Could lower collision size fix pathing?
ฝัง
- เผยแพร่เมื่อ 30 ก.ย. 2024
- Original video: • Could THIS Bonus Idea ...
Live Stream ► / theviper
Website ► theviper.gg/
Daily YT Streams and Gameplay videos: / @dailytheviper
Other Games and content: / @moretheviper
TheViper Merch ► teespring.com/...
Donate ► theviper.live/
Discord ► / discord
Twitter ► / theviperaoe
Instagram ► / theviperaoe
Facebook ► / theviper
Business Inquiries: theviperbusiness@gmail.com
TheViper | Professional Age of Empires 2 Definitive Edition Player for GamerLegion | Sponsored by Re-Bo
![ได้เวลา! พาน้องแฝดฉีดวัคซีนครบ 1 เดือน🫣 [cc] แดนแพทตี้ SS2 | EP.52 |](http://i.ytimg.com/vi/IzeLtaXBLNE/mqdefault.jpg)
Thanks for covering my video, Viper, and great commentary! I'm thrilled as a newer TH-camr to have my content catch the attention of the GOAT! I used a pretty extreme level of shrinkage for testing, but I agree that it looks odd if units are stacked too much. Something like 10-15% smaller villagers could be helpful while still looking fairly normal. I'll probably do a follow up video addressing some critiques, but I should have the trade unit version out later today.
I would be very interested to see the effect of .33 or .32 tiles wideness (17%-20% reduction), so one villager can no longer block a tile alone. I think it wouldnt change that much in raw ressource gathering but could significantly reduce the bumping situation.
@AdmiralWololo Out of interest, why are you using an AI generated voice? I like the content but I can’t get myself to watch a whole video with just AI talking over it, did you ever think about doing a voice over yourself? Never the less I like the content so I am a bit torn in between 😅
@@Schulzenberger Funnily enough, I feel the same way as you do about AI, which is why I don't, and will never use an AI voice. My voice is naturally pretty low and admittedly can be kind of monotone if I don't warm up for a while, although in at least some of my videos you should be able to tell pretty clearly- the Aztec pronunciation video for example.
I do some minor post-processing (audiate's denoise filter to remove background noise), and clip out highly audible breathing noises or speaking mistakes as I find them distracting, but nothing that changes the core sound. I'm definitely open to suggestions regarding how to not be confused with an AI voice, of course, but the baseline is that my phone/recorded voice apparently resembles Obama and/or Darth Vader since I was 14 or so. 😮💨
@@AdmiralWololo I never thought this to be an AI voice. You just have a lower quality recording equipment than the richest youtubers. Still, you have a good voice.
I guess one could think it's an AI voice because it's very fluent and calm, yet quickly spoken, a bit like a professional narrator, which is not common for a TH-cam video.
Ok hear me out: Dynamic collision sizes for villagers. Shrink it down only when they are carrying their max capacity for the given job. This means that we still have the same work count per tree/goldpile/etc. but the pathing will work correctly on the way back to the dropoff. This also means that there are no problems with blocking vills when raiding or stacking when building a wonder (not the focus of pro games but you never know). I think this gives best results of both.
I thought of this as well. Didn't Starcraft 2 use something like this? If I remember it right, the workers did not check for unit collission, as long as they were assigned to mine minerals? Of course, this will be used in not intended ways, you could click a villager on a resource near where he should go, so he could sneak more efficient through enemy bases.
Something like having the current hitbox size when actively working on a resource (or idle), and a smaller hitbox when moving could work too.
Great idea ! Surely this won't introduce any bug at all !
Eh I kinda dislike that idea. It will look inconsistent that villages carrying resources suddenly take up less space. I think if there was to be dynamic collision, it has to be a constant dynamic collision, like units having a(smaller) collision box for colliding with units, and a different (bigger) collision box for colliding with anything else (buildings, resources)
Programmers are sweating reading this suggestion. The edge cases this would cause (what happens if the collision is increased while being near a building or other unit, for example) would probably introduce more bugs than it's worth.
If they decreased the collision size from 0.4 to just 0.33 there would be no way for a single villager to block a single tile. That alone could be a huge improvement.
