This should be part of the Ride Test for the Photo section, where it checks the speed of the coaster going by that point, and how far the camera section is from the beginning of the station. Th faster the coaster is going, the more the photo is worth. The closer to the end, the more the photo is worth. Initial values would be zero each, and only changed when the coaster goes by the photo station. Might have to set it where the photo section looks for the slowest speed before and after the Photo section based on the max length of the coaster (to avoid a case where you put the Photo station right before the brakes so the first car is going 100 kph but the last car is going 10 kph).
This is the kind of statistics that is within reach of an excel sheet to optimize, so I took the math a few steps further to see how you can maximize your profit. In general: happier guests mean you can ask more money, and guests riding more rides also means you can ask more money. - If people ride only one ride of that type, you need a lot of happy faces to consider raising the price above 3 euros. Even with 70% happy faces, your profit will get only marginally higher at 3.30, and your profit will drop off if you raise it further. - If people ride multiple times on a ride of that type and the happiness does not change, the optimal ride price raises slightly, depending on the amount of happy faces, but not much. The optimum is around 3.50. - If people ride multiple times on a ride of that type and the happiness is random each time, the optimal price goes up drastically with the number of rides the guest rides, as the probability of a guest being happy at least once increases, and the probability of the guest being willing to pay the price therefore also increases. If guests ride 5 rides and they are happy 70% of the time, the optimal price jumps up to 4,90. With low happiness and low ride numbers, this remains close to 3 euros of course.
So... if you want to maximize the number of photos bought, do a flat 3 euro, but if you want to slightly bump your profit and you have a park with a lot of rides, you can go up to 3.50 but no higher. Seems simple enough. ...Certainly makes more sense than a five-euro "IT'S A FEATURE" bathroom ;)
@@WackoMcGoose And if you want to be lazy like myself, you could always just assume max happiness tier (since that's the goalpost for standard moneymaking anyways) and then base your values off of that. If you have few rides, charge the guaranteed 3.30 since that will always net you a buy. A medium number of rides would allow for a bump up to the 3.70 tier, and if your park is huge with lots of photo-able rides, the max I would personally go is 4.50 but a coinflip per ride with 4.90 is easily fine too (just slightly more prone to shite luck).
2:08 I love how you thought about this video. Constructing a rollercoaster with the part of the photo section in frame, while using the nice green tiles as background for text. And while you explain, the rollercoaster passe by from time to time. It's a small detail, but a really nice one and very appreciated.
Wow, I did not know there were 4 types of on ride photos. I had intuitively or accidentally stumbled onto the 3 dollar price through my years of playing, but noticed alot of guests saying I already have one. Thanks!!
I quickly whipped up some code to simulate a few situations. I don't think it's simulating 100% correctly the guest behavior as prices above 6.1 were never considered by any guest. It also doesn't consider duplicate photos, it only considers a guest entering a ride once and deciding if they want a photo. In any case, here's what I found: For 100000 guests with evenly distributed happinesses, 3 euros was the best, and it went downhill from there. For 500000 guests with normally distributed happinesses of mean 249 and stdev 50, 3.4 euros was ideal. For 500000 guests with normally distributed happinesses of mean 180 and stdev 50, 3.1 euros was ideal, but 3.0 was very close. For 500000 guests with normally distributed happinesses of mean 128 and stdev 50, 3.0 euros was the ideal. For 500000 guests with normally distributed happinesses of mean 100 and stdev 50, 3.0 euros was the ideal. So I guess that if you have a park with a lot of happy guests, you should charge 3.4 euros. If the guests are more gloomy, you should charge less.
Even if all the guests in your park are at maximum happiness, your expected profit from charging more than 3.30 won't exceed 3.30 3.70 * .875 = 3.2375 4.10 * .750 = 3.0750 4.50 * .625 = 2.8125 etc. So there's no reason to charge more than 3.30 unless you're feeling really lucky. The optimal pricing would be between 3.00 and 3.30 depending on how happy your guests are. Although, since guests can only have 1 photo per ride type, its possible if they ride enough times, they'll eventually buy a photo at the max price their happiness allows. If they ride a ride 8 times with a 3.00 photo, they will buy one 3.00 photo. If they ride 8 times with a 6.10 photo, there's a high chance they'll buy the photo on one of those rides.
