they were and the lower rolls were even lower. A 90% is more around 99% and a 10% chance is more around 1%. This was done to help create a sense of stability to the game so that if the player was told their chance to hit was high there would be less of a chance RNG would give them those random rolls that screw them over despite that they did everything right. This system is often heavily criticized as hiding information from the player and tilting the games systems even more in their favor. Edit: It's comes to my attention that the double roll system primarily applied to hit rates rather than secondary rolls such as crit rates which function off of single roll.
+In "Fire Emblem: Fates" released in 2015 and "Fire Emblem: Shadows of Valentia" in 2017 tried to marry the two systems into a "hybrid roll system". Its formula is much more complicated than the past two systems but what it effectively did was for chances that were 50% or below the game functioned on the single roll system, this giving lower odds to have a greater chance of surprising you. Odds 51% or above functioned on less exaggerated version of double roll. The number the game would tell you wouldn't be your actual odds and would actually be higher than the listed number but less so than in the double roll system. Less 90% is actually 99% more 90% is 95%. This was done so more likely events were more consistent and could be accounted for and players knowing and engaging in the games systems would be rewarded but without as big of a player advantage as the double roll system game. The hybrid roll system would only stick around for "Fates" as the next main title in the series "Fire Emblem: Three Houses" released in 2019 would return to the double roll system.
Edit: Grammar
Edit #2: New information added
Correction: Echoes, released after Fates, also used Hybrid Roll. Also worth stating that crits and Status Staves always use Single RN
Why don't they just put a "90% - 99%" on screen, and roll a dice that's actually between those numbers. So when it rolls 90 they can't complain and when it rolls 99 and still misses, well they might think it's the "10%" that missed, even though it was the "1%" chance of missing.
@@papersonic9941 Thank you for the help, i've added both bits of info to the post.
interesting!
The "double roll system" doesn't make sense as you described it. If you averaged two dice rolls, a 90% or greater would be less likely, because out of all the possible numbers you can roll there are more that are outside the success range (90-99). If you roll two numbers and average them together, even if one of them is within the success range the other is probably not, which would drag down the result. Likewise, extremely low rolls would be more rare for the same reason. It would generally be pulling all results towards the middle. Like how a single 1-6 die roll can result in any number equally, but TWO 1-6 die rolls more often result in numbers around 7.
This is like saying you roll two dice, so that you're more likely to get a 12 (compared to a single 1-12). It can't work that way.
I haven't played the game, and the only description I know is in this video and this comment, so that's the information I'm working from. But it sounds more like an "advantage" system like in D&D 5, where there are two rolls and you just take the better roll... or if low extremes are also more likely, you'd have to take the better roll if it's above 50% or so, or take the worse roll if it's below.
"You can spend 10, 20 minutes analysing the ramifications of every move" - so, a standard chess game, then.
That's why chess is often played with timers!
@@shadyarian Exactly, I really enjoy playing speed or bullet chess, but I can't bring myself to ever play I game of normal chess.
Yeah, except not for boring old farts
@@All4Tanuki Chess is one of the most popular games, and is growing in popularity, exactly because people are starting to lose the mentality you display here: a 12 year old thinking it's edgy to call board games boring because it's not in a videogame.
@@RedFloyd469 tl;dr
6:06
Randomness has a "roll" in game design.
Nice.
Dice*
I was about to correct your grammar and then I realized- good pun lol
What game is that?
@@lezbadeez3692 Dicey Dungeons
"optimizing the fun out of games" Basically my problem with every multiplayer game at high level. I don't like it because it becomes a contest of being the most efficient at executing the same optimal strategy.
That and assholes who shove people around because they don't have the same amount of hours in the game as them.
This is why I quit League. I'm just keeping my account so I can rip champion and skin models. I don't see what there is to actually achieve in games that are heavily multiplayer dependent.
Thats why I typically like to try and find a strategy that specifically beats that one. Even if it isn't "Good" if it beats the #1 its my favorite way to play.