I'd love to see a test between 0.4 vs 0.33 vs 0.3.
Maybe just go from 40% to 33%, so 3 units can stack on a tile and one unit can't block the entier time (two vills can always go on one tile, since you can't position one vill to block a full time), its small, but enough to fix the edgecases that triggers most of these bugs.
Or 30% just in case
I now need Spirit of the Law to make a video comparing normal collision (0.4), then 0.35, 0.3, 0.25, 0.2 and putting them through multiple tests.
How big is the effect on indeed going into a town center, on normal wood gathering, chasing with melee units, etc.?
Please, Spirit? Pleeaase?
Admiral wololo made it big. Congrats bro.
Admiral.....WOLOLO!
Still working on it, but it's an honor to be covered by Viper!
FWIW, Starcraft 2 does have a type of dynamic collision size.
When moving the collision size of units shrinks; when they stop it grows back and because units can push each other it causes a tight ball of units that moved to diffuse out a bit.
You can see it the most with a large group of Zerglings.
Granted, SC2 does have many mechanics related to movement and pathing that cannot be translated into the AoE2 engine, but it's still an example worth considering.
One other change could be a 2 values for collision size - smaller for ally smaller and larger for enemy ones
It would give benefit of more natural pathing for villagers and melee units while allowing bodyblocking for enemies
(It also makes intuitive sense, since troops prefer to keep a distance from enemy while fighting, while allowing friends to stay back to back.)
I have no idea what AOE pathfinding algorithm looks like though
This idea could could have significant performance cost, depending on implementation
Cossacks did it perfectly in 2001, with armies at least 10 times larger than AOE2. You could literally have ten thousand soldiers on the map without performance issues.
I think reduced collision size contains the danger of less aggression in the games, when it's getting harder to block units so they cant escape and the eco beeing super efficient it will be really hard to get other value from your military then map control if youre opponent is not super potato and doesn't notice early enough. For the eco and maybe large scale melee engagements this can ne great, but scout or maa openings loose lots of potential
Oh no, maybe siege options will have to be improved then!
He has something but Vipy is right that stacking is too much.
Feels nice when you control them but if you can't attack a villager because it's one pixel wide in the stack... Eh.
It also will make raiding waaaay more difficult if you can't block the path of the villagers.
Maybe the vills dont even need to run when they can just fight back with 5 stacked into each other and 2 shot a scout
They could have different collision boxes with friendly units than with enemy units. Then you can still block them but they wont bump into eachother
Tbh I think a little nerf to raiding is ok. Especially lower elo players often complain that instead of having 2 impressive armies properly fighting in battles, the game ends up with a bunch of villagers rushing your city. Small consistent colision reduction should be sane for everything. Also improving infantry where it suffers the most, making it especiialy good in dense urban/forest fights in comparison with cavalry which is accurate.
14:59 I don’t think anyone can use the realism argument for collision size when xbows can hide in a corner stacked so close that knights can’t engage
Yeah and that should never be possible either. I have always thought it was a bug and consider it as much bug abuse as the new 'stacking' of units that was discovered.
Archers are already too good IMO and fire are ludicrously high rates compared to how it is in reality. They don't need to be able to stack with the stand ground trick.
Realism is a bad argument for most things pertaining to gameplay. If we cared about realism then the archer line would be totally redesigned to have a slow-training archer unit with a high food cost that remains viable for rapid fire against low armor targets into imperial age, while crossbows would be a fast-training gold-intensive unit that fires very slowly but hits a little harder against armored targets. Crossbow adoption was largely about the decreased training required to wield one after all because traditional archery is a skill learned from a young age whereas crossbows shoot like a gun, but it takes more industrial capacity to produce crossbows in the first place.
It's always a tradeoff. sometimes game feel and balancing need to come first while other times realism is an important factor. I don't think most units should be able to stack but it's important for ships because of their size. I also don't want my villagers to just walk through buildings.
Just look at how often stupid people cause traffic jams IRL lol
It's always a tradeoff. sometimes game feel and balancing need to come first even though overall the goal is to stay on the more realistic side.
I don't think most units should be able to stack but it's important for ships because of their size. I also don't want my villagers to just walk through buildings.