Yeah i would assume that 4.5€ or 4.9€ might be the most optimal price in a park with happy guests since they have multiple opportunities to buy a photo
I did the math in my comment as well but here's what I calculated it out to be... At an average peep happiness >180 peep will 1 ride in each category before exiting the park - $3.30 2 rides - $4.50 (expected return is $3.87) 3/4 rides - $4.90 ($4.29 - $4.59) 5-8 rides - $5.30 ($4.79 - $5.18) 8-20 rides - $5.70 ($5.27 - $5.68) 21+ rides - $6.10 ($5.73+) aka peep has to fail a 12.5% roll 21 times for you not to profit here Assumes peep will remain at >180 happiness when riding each ride.
@@michaelt5188 That's a good calculation, but tbh I think whether it applies would depend on how the game calculates if a guest wants a photo in the first place, which the game does before it rolls the price range they'll buy it in.
It always surprises the child inside me how much you can charge for things in this game. I only recently started charging $2.50. I also assumed ride stats affected it (and not just indirectly by making people happier).
The combination of Marcel and Duerklinks contribution to the breakdown and mechanics of this amazing game is astounding. I learn so much from you guys! I dont want to minimize what anyone else has done, I know Bassie making plugins and X7123M3-256 making ride types!
Where can you find these plug-ins? I've tried to find internet resources for new rides and scenery but a lot of the websites are either down or filled with malware
My guess based on looking at the table is that 3.30 is the best price in a "typical" park, since the odds are negligibly different except for unhappy guests (which you shouldn't have anyway). But also that the revenue difference between this and a simple 3.00 is probably negligible anyway. Would be a little bit interesting to see some empirical results in some sample park, even if it couldn't be taken as representative of all parks. Or at least an expected value calculation based on SOME sample distribution of guest happiness, arbitrary or not.
3.30 would be 10% more income per photo for -12.5% as many photos sold for moderately happy people, so I'm still not sure it would be worth it. I'd only try it if at least half the guests were very happy
@@steepslopesmm2 Maybe 3.10 otherwise? I think that option, at least, is almost strictly better than 3.00 in practice. Also many others in the comments are talking about how re-riding withing the same category causes the -12.5% to not actually be -12.5% in practice, which supports 3.30 or even higher.
guests won't buy on-ride photos at all below 118 happiness and their chance of buying one increases linearly from 118 happiness to 242, so higher prices are probably better to a weirdly high amount
@@steepslopesmm2 While that's true, the question becomes how many guests are only moderately happy (or less) when leaving a good roller coaster ride, though? How much would 12.5% *of* moderately happy guests represent as a percentage of the *total* guests leaving the ride?
I often see guests riding the same ride over and over. This would also be important in determining the optimal price? Since each guest can have multiple chances to buy the photo this may favor putting the price at something below 100% chance.
Good stuff. I've always charged 2.50. Good to know I can up it some more. Also, you can't just look at the distribution of guest happiness in the park, but rather their happiness when they come off the ride, which significantly boosts happiness when they have the opportunity to buy a photo. I think I'm gonna run with 3.30, since my rides are just awesome.
Woah those categories answered life long questions, I always wondered why guests sometimes said 'I already have a on-ride photo' even though I could see some guests have multiple on-ride photos in possesions.
Logic's telling me that a pretty good way to do this would be to charge $3 for wooden coasters and water rides since chances are guests won't ride them as much, but for steel coasters to bump them up by a bit, maybe to around the $4.10 mark since there's a far better chance for guests to roll multiples. Of course, this depends on the odds of a guest even rolling a photo in the first place. If it's quite low it's probably better not to push to much and maybe go down to $3.30 or $3.50.
I haven't searched TH-cam before commenting this, but with your encyclopedic knowledge of RCT2, have you ever made a fully optimized park? No other goal than to exploit the games program through standard play to have the mathematically best possible park.
This gives me a weird idea for a park goal. Make one coaster from each category and charge 6.10 for all 4 on ride photos. After a while, start checking guests inventories to see if you can find one with all 4!
What I like to know is how much the excitement rating affects a guest’s happiness when they ride it. Maybe that’ll give me the edge to charge at least $3.40 per photo.