@@battleland7263 and watch as pros finally showed everyone how to abuse it, followed by everyone using it in all situations(even if they lack what made the pros successfully run it)
I feel the pain, but the fact is, you can pull a lot of victories using very wild strategies, because your opponent has no plan for it, as they are expecting you to use the "optimal" one.
Randomness is fun when you just passed a save point and engaging an enemy/event that presents just enough of a challenge that if randomness makes it harder, you just adjust your strategy and make up the difference.
It's not fun when you're on the third stage of a final boss that is already immensely overpowered and they use a multi-attack that lands with critical damage and all of a sudden you've been put in the tailspin of having to heal every turn. That can get real annoying real quick.
The talk on probabilities in civilizations was really interesting, glad you mentioned it. It's funny how people think that with a 1/3 chance, if they lose the first two, then obviously they should win on the third try. Although, I remember thinking that way as a kid myself.
I never knew that about Civ, and you better believe I'm gonna abuse it in the future.
just gotta store up a ton of bad luck then go in and win the lottery.
Incredible insight, it's reminding me of how screenwriting is often taught, i.e. something must be unexpected. Whether the unexpected elements come in at the front end of a story or at the back end, it's always a reliable way to give your story gravity. Additionally, the part about information flow is very reminiscient of dividing the narrative into scenes or acts. Making sure new information is drip-fed is a fundamental element of successful storytelling. Surprises are necessary to keep audiences engaged, and the same can be said of the randomness you discuss.
An immensely good video, one that I find myself returning to for.. probably the 4th time now. I went from making games in my spare time to actually studying game design. Much thanks to your immensely informative and engaging videos. Thank you
Output randomness: Everyday there's a tiny chance a new GMTK video is published. If there's no video, no matter, I can wait. If there *is* a video, YEAH!
It's input randomness. You act *after* the fact whether there's a new video. Also, releasing a new video is anything but random. There's no chance he can make 1000 videos a day. It's also unlikely that he'll wait years before releasing a video, unless he quit this job.
@@y.z.6517 No need to take it seriously. It's just a silly joke
@@y.z.6517 Anyway, thanks! You may be correct and I have something to think about
I've got to say, the game that taught me about randomness and risk mitigation was Blood Bowl. The core mechanic of the game being that you and your opponent alternate turns, but your turn ends abruptly if anything you do goes wrong. So you've got to perform actions in a sequence to maximize upside and minimize downside.
surprisingly enough: Blood Bowl demonstrates how you can have ridiculously output randomness which also are crazy punishing and still be a fun game.
Praise Nuffle!
The turnover mechanic is brilliant in BB. I also think it's worth noting that BB allows players to buy a small number of precious re-rolls which can be used to counter the worst rolls, and all this combined with a full roster of units on your team to control -- and all the chaos that can occur each turn for both your team and the opposing one -- gives players a boatload to think about each turn. Plus it can be hilarious even when we're really unlucky. I think that's often helpful with games that chaotic from the RNG is that they're sort of funny and generate lots of amusing stories to tell.
0:58 "Do we just like luck when it lands in our favour, and hate it when we lose? No."
17:40 "We found that if there's randomness where you're expecting something bad and then you get something good, no one ever ever complains."
The second citation is about output randomness, not randomness in general though.
Man, i appreciate the fact that you talked about Keith Burgun. I really like his discussions on strategy and tactics.
8:53 "carefully tuned output randomness" the importance of this tuning, especially in competitive multiplayer games, cannot be stressed enough.
One example of bad output randomness that comes to mind is the random crit system of Team Fortress 2. In theory, it was created to break stalemates and give players of any skill level a fighting chance, but in practice there are better mechanics that solve stalemates, and it ends up helping the players that are already doing well (since your crit chance is increased based on damage done recently). Uncle Dane's video "Remove Random Crits From TF2" goes a lot more in depth on it, and even touches on the fact that even the player dealing the crit can feel unsatisfied about it (I understand, though, that 17:26 - 17:45 is about single player games).
Also, it's worth noting that competitive games already have an irremovable random factor: the other players. The way a player acts, reacts and how they expects other players to act has enough randomness. Using poorly implemented output randomness here can severely undermine the players' skills and mind games.