Just look at how often stupid people cause traffic jams IRL lol
Giving melee units partial range helps with this a lot. Even 0.1 or 0.2 is a huge help for swords and knights, just to reduce some of the bumping. Spear line can get 0.3.
It's not as good as a full 1 range like lancers or Kamayuks, but it makes a difference and feels so much better.
To balance, making economic units more slippery with a smaller collision box would work really well imho.
I am sure the devs have considered all of this, pathing has been an issue for a long time. So hopefully they're actually working on a solution, otherwise there may be some other mechanic that fundamentally breaks the game, which is why they haven't done it.
I really like the idea, but I wish he experimented a little more with different sizes. 1/3 is just overkill and kind of hurts the identity of the game, apart from the things you mentioned. I just want workers to not get stuck consistently, even if there are just 3 workers on the same side of a lumber camp. Reducing the size by like 20% or something would probably be more than enough to fix that problem while not having such a big impact on the gameplay and optics.
Fair points, I chose 1/3 for testing because I knew it would have a very noticeable effect, but that's pretty extreme. The global change I suggested at the end was around a 20% reduction (.4 to .33), as you suggested.
I wonder how 2/3 of the size would work out. 1/3 seems really extreme. Great idea!
That's what he did. He didn't cut villager size TO 1/3, he cut it BY 1/3. Villager size went from 0.20 to 0.13, which is 2/3 of the original size.
@@kekker_ he says "I've reduced the hitboxes to 1/3 of the normal size, meaning their linear blocking potential is 13% of the tile instead of 40%" (diameter, not radius) around the 3:55 mark. either way I bet there is a middle ground between the current state and this to help blocking without making ridiculously efficient stacking vils.
I think it will be a good compromise if the villager has this smaller collision when they walking but when they gathering resources they have the normal collison size like now. So when a villagers chopping wood they don't stack like in this video because is so unrealistic looking and visually bad. But when the villager collect the full amount of rescource and go to drop off or task to walk away, attack, ect they become smaller so they dont bump in each other. I think it is good middle way.
I agree that could have the best of both worlds. But I can see it breaking if you gather a lot of villagers in a small area close to a node. Then task everything to start gathering. How will the game handle everyone suddenly has a larger hitbox. Would the push each other out of bounds?
Decreasing collision size like that would break blocking vills with scouts and stacking would introduce problems when attacked by mangos. Maybe if the collision size decrease wouldn't be that significant? 🤔 The first problem could be ameliorated with two classes of collision sizes: one for villager-villager interactions (smaller hit box) and one for villager-other units interactions (the normal hit box).
Villagers should have a secondary, smaller hitbox which is used only by villager-to-villager collisions of the same team. Keep the current hitbox for every other interaction. Easy!
This might be an unpopular opinion, but here it goes: I think some "erratic" behaviour of villagers is acceptable, if not desirable.
Well... villagers are not trained units as military ones, in the realistic sense of it. So, some villagers in the Dark Ages are gathering lumber when a group of an enemy military units arrives. There will be chaos. Yeah, I know not everything in AoE is realistic (herdables reporting on map exploration, for example), but I think we shouldn't deviate too much from it.
Now, in the economic side of things, it is the player's responsability to look after its economy and its efficiency. Build more lumber camps, mining camps if necessary. Some villagers got lost along the way to a far-away location... seems reasonable to me, even though it is not specifically done by design.
The low collision of villagers, IMHO, will have major impact in how the game is played. First thing that come to mind is trying to deny an enemy building, lets say a castle, while the villagers are stacked. You won't be able to target-fire them. You can, one by one. In an army of 20 knights, maybe you get two of them hitting ONE villager in a stack of 10. Horrible.
This villager stack thing reminded me of AoE4, and that always looked odd to me.
Thanks for sharing this viper hopefully they listen
IMO, more Daut castles are possible with this change when all villagers building a castle are stacked up and a mangonel shows up 😂
Also having 4 more units on a tree instead of 2 being just a minor footnote, that guy may not have much knowledge in terms of balancing their proposals.
But the initial idea with unit size may be something for the devs to think about.
Seems like he is primarily interested in maximizing the theoretical efficiency of units.