When it comes to pay-per-entry parks I charge $3.30. when it comes to pay-per-ride parks I don't really adjust it, if I do it's to $2.50, that way (particularly in parks lacking the Cash Machine) the guests spend their money where it really matters. In hybrid parks I typically change $3.00 for them, that way I'm compensating for the reduced ride prices a bit.
I think it's interesting just how much nuance is behind something as simple as charging for a photo. It shows just how intricate game design really is.
A couple seconds after you reduced the Go Karts to 1 and things started going absolutely insane, my computer randomly got a Blue Screen of Death, and for a split second I legitimately thought it was part of the video!
The question now is, does riding the coaster changes the happiness? I mean when a very unhappy person gets on the ride and when he leaves the ride but just before buying the picture, does he get more happy and maybe so much more, that he would pay more?
If the stats of the coaster are good you can see their happiness gradually increase over the course of the ride, unless the ride is too long, at which time it will start to decrease instead.
Guest: I’m not paying £6.10 for a on ride photo of Roller Coaster 1! **Rains** Same Guest: Here is £20 for that umbrella I didn’t want! Conclusion: Just charge what you like 👍
So what determines if a guest wants to buy a photo or not? is it exclusively based around the ride category grand slam or are there other factors at play?
It sounds like they always buy if they have the money AND get a hit on the random chance. What I'm surprised about is that the position on the coaster doesn't matter. It sounds like you could put it right before the first lift hill, and they'll still get the picture.
@@Inuvash255 he shoewd in an earlier video that the photo section doesnt even need to be on a part of the track the guests go through and theyll still buy it. So theyll just pay $3 for some scenery photos
That explains why my one obligatory inverted makes bank with photos. Only knew 3 types. I'll try 4.90 up from 3.50 in future, my parks are particularly pleasing to guests. Also, rides give a big happiness boost anyway.
Another great video, Marcel! When I watch some of your videos on some of the strange algorithms behind every system in RCT2, I start to wonder why Chris Sawyer made some of these calculations so random and cryptic. Ultimately I think it comes down to the fact that he wanted to model behavior of guests, and given that human behavior is so random, doing some somewhat arbitrary calculations that vaguely mimic reality makes a more believable simulation while using little computation power.
What a complex system! I am now quite curious what determines whether a guest wants a photo or not. I appreciate that you said to divide by two again, rather than to divide by 4!
Its crazy how on such an old game something as simple as the optimal photo price can be considered through so many variables that its almost impossible to compute. One thing is for certain, it should never be lower than 3.
I do hope that open rct2 has/gets a feature where I can see the total number of guests for each mood: how many are unhappy, happy, very happy, etc. Having a quick glance feature like that could really help inform what prices things should be at.
Not sure if I misunderstood but I believe you mentioned that all of this math aside, not all guests actually check if they want to buy a photo... Aside from owning photos from the same category of rides, is there any other math regarding what guests check if they want to buy a photo? Or do all guests run these numbers to check if they don't own a photo from the same ride category?
I ran the numbers for a single ride with photo booth. - if less than 23/31 (75%ish) of guests are happy, then the ideal price is $3. - Above this, the ideal price becomes $3.10, and as you gain more very happy guests it goes up further. - If all guests are at least happy and more than 17/33 (52%ish) of them are very happy, then $3.30 is the ideal price. - It never goes above $3.30 for just one photo opportunity.
The category segment of this video has given me a new money making strategy for when I'm starting a scenario. Try to have at least ONE of each ride category with photo section before building more rollercoasters, so that you may maximize photo income (as well as ride income) early and give yourself time to gain enough wealth to build a better park and more rides down the line.
As a kid, I always assumed that the true value of a photo was based on the speed of the vehicle going through the on-ride photo section, but here comes Marcel to prove me wrong again. :)
I'd love to see a park that implements all your knowledge of shortcuts and exploits. From the sky it'd look like a troll park, but it would likely be as lucrative as can be.