"In theory, it was created to break stalemates and give players of any skill level a fighting chance" - on this note, you're actually incorrect- a developer message on one of the maps states outright the ACTUAL intent was to reward those doing well. "This creates those rare, high moments where one player goes on a rampage." "In short, the better you do, the more likely you'll continue to do well."
This was the developer intent, and sadly, it works out exactly that way. And it's really not fun. For either side.
@@NimonoSolenze You're right, I was misremembering it. Maybe I read somewhere that it had that function (on top of creating "'rare' high moments"), but the dev commentary says otherwise.
Which is a shame, because then random crits have one less good reason to exist. It's a mystery why they weren't removed after all these years.
League of Legends is rumored (I couldn't find devs confirmation) to have a biased crit chance that helps keeping random system fair - adjusting each subsequent roll based on your no-crit/crit streak so overall it adds up to your actual crit chance as fast as possible: if your character has 25% crit chance and first attack doesn't crit, it'll up the crit chance with each subsequent non-crit until crit happens, then reset; doing similar crit in reverse. It still keeps crit system random, but at the same time pushes damage output much closer to theoretical calculated value over much shorter period of time.
I was enjoying this video so much the end snuck up on me and now I'm kind of bummed it's over. I'm 4 lines away from 2 pages of notes from this video, there's so much value here thank you so much for sharing it! I can't wait to dive into more videos :D.
Considering that the entire Pokemon series is LOADED with output randomness, I'm shocked that you didn't even give it a passing mention.
As a competitive pokemon player I felt that
Pokémon has input AND output randomness. The whole game relies on randomness for every player.
It would be the perfect example, and yet...
@@polkadi ...its not metioned, and surprising fun
"If it's not 100% accurate, it's 50% accurate"
@@SemiHypercube yes
Drinking game: take a shot every time Mark talks about Splunky
Take a shot every time he talks about Slay The Spire and you'd be dead.
rain world absilutely deserves an honorable mention here!
the world itself is static and meticulously crafted, and the mechanics are very minimalistc. but all other creatures, most of which concider you a tasty morsel, and some of which you can eat, are all free roaming AI with vastly different behaviors, temperaments and intellects, and they fight over territory prey and surviva. .. or your body. things never play out the same twice, and you are always on your toes.
that one really opened me up to the mindset of hardcore gaming
(7:22) "pre-luck" and "post-luck" feels like better terms.
"input-randomness" sounds like you're inputting something into a randomiser, and "output-randomness" is the randomness you get from the randomiser. Both "pre-luck" and "post-luck" would be the result of said "output-randomness".
idk it kinda makes sense. the input randomness is what goes into deciding your next move or step,output is the output of said move/step
@@kablamo9409 I'm mostly thinking from a developer perspective, with inputs and outputs of functions.
@@Liggliluff Yes. from a programmer's perspective, we (are supposed to, in computer language) treat 'functions' and 'methods' i.e. getting a random number as machines, so input and output are bad analogies in this case.
@@jr.p.8429 So because this is about game design, where you have developers; avoiding input/output is a good idea in this case. Plus people are more familiar with pre/post than input/output. We say prehistoric, pre-launch, pre-workout and postpone, post-sync, post-event for example, not input-historic, input-launch, input-workout, and output-pone, output-sync, output-event. Pre/post are prefixes ("input-fixes") well known by the people.
@@Liggliluff well yeah.. because "pre" doesn't mean the same as "input" and "post" doesn't mean the same as "output". your input your credit card into the ATM, and the ATM output the money. you don't pre-put your card into the machine and post take the money. Same goes for the luck thing, luck doesn't equal randomness. pre-luck and input-randomness doesn't mean the same at all. pre-luck/post-luck doesn't even make sense in my head. luck is a result, it's not something you can have before (pre), or after (post). It's by seeing the result that you know if you were lucky or not. randomness is much better than the word luck in that aspect. Randomness is not the result but what lead to the result. Throwing a dice give you a "random" number and getting a 5 would be one possible output of the randomness. and THEN, whether you wanted a 5 or not, you know if it was a lucky throw or not.