It was a well-researched video, but it lacked in the discussion on balance implications and edge cases.
I still think the idea has some merit and wouldn't mind a small decrease to collision (nowhere near as big as his suggestion).
Steppe Lancers & Kamayuk are looking forward to this change, also Chakram Throwers and the whole Mangonel line =)
My idea fron a while ago, which I think should not cause performance loss, is to have a "hard" collision box and an "avoid" collision box. Units will path around each other's avoid box normally, until they are unable to find a path to their target. Then they both switch to their smaller hard box, likely moving past each other.
Ah, and there at the end someone has already suggested this idea.
I would definitely research pathian tactics!
Wololo is a seriously underrated channel, he has some real bangers
It's very easy to protect a wood line with a single tower, wall in villigers, etc. If done wrong this would make aoe2 less interesting to watch
Is collision used as a hitbox for archers as well? That would be horrible
About buildings:
It would be neat if AOE2 got an update to building that does the following:
- Change how buildings like drop site look by adding the droppable as decorative around the actual drop off site (i.e. still occupy a 2x2 tile but a mining camp can have the gold/stone around it like just decorative lush so visually it can make sense friendly units can walk within it).
- Build connecting road around it.
- More buildings like town center that can be walk underneath/within. I like and dislike AoE4 buildings that can be build tighly and not act as walls so I don't know at the end.
Dynamic size:
- Villager size when marching is Y - say 1.
- Villager size when working is X - say 1.5 (so they don't stack).
- Marching unit size is Z - say 1.1.
- Fighting or not in formation unit size is A - say 0.9 as they are not a peasant rabble (so they can fight without blocking each other).
- Add push priority, so when unit bump into each other SOME like cavalry can move infantry to the side, and so equal size unit can push each other and they get back into position UNLESS they are on hold ground stance. Also this would push villagers if they stack on the same resource. Units will try to return to their original position if they were idle or commanded to get there before (like attack stance 2). Also add push priority for idle vs working so villagers can push slightly military units of the same player or ally.
- Push priority should not exist between enemies so they can body block enemies like it happens now.
I have previously made a workshop mod for AOE2 (2013/HD/Retired steam version) called "AOE2 Alternate" where, among other things, I played around with the idea of drastically reducing collision of various units. Through testing, however, I discovered a glaring problem, not only are units "harder to catch" for enemy melee units, but they also become a lot more difficult to hit with ranged units too.
On the other hand, reducing it by a small amount doesn't improve things in a noticeable way, any less than 20% collision reduction didn't feel like it fixed villagers getting stuck>idle or carts bumping into each other (it probably alleviated it on some scale, but it was no fix)
That's pretty interesting! I'll be doing more tests with villagers with less extreme shrinkage, but I wonder how things are different in DE. I know there are some pretty big differences with trade carts (and collision generally) between HD and DE.
@@AdmiralWololo another thing to consider is their movement speed which maybe should be fractionlly reduced (i.e small nerf to wheelbarrow+handcart from 10% ea to 5% ea) to compensate for that extra efficiency of a reduced hitbox. i think the sweetspot will be the 0.30 mark where 9 vils could fit on a tile comfortably- same for all infantry units. stable units should be less than half, 0.45, trade carts/cogs maybe 0.10, ships would be annoying to balance, but tbh i think their models need to be reduced.. galleons look too big, like a castle looks too small?
if all the numbers were revised reasonably, i'm sure pathing vs efficiency vs realism could find a golden ratio
Is it possible to implement this new collision size only while the villagers/infantry are moving? After they start to work or reach their destination, they go back to occupy their previous size, so that the economical balance remains the same? Change completely collision size would mean rebalance the whole game, since times would change dramatically... We just want solve the path issue...
14:34 a little correction: when workers are tasked to mine or collect gas (aka mineral walking) they don't lose their collisions, they'll simply ignore all sort of unit collision. the difference is that while workers will walk through any unit (including themselves), other units are still unable to walk over them, u can see it works like that when u raid an enemy and ur melee units like lings or zealots will still have a hard time getting in the middle of the mineral line, while the workers still walk in a perfect line.