So something to consider is also your ride excitement and queue length. Rides generally leave guests happier when they leave-- I've never really seen a guest come off a ride particularly angry, but I have seen them become unhappy from a long queue and then only become marginally happy by the ride. Considering guests can't buy a ride photo every time, it seems to me it may be a good idea, in terms of maximizing profit, to raise the price to maybe something like $4.50? Because if you've got a lot of ride photos in your park, chances are they'll probably roll a number that lets them buy it at that price at least once. At the default price, which I think is lower than the $3.00 you list here, usually every single guest in my park ends up with a ride photo quite quickly...
it looks like its almost never a good idea to raise the prices oh but you make a great point. they arent gonna be able to get a picture on every ride, so it does make some sense to raise the price.
€3 seems cheaper than the on ride photos in most real life theme parks I've been too; but then, I'm not sure I've ever bought one, because they're so expensive. Do we know what the chance is a guest will want to buy a photo?
@@kotor610 Even then they were low. I remember on-ride photos costing € 8 at Efteling in 2003, only one year after RCT2 came out (and 4 years after RCT1).
I did some rough calculations, and it seems the best play is to price the photos at the highest value where everyone will buy them, i.e. if all or almost all of your guests are very happy, then price them at 3.30, if there are lots of guests that are somewhat happy price them at 3.10, and if there are guests that are unhappy then price them at 3.00. If even 12.5% of guests don't buy the photo then it's not worth the extra price I didn't include guests visiting multiple rides where they could get a photo because it would require knowing how many rides are in your park and how long the guests stay there, but for a single ride in a vacuum my method is the correct one. Since we're only talking about a couple dozen cents here anyways, the sensible play is probably to just set the price at 3.00 so you don't need to worry about average happiness or stuff like that
Even with just a single ride, guests will usually ride that same ride multiple times. You are losing out on a lot of profit if you just take the guaranteed money.
I think (if I understand this right) the photos have the souvenir flag, so whether a guest buys one is decided on line 1510 (at the time of writing) of Guest.cpp, which says a guest will buy a souvenir/photo when they have been on at least 3 rides and are happier than a random number between 115-242. Of course I am nowhere near as carefully researched in my approach, I ended up mostly charging £3.20 as that's the highest I could get away with before I start regularly seeing "I'm not paying that much for an on-ride photo" in the guest thoughts.
Video idea: You could show, how all of the properties of coaster are counted and all of the tricks, to make it better. I for example sometimes wonder, if putting breakes before the loop would help with the air time, if I manage the speed of cart correctly.
"It's nearly impossible to weigh all of these factors" Every mathematician, programmer, statistician, and computer scientist: "Hold our beers" *complex mathing noises*
I did the math in python, averaged over 10k tests. Charging anything over $3.30 is guaranteed to be sub optimal under all conditions. Price | Average profit based on happiness level $3.00 | $3.00 | $3.00 | $3.00 $3.10 | $2.73 | $3.10 | $3.10 $3.20 | $2.41 | $2.79 | $3.20 $3.30 | $2.06 | $2.87 | $3.30 $3.40 | $1.72 | $2.54 | $2.95 Conclusion
A cursory number-crunching shows that $3.00 is the best bang for the buck. Happiness Ratio [H] = (255 - Guest Happiness) Profit Ratio [P] = (% to buy photo * Picture Cost) Average Profit = (H * P) At $3.00, you average $3.00 per picture, since everyone that can, and wants, to buy a picture will. Even just going up to $3.10 drops profit down to $2.91, as the 87.5% chance to buy at low Happiness drops the average, as it's the widest "band" along the Happiness gauge. This does assume that Happiness is evenly distributed across all values, so a park with a higher average Guest Happiness could get more money. It seems that lower is, to a degree, optimal; as lower costs increase purchasing odds across all Happiness levels, meaning more pictures are purchased when the game decides to look at the price.
The question this video raised for me is, why isn't the mine train coaster in category 3? I remember it looking very wooden whenever I built it in my parks.
After watching this and weighing my own analysis into it I would say pricing at 3.30 is a plenty safe default for most parks of a decent player. Still get that 100% happiest, and still really solid for what are hopefully outliers of ':)' to 'meh' or less. Otherwise the math on chance starts getting REALLY over the top and 'if then' happy.