Input-randomness and output randomness make more sense to me. Although "input" sound a bit weird because like you said, it almost sound like the player had something to do with it. But It is still better than pre luck which sound like the user "own" luck in a tangible way (a luck stat maybe ? But even then, it would only be a variable in the randomness, not the randomness itself)
Sometimes, developers can make something seem random when it isn't, or only make it partially random but have clear trends and tendencies. In Freedom Planet, almost none of the enemies and bosses behave randomly - when it seems random, it's usually the enemy doing something based on your own movement. Spade, a miniboss in the Sky Battalion stage, appears random at first, but with practice, he can be completely controlled. The final boss, Brevon, has a final phase that is only random on Hard difficulty. He seems random on lower difficulties, and still kicks your butt, but you slowly pick up on his predictable pattern and learn to use it against him.
Final Fantasy Tactics has plenty of randomness, including zodiac signs, hit chances, enemy secondary skills and equipment, and random encounters. However, the AI has some very clear tendencies that you can learn with experience, and a seasoned player can often exploit these tendencies, setting up situations to strongly motivate the enemy to walk into traps.
Chrono Cross has your typical JRPG steal system, with a few key differences. You can only steal once per fight, and you aren't guaranteed to steal something, but every battle - even bosses, even the final boss - can be fled in order to start fresh from the beginning. This means you can start fights with a steal attempt, and flee until you get the item you want.
Reminds me of some enemies in Slay The Spire which just make a sequence of predetermined moves, and repeat, in a non-random cycle. I wasn't ever good enough to take this much advantage of it, but I presume it's the way the game intends you to survive when you use the relic that hides enemy intents from you.
No games are purely random. In theory, it's possible to manipulate all RNG in all games, because of how computer works and how a "random" value is generated (Which is actually a bad term, the correct is Pseudo-random). It's possible to manipulate what stats your Pokemon will get and stuff like that, if you know how to.
The chrono cross thing seems tedious more than anything. Why even make it possible to fail if you can just retry infinitely until you either get it our give up from boredom.
"I never said the word epidemic so many times..."
Thanks for the jinx man xD
2:29 nice edit damn.
All of your stuff really, but i loved that and the opening of this video in particular!
Thanks!
I didnt even notice that! Thanks for pointing it out :O
When playing RPGs, I like dice rollers generally, but when I'm making a roll I care about, I do find I can accept failure better from a dice than from an app.
So, I use dice rollers when DMing, and dice when playing.
I sometimes think the physical dice are loaded or damaged when they don't roll in my favor for a long enough time.
@@takatamiyagawa5688 Most cheap dice are badly balanced. They're manufactured by the injection mold process, which often leaves air pockets in the final product. The only dice you can count on to be as unbiased as possible are dice which are carved down from a solid material, but those are a lot more expensive.
Another awesome video! I'm so glad you mentioned Hearthstone and esports in there, because that is a perfect example of a game that I LOVED the randomness - both input (card draw) and output (cards with varied effects). Some of my favorite memories in that game involved Yogg-Saron, which would cast a completely random spell with random targets for every spell you'd cast in the game so far - with often hilarious effects that could instantly win or lose the game, no matter the current state otherwise.
...until I got highly skilled and competitive at it. Then even minor sources of randomness became pure frustration, to the point of screaming at the screen and slamming doors when it "always" gave me the "worst" draws or outcomes and "stole" the game from me. Sometimes this was actually the case, but usually it just felt like it and left me constantly wondering if I made the right play and got screwed, or really wasn't as good as I thought.
I've always looked at randomness in general like a spice in games. I think you have to use it sparingly, it can't be the meal. A good chef knows the right way to season something. When you get it just right you get something special. I think games like Diablo 2 used it in a way that stimulated a lot of people just right. When you get it wrong you get an inedible mess. Like a F2P mobile game that just says "make it random" at every step in place of actual considered game design and satisfying progression.