Could lower collision size [TURN AGE OF EMPIRES INTO STARCRAFT 2?] -- Yes, yes it could. Sacrificing collision for pathing is a trade-off that didn't help make sc2 any better than sc1, it makes individual units less valuable.
I dont think that dynamic collishions size is possible on current version of aoe2 engine. The AoE2DE is actually unstable. U can face random lagging or game crashes or smth else and they're (devs), probably, dont really know why does it happen or where is the source of problem. And if we think a little bit we face the problem that we need to track ~280 vills and their condition in game and we need to know condition of EVERY object on map that theoretically may affect to their path (trees, buildings, other vills, other units, animals, gold and stone tiles and even technologies). That is supermassive amount of data for calculation and i dont think that even rewritten engine of aoe2 can do this right now correctly. And we just spoke about 1v1 game what if we gonna talk about 4v4 game with ~1140 vills and map is bigger that means match match more objects for intaractions. I'm pretty sure it won't improve the stability of the game. Perhaps the best way (for now) is to just reduce the size of the vill collishion by 5 - 6% until they get to an acceptable result.
Take a look at the original Cossacs game from 2001! There an army of 500 was considered a relatively small army, and you could have many thousands, or even tens of thousands of soldiers on the field, with better graphics than AOE2 had at that time. In 2001! Very long distance pathing was sometimes an issue, but short to medium distance pathing was perfect and smooth. And indeed they had very small collision boxes.
The solution is small collision boxes for friendlies, but still keeping the larger collision boxes for units of opposing sides, so that melee units can block enemies from simply sliding through.
"Realism of having a body"
Okay, but, Vipey, people can also turn their bodies to the side so they have a narrower profile and don't run into each other, when passing alongside each other. While it's somewhat realistic that people would have to walk around the others when working in a relatively small area, well, people ARE able to do that. Also, currently, the villagers occupy a larger space than they actually look like they do. A lot of units do. Almost all units can block an entire tile by standing in the middle of it.
Collision size kinda needs to be reduced across the board for melee units, this much I agree with. Especially sheep. Sheep occupy a freaking MASSIVE area.
As for walkable dropsites... Well, it wouldn't actually be a bad bonus, make it an unique "storage tent" that sort of looks like a TC section.
Dynamic collision size the way as described seems impossible, what you CAN do is change the collision size of different kinds of vills. For instance, lumberjacks, miners and builders could have a reduced collision size so they don't bump into each other, while regular untasked vills and farmers have their regular size so they're easier to block and hit.
Trade carts could be changed so that carts that aren't carrying anything have a much reduced collision size, which would actually massively reduce the bumpage because one of the primary reasons for the bumping is carts taking the exact same path in opposite directions. Reduce the unloaded Carts' size and that bumping problem instantly becomes way smaller because the loaded carts will just be going in a straight single file line.
Reducing villager hitbox size is okay; everything else he suggested seems like a terrible idea. Hitboxes should NEVER change civ to civ; it is a core mechanic and needs to be universal. No civ bonuses to hitboxes allowed imo. I dislike the idea of walkable dropoff points too. Changes raiding and walling dynamics too much.
The thing is it buffing all other resources but farms which makes clowning strat even better.
The game should maintain 2 workers/(per tile per side) gathering on average just for balance purpose (yea, you can make 3 workers work on 1 tile of gold but its not convenient if you don't babysit it and it's fine).
Basically thing change nerfs the idea of plans that extent later then 5 mins into the future: nowdays you can task 20 vills on stone but you won't have castle ready in 1.5 min after it cause it will be not really efficient. While clever player nowdays will task 8 vills and get hes castle 4 mins of gathering convinietly (for example if he sees enemy planning castledrop in the future and not in the last moments when villager almost at the march to his base).
However on the bright side it nerfs bodyblocking in small microfights with 1-6 units involved and makes the game more strategy and macro-oriented rather then micro-battles in early feud. It might be only me but in that case it's a good change (when insane micro can win you the game even with lack of thoughts or right moves...in rts!).
a big part of what encourages expansion in aoe2 is villager bumping that reduces their efficiency and makes you want to build a second and third lumber/mining camp. really the only problem i have with villagers' collisions currently is that they inconsistently get stuck indefinitely. This creates situations where both you and your opponent tasked 8 vils to gold, but in practice you only have 7 gold miners working, so you effectively lost a villagers due to pathing while your opponent who has done the same thing didnt.