I'd argue that in a well functioning park you're ideal price would be $4.50 or $4.90, assuming you don't mind the happiness hit. I've been charging $5.50 so I'll be making some adjustments from this video because of that multiple on-ride photos category allowance I did not know about! Edit: I did some math and it looks like when you adjust for multiple rides in your park for each category, the ideal price will be around $4.90 - $5.30, assuming peeps will ride 3-8 rides in a category prior to leaving your park and remains above 180 happiness when riding your rides. At an average peep happiness >180... peep will 1 ride in each category before exiting the park - $3.30 2 rides - $4.50 (expected return is $3.87) 3/4 rides - $4.90 ($4.29 - $4.59) 5-8 rides - $5.30 ($4.79 - $5.18) 8-20 rides - $5.70 ($5.27 - $5.68) 21+ rides - $6.10 ($5.73+) aka peep has to fail a 12.5% roll 21 times for you not to profit here Assumes peep will remain at >180 happiness when riding each ride.
@@TheWatchernator I got into a bit of a rabbit hole on it and quantified the expected returns using probability :) Essentially the argument to be made is you have the opportunity to adjust for # of expected rides a peep will ride prior to exiting your park. Because there are varying % of success, each failed on-ride photo purchase will get another reroll chance when the peep rides another ride in the same category. The current recommended price only assumes a success rate from the FIRST attempt at an on-ride photo purchase for any given category. We obviously can have multiple rides in each category and can expect peeps to ride them as they experience your park. This is akin to being able to flip multiple coins at a time, but in this case, all we're looking for is just one heads result. Intuitively, you can see that the chance of NOT flipping heads diminishes significantly the more chances you get to flip the coin. The same applies here and the high level math is below... At an average peep happiness >180 peep will 1 ride in each category before exiting the park - $3.30 2 rides - $4.50 (expected return is $3.87) 3/4 rides - $4.90 ($4.29 - $4.59) 5-8 rides - $5.30 ($4.79 - $5.18) 8-20 rides - $5.70 ($5.27 - $5.68) 21+ rides - $6.10 ($5.73+) aka peep has to fail a 12.5% roll 21 times for you not to profit here Assumes peep will remain at >180 happiness when riding each ride. Once I clean up the file, I'll publish it for further scrutiny.
Would it be possible to create a park that could be stable in a condition where everyone is either very happy or very unhappy? Someone with happiness below 128 would consistently be mad about $3.70 photos, but someone with happiness above 180 would mostly be happy. Are there enough mechanics like this that you could create a park that could sustain people as both very happy and very unhappy simultaneously?
There's a small amount of random chance that a guest will just wander for a long time, never going on rides. I think you'd need a very optimal park design to keep ALL guests happy. Even when they have a map, all that really does is make their next turn aim towards their selected ride, it doesn't ensure perfect pathfinding.
It's quite rare for me to drop below 999 happiness in parks these days, exception being when overcrowded so I just default to and have to a while to 4.50, although I'll probably go up to 6 now given most guest except in pay to entry parks tend to ride several rides or rides several times. Rather just tank the minor happiness debuff. If I'm not capped on park rating it's almost always because I haven't got enough drinks/toilets/food so it's better just to slap down a few more of them around.
For unhappy customers, the price with the best expected utility is €3.00. For moderately happy ones, it is €3.10. And for very happy ones, it is €3.30. Basically, the base price of 3 euros is high enough that as soon as the guests needs to randomly determine if they want to pay, you're losing out.
I was kinda hoping there would be a variable of where the onride photo was actually placed. For example, guests would be more likely to buy an onride photo if it was placed at a more exciting part of the ride like the bottom of a big drop, rather than an onride photo that takes the picture at the brakes.
before i try it myself, assuming a park has very happy guests and the chart of chances guests will buy photos is correct, this is my prediction for an optimal photo price: 2.90 + 0.40 * min(x, 8) x is the amount of rides in the park and the same category that have a photobooth (yes i call it that for some reason). for example if you have 7 wooden coasters in your park and 5 of them have photobooths, x is 5 when calculating the theoretical optimal photo price for said wooden coasters.
tried to link a screenshot to a quick table i made, but for some reason youtube lately has been automatically deleting all of my comments that contain a link, so, guess i'll go cry about it.
seriously why cant they just disable posting comments with links? why instead say "returned error" when i try to edit my comment only for it to be gone when i reload the page with no explanation? fucking wish this site had a competent competitor.