"I've never said the word epidemic so many times" - Well that sentence didn't age well xD
Manfredo Stagno d'Alcontres why?
@@ringabell80808 I wonder.
It was a pandemic pretty soon though
You're missing a *massive* part of randomness here: the minimum possible number of bad rolls it takes to have irrecoverable consequences.
The problem in XCOM isn't the output randomness vs input randomness; it's that one bad roll can have game-killing consequences. If an enemy had 100 starting heath and the output randomness on a shot is 10-20 damage, that's an adaptability test -- especially if sampled without replacement (or with delayed replacement, or any one of many such systems). But instead you end up with enemies with a base health of 6 and a sniper with a 99% chance of 12 damage but a 1% chance of 4 damage, against an enemy that has a plausible chance of a 1-hit crit on your units. So one or two bad rolls -- with no chance to adapt between them -- can cause huge problems, which is why it feels so bad.
This is the same reason that FPS players are fine with randomness in the firing of automatic weapons but not sniper weapons -- over 20 bullets from an automatic the expectation is still narrow, whereas a sniper that can just randomly jam or completely whiff feels terrible.
Pandemic is an interesting example because the epidemic cards have a 2-bad-rolls-loss possibility, if you get unlucky enough with what's on the bottom of the deck. But the investment in a single pandemic game isn't that high, so it feels fine to lose to that randomness occasionally -- you can just play another round. Whereas in XCom that 2-bad-rolls can come after playing a save for weeks. (If mission consequences weren't so devastating on your team in XCom then it'd be fine there too -- you'd lose some unavoidably, but you'd pull together for the next one. But that's not how things go, so they just incentivise loading a save instead.)
yeah. he says like we should be happy about it just for being ~the facts~
tldr - realism isn't fun.
The problem with output randomness isn't output randomness. It's fail states.
Imagine XCOM, but instead of "70% hit chance" you have "30% suppression chance" so 70% of the time you deal damage (same as now), and 30% of the time you don't deal damage, but apply a debuff to the enemy that makes them weaker on their turn. The fact that missing is no longer a pure fail state changes the dynamic.
By changing the fail state to a less beneficial success, you're making players feel more agency to their actions.
@@a-blivvy-yus great idea! Keeping the sense of impactfulness reminds me of alternative uses of cards, or dice manipulation
@@revimfadli4666 The idea is based on an old tabletop wargame I barely remember playing almost 30 years ago. Don't know the name, but when you attacked enemy units, your shots either inflicted damage or suppression, with both having value in different ways.
An iconic form of Output Randomness: The dreaded/beloved Random Crits in many video games, such as the often hated on ones in Team Fortress 2. One Crocket can wipe a quarter of a 12 man team or more if placed well, but many players will get easily tilted or even enraged over this even if there's only a single kill made out of it.
Now I'm gonna hide, cause speaking of said crits often sparks controversy borderlining a war.
Random crits are especially frustrating in TF2 because it's a game built around player skill. Whether you miss or hit a shot depends entirely on your own aiming and the other player's dodging, but you have no control over a lucky crit that randomly ends your life.
@@katerwaul9574 On top of that, contrary to what the video says about randomness in multiplayer games being meant to lessen the value of skill in a game so everyone has similar chances of winning at all skill levels (something I vehemently disagree with being a good design choice), TF2's random crits were designed around rewarding the skilled, as stated in-game by the developers. They wanted to create a sense of specific players being able to go on rampages, so they made crit chance go up the more damage you're doing over a specific period of time. Crits do high damage, especially crit grenades, stickies, and rockets, so crits with those (especially the rockets) tend to snowball into more and more and more crits until there's no more targets to hit.
Uncle Dane has a great video on random crits in TF2: th-cam.com/video/WHvwijT2ss8/w-d-xo.html
(yes I am passionate about random crits in TF2. yes, i do believe they should be removed BECAUSE of the explicit intent to turn skill or even just random chance into a snowball team wipe. it's not fun for anyone.)
@@katerwaul9574 I mean, Crits are BS but it's not really true you have no control against them. You still need to, you know, get hit by them in order to actually die from them.