I think in voobly villagers rarely if ever got infinitely stuck while gather gold/wood, so, DE devs should really look into what how the pathing there has solved this.
At the end of the day any inefficiencies you experience is also experienced by the other player. I've seen so many "ugh pathing issues" that look more like miss clicks or don't realize the optimal (usually shortest) path they sent their villagers requires back tracking through an enemies army. The pathing algorithm (likely) doesn't account for danger or another player's ability to realize they can trap your units. I feel like this is a need to over-optimize a fairly solved game. Do you really need the extra 50 wood per minute because villagers bump each other or the odd villager stands still for a minute or two to have fun or play effectively? I really doubt it.
what about vill war?
the main issue: game calculate path at once. You click to another side of the map and it already knows where is walls.
with gathering resources- it know the path to lamberjacks and stick to it.
anyway, reducing size it would harder to block vils and kill them. ALthought 5% is small number, less skilled people wont able to execute it.
I think they should reduce the collision size by a big amount like his 30% BUT only when the villager is carrying resources, when it is just walking empty handed could be 10% for a small general fix
If the vils were more efficient it could lead to larger armies in late game which could be fun. Might allow for armies to include a more diverse set of units or perhaps easier / faster tech switches into a different unit. In team games it could also help with players slinging if they're in a safe position which would help keep weaker players in the game or help a flank player try some crazy strat since the pocket can sling a little more. I dunno, seems like it's worth some experimentation at least. And 100% agree on the trade units; like maybe have their collision size reduced if they're traveling more than 15-20 tiles.
We could have a few civs that would have walkable res drop-off buildings as a bonus, either some new ones or just some of the existing civs. There are already Huns that can't build houses and Georgians & Armenians with mule carts, so the idea is not that far off.
Eco and melee unit hitboxes could then be a tiny bit smaller globally.
The walkable mining/woodcutting camps would work very well as a unique bonus for a new civ with a unique architecture style. Off the top of my head, I know the pre-colonial Maori used storehouses raised on sticks
Seeing how problematic low collison became in sc2, where armies practically group up together into a deathball and the fights usually end in 5-10 sec, I'd rather have jankier pathing.
Too much collision reduction would make AOE2 feel like AOE3. Like Viper said someone in between would be perfect. Like Viper said, you still want it to be somewhat realistic.
I'd very much like a smaller collision size for vils.
But I'd also like the 'stand ground' trick to get patched out, where units (esp archers) can stack on top of each other.
I have always thought it was a bug and consider it as much bug abuse as the new 'stacking' trick of units that was discovered.
Archers are already too good IMO and fire are ludicrously high rates compared to how it is in reality. They don't need to be able to stack with the stand ground trick and hide on a single tile in a patch of trees or something.
While this looks nice, and bumping is frustrating, i feel like this would sacrifice too much in terms of readability in the game, especially for beginners. Walkable dropoff points will make no sense for a beginner, if mills markets, and military buildings arent also walkable. (These buildings also often have actual walkable ground in their sprite)
Also the stacking is just too much, i like being able to see how much eco my opponent has, in the blink of an eye.
Lastly this is only a band-aid, the pathing itself will still be poor, only less noticable. Villagers will still take less optimal paths
You can have best of both worlds by having reduced size only while moving.
It would massively help pathing, but not create funny stuff like several people swinging an axe through each others heads.
20% reduction in vill collision radius to 0.32 down from 0.4 will ensure that a single vill can never block a one tile gap which is especially good in case of wood vills getting idle if I had to guess. I've GGed a non zero amount of games in the dark age after discovering that my second or third wood vill has been idle the entire time when I go to build my second dark age building and am short on res.
Damn I would love to see a bit of that reducing collision for infantry, that could be what they have and knights don't :) It could be a nice civ bonus/unique tech for villager too if we needed more civ 11
I agree tho that realism aspect is very important to the game visually
Slightly adjusting collision size(s) makes sense to me but as a civ bonus... hell no. Less villager-, infantry- and trade-bumping would be nice for sure.
Doesn't like that at all. It doesn't solve the fundamental problem. In Reality, someone would make space to let the other one through. Could be tough to design and implement an algorithm for that, but it should be doable. Atleast us humans can do it ;) As long as there aren't hundreds of people pushing through a small corridor..
the realismpart of viper is a good point, I'm fine if they stack un groups of 4 or 5 on each other, but just inside the TC, while creating new vills, not while working, choppa-ing the heads of other vills
His change in the video is drastic, but something as smaller % change can drastically improve the issue, without endangering the game realism or creating an exploit in the game. BTW walkable lumber camp is good idea, while the villagers stacking like that is too much.
Smaller collision size does not solve the underlying problem of units blocking each other. There will always be too narrow corridors or undesirable compromises. This is also related to the pathfinder, which was already touched, fixed and broken elsewhere so many times in the past.
I believe the root of the problem is the stop & rotate motion sequence of units when they collide with another. The idea:
1. When planning the path,, assume that the units with the same target position have lower collision boxes (this ensures that units walk around cluttered corridors).
2. Do not recalculate the path when bumping into a unit with the same target position. Optimally, slow down before the collision occurs to make the movement feel smoother. Even halting for a few milliseconds and resuming after should be more time-efficient than rotating and walking a new path.
(or bring back the "Pathing" technology. I'm sure that would solve it! /jk)
the real question is targeting vills under walkable buildings with infantry or even archers....selecting the right target becomes impossible. You also couldn't target weak units easily.
I think SC2 does is right. Working workers (i.e. those that are collecting a resource, moving to the drop off point or back to the resource) have no collision. The game instead has a worker limit per resource to avoid stacking infinite workers.
Instead of reducing collision, the way to improve melee performance is to give them a little attack range. Kamayuk and lancer are so smooth with 1 range. If all other melee was given .5 range they would perform much better in battle
dude is overdoing all changes... relax bro, don't need to touch gather rate, otherwise will change ALL in game dynamics. Just minor decrease in collision box is enough.
Yeah the stacking shouldn't happen, I feel like this also has a lot more of an impact on balancing than you might expect, especially if you want to block enemies or a boar with a few sacrificial units.
I do agree that it's an issue and reducing the collision size slightly could work but you need to account for the balance of the game and also the realism it tries to convey.
As a kid I would go back and forth from playing aoe2 to stronghold crusader. Dealing with economy in stronghold was much less frustrating because the collision size of all units is zero in that game.
Villagers clumping up in a spot would work wonders for Hoang mangonels
Edit: walkable camps would be useless when walling, or assuring your vill security vs scout raids
My only concern with this is the mechanics of blocking villagers with military units. Like in dark age with your scout when doing a drush for example.
Alternatively, there are solutions to this using smarter pathfinding systems. Eg. If the dude you bump into front of you is going the same direction as you are, just wait a moment or two and continue walking instead of repathing like it does now, which ends up making a weird ass path. Then for the stuff like when the villagers are mining/chopping and trade carts, there can be a system to path a single time while avoiding crossing other villagers’ paths (think of circuit boards, and how the traces don’t intersect with each other). Then the vill uses that same path every time it carries stuff until the resource is depleted and it needs to make a new path. I’m sure this is a lot more complex to implement than I expect (when is it not?), but if done correctly it could solve the issues with repeated pathing like this for good.
I am unfer 800 elo, so it sounds nice for my eco, but trying to get better at archer mico sounds tough with that.
Though i suppose i could have a patrol hotkey i actually remember hahah.
Just addressing the melee unit idea. Wouldn't changing the collision size allow more units to attack at the same time? That would cause a balance shift for fights and probably cause a whole lot of other changes to get balance back to normal.
maybe a global change of 5-20% decrease in collision size for all land units will help pathing throughout the game without changing balancing by too much.
I’m only halfway done with the video, so this might get touched on, but remember that they actually increased the collision size for step lancers because they stacked too much and were op
this thing could really turn off MAA and Drush.... also it decrease variability of strategy and the effectiveness of the raids...