This is nothing. In D&D 5e if you play as a crit oriented paladin you can one shot creatures that should rekt a party of four players. You have than one in twenty chance of evaporating almost everyone on your level. It is more powerful than crit rockets in TF2 and can make a deadly encounter a piece of cake.
@@GrimmerPl As someone who is a DM, has played exactly this and made a Paladin/Monk One Punch Smite character, I am proud of you for reminding of this. This is indeed fun. Additionally, the Sneak Attack of Assassin Rouge's also work.
One counter to the dice argument I’ve seen: when watching a stream of Dicey Dungeons I noticed a lot of people putting extra stock in rolling a six and getting annoyed when the streamer didn’t respect a six roll even though the function of the first two characters are to make you value high rolls and then tell you a six is just as rare as a one
I'm pretty sure the second character values *low* rolls (assuming we're talking about Thief here).. though yeah, having a build which can manage both low and high rolls is very good. (Though Warrior's Combat Roll basically does that for you, but to be fair - he's the easiest character.)
15:33 This is where the math "knowledge" of the average player sadly stops. 33% chance to win, I lost 2 in a row, I should win the 3rd, no matter what. They'll throw the mother of all hissyfits if they lose, but we know for a fact they won't bat an eyelash if they win five times in a row with that same 33% chance. Explain that brilliant logic to me...
The problem is probably that real world parlance (how people actually speak and what they mean) and mathematical probabilities (how the numbers are calculated and what they mean) are an entirely different thing.
Games express mathematical probabilities of more or less independent events, but players think in terms of success rates. Hence, mathematically speaking 33% chance means that you can fail thousand times, and it is fine, but in real world parlance that is completely wrong. You cannot have 33% success rate and miss thousand times.
I think it is games that get this wrong. They should speak the same language as players.
My god, I always wondered why I was never super unlucky in Civ
@@janimiettinen74 Your argument is that math is wrong about math. I think it's more accurate to call the players ignorant instead. They do not understand math. Period. If they did, they'd know that a 33% success rate means even after infinite attempts you are NEVER guaranteed to have succeeded. In fact, every time you try, you are more likely to fail than not fail. Period.
@@dontmisunderstand6041 My argument is that "30 % chance" is ambiguous expression. It can be understood in two ways.
"30 % chance" is not exactly a mathematical expression because it does not specify in any way what that 30 % chance pertains to. Hence the ambiguity, therefore the misunderstanding.
It's weird to say Hearthstone "allegedly" has a pity timer, because it's been widely proven to exist for years now.
Allegedly just means the devs haven't out and out confirmed it yet. You can't expect them to drop all their secrets on demand.
@@AmeshaSpentaArmaiti They have confirmed it. They said the hard cap on the timer were 40 packs... Years ago.
They've also *allegedly* made a stealth change to the pity timer so it only works on the newest pack, so that take for what you will
@@SuperHipsterGamer Ah, never saw it myself, and it didn't seem like the kind of thing a company like AB would just put out there.
It's also possible that he was a point that he mostly just has out of hearsay, a fact that everyone knows, but he's not going to put in a few hours of combing in order to find that one patch note that confirmed it.
There is an old adage in military doctrine... All skill is vain when an angel pisses in the flint of your musket
Wow, phenomenal video. Thank you for all your effort in researching, writing, organizing and presenting this info. The result is both accessible and very understandable, without wasting any of the viewers’ time. Bravo.
Certainly a thing that helps when engaging output randomness is allowing reasonable ways to trade out a chance for a 100% surefire thing. I feel like Xcom would have been less frustrating if more hardflanks became 100% chances to hit more readily. The mario and rabbids figured this out by mostly doing round values and getting 100% chances vs enemies without cover against you.
Also holycrap sorting out good success rates that then end up being foiled by ANOTHER chance on a chance you can't readily see, like snakemen in xcom2 dodging, turning a hit into a grazing hit, and throwing your plan out the window. Though Xcom generally is just an awkward case with all of this because it wants you to be so aggressive yet your ability to react to new threats feels generally too low with how fragile your soldiers are.
I played XCOM for a while, it got really frustrating to play. The game constantly pushes you forward but punishes you for moving forward, your soldiers have way too low an accuracy (to the point where you're relying on 40-50% shots regularly), the enemies always get to move as soon as you see them and your soldiers go down fast and hard (which is especially frustrating with how unreliable their accuracy is).
@@Nukestarmaster yeah I agree. The mechanics are much more punishing for playing aggressively because of that. It probably would feel generally better with better accuracy (or just use the mario rabbids model, that one would make flanks godlike and thus encourage having brutal fast flanking teams all up together) or have your soldiers durable enough (or means to delay) so that you have a reasonable chance to deal with multiple pods before deaths roll in.
The timer was certainly a poor way to force fast play with how the game was setup.
@@Adalore The timer was what made me go "Fuck that, I'm not buying that" about XCOM2, Enemy Unknown was bad enough without an arbitrary timer.
Heh, i was fortunate enought to be able to play 2 on PC and having read about the timer beforehand the first thing i did before even starting a campaign was find and install a mod that removed it.
The Second Wave option that I think was called Absolutely Critical in XCOM EW changed flank shots to be 100% crit (not hit), which made the risk/reward involved in moving into dangerous positions to flank enemies much more interesting (40% just feels inconsistent).
I was watching my dad play Xcom 1 and the “Last” enemy in the kill all mission he was on was the Stealth squid thing(can’t remember the name) so he put all his units on overwatch and some poor thin man had dropped in the circle of death and instantly triggered all six overwatches at near point blank range, taking a total 20-25 damage.
And to help those who don’t play Xcom understand the intensity of this event.
Thin Men have only 3 health and Dad had endgame level gear.
That's what I dont like about the new xcom remakes. In the old xcom, they only spend as much time shooting overwatch as they need, and can shoot more than once
@@sol2544 Or someone drops way in the back and exhausts all your overwatch shots, which all miss. Afterwards someone moves at the front.
XCom is just so screwed up... a thousand badly thought out mechanics.
What I want to point out is that players report their bad experiences, like missing 5x 90% shots. Which then gets excused by "well it can technically happen". But XCom RNG is screwed up when it works in favor of the player, too. I had so many maps with about 60% hit chance, but 95% of my shots hit. Several maps in a row.
These are just 2 of XComs thousands of issues. When a game is bad, it's usually bad in several departments. XCom is no exception.
I just discovered this channel and I cannot have enough. Im addicted to this. So informative & inspiring. Thank you, this is gold!
Man
The idea of luck only being satisfying when you expect a bad thing to happen, but a good thing happens makes me reevaluate all my rpg design choices
Edit: thanks for this video I didn't know I needed
Mutant Year Zero: RtE is a good example of a game ruined by a bad "output luck" design : you have WAY too few % rolls per turn and each of these rolls has WAY too much consequences - worst of all, the kind of consequences you can't repair or at a very high cost which puts at risk the following of the player progression. Thus, perfect fights can be ENTIRELY ruined by only one bad roll, which leads to terrible frustration and having to reset all the time, instead of having to improvise to the situation.
exactly!!!! I tried the game and found it frustrating exactly because of that!
In every single encounter in Mutant you can achieve 100% to hit. I rarely went for a shot that wasn't 100% to hit during my playthrough. And if I remember correctly I had a lot of fights where by a proper setup I had zero RNG elements - just a clean execution.
Love it when I see great not well known game in a YT video, just feels good, seeing your favorite games
"That's how we think numbers work." Absolute Savage.
14:06 This was my main thought throughout the video - unless you’re playing a one-shot game, there’s no meaningful difference between “input” (i.e. anterior) and “output” (i.e. posterior) randomness. I think a better way of thinking about randomness is to separate it into:
- “Player-Random” - the game includes actions that can be taken by a playable character which do not yield 100% predictable results.
Example: moves with less than 100% accuracy in Pokemon; drawing a Development Card in Settlers of Catan.