I think 0.15 (¼) is the best, that way a villager cant block a single tile and that alone is huge and you can fit 3 villagers in a tile instead of the usual 2.
To be fair, I think every Aoe2 game can be improved if both players had to bring two villagers to the center of the map and there was a villager dance -off at the beginning of the game
It really silly to think that devs who actually made the game have never thought about such a simple trick lol. You test the game inside out while making it, they most likely tried much more things and came to a conclusion that it makes the gameplay worse.
As a civ bonus it would be fine, but not as a change across all civs
I wouldn't mind giving this bonus to the Bulgarians, they need something
For greedy boomers vil pathing is an issue, for the rest agressive ones doeant matter. Let them complain to the "devs" 111111111.
How about a shadow effect, where we would see the 'shadow' of the troops at their current size while the 'real' size would be smaller and only the game would see it?
I'm happy to learn Chris Evans is finding voiceover work after Captain America.
I agree with Viper just lower it a bit and then it would be great, i dont like the walking through mining camps though, but apart from that, i think this has to be tried!
When the developers can't fix the problem they created that's when you know it's fluffed
Wait, why does the clip with smaller units look so good. It looks more realistic and immersie to my eye. Really cool!
I prefer good gameplay over realism.
Pathing has been such a bad problem for so long, i'm happy to suspend my disbelief for my vils to not get stuck.
If you're willing to degrade a quality of the game to improve another then what the hell does that do? Only makes the game worse long term, sets a precedent and also gives the green light to the devs to do questionable updates just to fix this or thart. Completely wrong imo.
@@PotDylan I dont see lowering a little bit of realism for making the game play better as a bad thing. I mean, we have roaming elephants that kill things, we have monks that get armies to change sides by waving arms and shouting at them and we have ships that hold at maximum 20 slim villagers OR 20 PHAT elephants. Realism is not exactly highest on the list for age of empires anyway.
If the vils look a little more stacked or close together, thats not a big deal.
@@bitemyassets It's not so much about realism, it's just not AoE... I can't pinpoint what it is but it bothers me to see villagers stacking like that...takes away some character I feel
I don't think realism is the right term for this. It's one way to approach a singular gameplay mechanic versus another way.
Reducing the collision radius has many gameplay implications, some of which were mentioned in the video (like the buff to siege and nerf to melee units).
Stacking can be especially jarring when you, for example, have 5 scouts and engage a single spearman but, surprise, there were actually four spearmen on top of each other and what looked like a great fight turns out to be a game-changing loss. There is no way for you to drag your mouse over a group of enemy units to see how many they are.
In the case of villagers you often want to target the weak ones, but if they're on top of each other you can't.
How about standing vills having standard collision box size whilst moving ones get smaller collision boxes.
Start slowly lowering until the sweet spot is found. That was we get continual improvement over a period of time.
What if one resource can have max one villager like with farms. Wouldnt that help?
There's potential for a middle ground option that will satisfy realism and practicality to be found here.
Trade carts just ignoring trade carts would be enough. So they're still raid able.
Agreed. Reducing collusion size is okay but not as much as making them look like stacking
1:30 anyways, can we talk about how fitting it looks that the units are at a smaler size? Looks way more realistic!
Imagine how many vills you can get around a boar, sheep, castle, or wonder.
the on in the chat saying reducing the collision size dynamically for units seems quite a nice idea if thats possible
Mom! mom! Viper is talking about a comment I left in a video, I'm famous!
Another suggestion is to give melee a tiny bit more range:
th-cam.com/video/ItQNlG7MR7o/w-d-xo.htmlsi=JPETtK1pQ7fUVe5y
Dude just seems hell bent on making it a civ bonus instead of fixing pathing
The overlapping villagers chopping wook look a lot like villagers in Cossacks.
If is walking reduce if working stay same colizion size. Easy.
Spirit of the law did show that trade units get smaller collisions with DE.
daut castle still exist as all vills gathered at the same position, 1 onager shot killed all the vils
smaller collision size of mele infantry means that celts will shred tc's even faster
make an african civ where vills cost less food and therefore are smaller 🌚
build wonder with 200 villagers stacked
As for title - ofc it could. It worked well for factorio biters.