- “Utility-Random” - where a game distributes utilities (i.e. resources, money, weapons, etc) that a player can’t opt into receiving (i.e. discounting when a player chooses to open a loot box), and some facet of the game’s distribution of utilities is random - e.g. when they’re distributed, who gets them, how many everyone gets, the nature of the utilities that are distributed, etc.
Example: resources yielded by dice rolls in Settlers of Catan; the location of “villages” (which give early access to good tech) and resources in the Civilization games.
- “Threat-Random” - where a game includes one or more NON-PLAYABLE threats that a player does not opt into receive (i.e. discounting cases like choosing to draw an extra card in Blackjack and going bust as a result), and some facet of the threat is random (e.g. when they happen, or their frequency, or their potency, or their identity, etc.).
Example: “Bowser fire” in the Mario side scrollers; summoning the Robber in Settlers of Catan.
You could further distinguish between types of “Player-Randomness” by comparing games where the player experiences a chance of error no matter what they do (e.g. passing error in FIFA), and games where the player typically/always has an error-free option but sometimes/always also has an option with a chance of error (e.g. moves in Pokemon, if the player ensures they have a move with 100% accuracy).
seems like bowser being dumb and only spamming fireballs on the upper area could kinda be seen as a "utility"
God I hate Catan.
Wouldn't one player's "player-random" action similarly be the threat/utility-randomness of another?
Nah, the "other" player would just be the recipient of the results of the "player-random" event. As defined above, utility-random and threat+-random apply only to non-playable facets of the game.
"84% chance to hit, 5 misses. How is that even possible?"
As a number-nerd, I haven't laughed as much as I had upon reading that for a long time.
anything lower than 90% chance is a risky no-go
Well TBF the chances of 5 misses are 0.01%. Which seems *_really_* low, but for anyone that crunched the numbers for card games... it's in fact 5 times better than getting a Flush in Five Card Stud. :-B
"Randomness" is one of the things that makes the dynamic ecosystems in Rain World work. Input randomness takes the form of the game's rain timer (which is decided at the beginning of each cycle) and enemy movement (which happens continuously). Players aren't given a birds-eye view of their current region - so the lizard that "randomly" pops out of a pipe to scare you is merely the result of your lack of knowledge. By understanding the specific traits of each ecosystem, players can more easily navigate what are essentially organic rube-goldberg machines.
This video is certainly going to get a bunch of new view now
I see you also immediately looked up gmtk videos about randomness
@@JonahHW literally my first instinct
I've never heard someone else talk about these concepts, so I ended up calling them front-loaded and back-loaded randomness when trying to explain the difference in various tabletop systems to other people, though there's sadly not many tabletop games that use input randomness that only use numerical dice.
Just wanted to say thanks, this helps answer some of the issues I am having with the game I'm developing.
I love it so much when you said "that's how we think numbers work" 😆.
I work in a casino dealing cards and roulette and I find it mind blowing how many people scream "THAT'S IMPOSSIBLE!" while running back and forth to an ATM when 5 red numbers hit in a row.
I used to discuss it with players but they seem to have it in their head that there's some kind of magic entity in the universe that'll pull that ball to a black number simply because of where it landed last and that 7 years of me watching the lack of consistent patterns on that wheel means nothing.
To be fair:
1) Gonzalo Pelayo
2) There is a mathematical strategy to always win, and that's to double your bet every time you lose (it just requires infinite money and a table that allows arbitrarily large bets, all in order to net-gain a dollar).
@@majorjohnson8001 I know, people get wrecked doing that all the time (in all games not just roulette) as the tables have limits.
@@Juice8767 Well. Yes. That's why I made the conditions I made.
5:10 "Drama is driven by the unexpected." What a great remark.
Spelunky: Mark's favourite game ever.
I always though about randomness in games as possibly being either horizontal, vertical or a mix of the 2, where with horizontal I mean that the different outcomes only add variety and don't change the difficulty of the current challange (or the entire game) in any way, while vertical is the opposite, it directly affects your chances of winning while not giving any variety, most random events are a mix of the 2