I think the category could dictate how much you can charge. Lots of metal coasters means lots of opportunities for a guest to go and buy a photo, so the risk of a higher price is a little bit mitigated and they'll probably end up buying a picture eventually. Wooden coaster only having one ride in its category means you want to make sure anyone who rides it buys a photo since.. Actually, do guests go back to rides over and over? Id like to believe they just pick any random one and ideally you just want to have as many diverse rides as possible to avoid the duplicate ride halfing.
I've watched a guest go immediately to the opposite side of the part to ride the same coaster like 4 times in a row, I think I had an ad out for that one
The optimal price (Maximum of "{chance that someone buys it} * {Price}" ) is €3.00 for unhappy guests €3.10 for somewhat happy guests €3.30 for very happy guests Pick whichever the supermajority of your guests are (or go with €3.00.. it is only a 10% increase)
I think 3.30€ is best. A guest might not be happy on his first ride, but he might return later with a Happiness value of more than 180 and so the chance of him buying an onride photo (if he decides to) is 100% at that point again
1:33 I've just now learned that it's when a guests exits the ride, not when they get on the ride and for some reason I didn't question it. like why would they buy a photo of a ride they haven't even ridden yet?
fun fact i happened to test yesterday: your train does not have to pass the photo section of the track in order for people to buy photos
That's indeed true and a fun trick, I should've mentioned that in the video.
@@MarcelVos You mentioned it a few weeks ago ;)
@@MarcelVos At 0:00:49 the Log Flume is listed as a ride with a Photo Section, However its not listed in any of the 4 categories, how comes that?
@@basti2955 Oops, that's a mistake. Well it's a water ride so it's in the fourth category.
This should be part of the Ride Test for the Photo section, where it checks the speed of the coaster going by that point, and how far the camera section is from the beginning of the station. Th faster the coaster is going, the more the photo is worth. The closer to the end, the more the photo is worth. Initial values would be zero each, and only changed when the coaster goes by the photo station. Might have to set it where the photo section looks for the slowest speed before and after the Photo section based on the max length of the coaster (to avoid a case where you put the Photo station right before the brakes so the first car is going 100 kph but the last car is going 10 kph).
This is the kind of statistics that is within reach of an excel sheet to optimize, so I took the math a few steps further to see how you can maximize your profit. In general: happier guests mean you can ask more money, and guests riding more rides also means you can ask more money.
- If people ride only one ride of that type, you need a lot of happy faces to consider raising the price above 3 euros. Even with 70% happy faces, your profit will get only marginally higher at 3.30, and your profit will drop off if you raise it further.
- If people ride multiple times on a ride of that type and the happiness does not change, the optimal ride price raises slightly, depending on the amount of happy faces, but not much. The optimum is around 3.50.
- If people ride multiple times on a ride of that type and the happiness is random each time, the optimal price goes up drastically with the number of rides the guest rides, as the probability of a guest being happy at least once increases, and the probability of the guest being willing to pay the price therefore also increases. If guests ride 5 rides and they are happy 70% of the time, the optimal price jumps up to 4,90. With low happiness and low ride numbers, this remains close to 3 euros of course.
So... if you want to maximize the number of photos bought, do a flat 3 euro, but if you want to slightly bump your profit and you have a park with a lot of rides, you can go up to 3.50 but no higher. Seems simple enough.
...Certainly makes more sense than a five-euro "IT'S A FEATURE" bathroom ;)
@@WackoMcGoose And if you want to be lazy like myself, you could always just assume max happiness tier (since that's the goalpost for standard moneymaking anyways) and then base your values off of that.
If you have few rides, charge the guaranteed 3.30 since that will always net you a buy. A medium number of rides would allow for a bump up to the 3.70 tier, and if your park is huge with lots of photo-able rides, the max I would personally go is 4.50 but a coinflip per ride with 4.90 is easily fine too (just slightly more prone to shite luck).
I would say, you get the the most out of 3,00 till 3,30 euro.
I don’t even play this game anymore but I still really enjoy learning how complex this game really was
You should play with the mod of open rct2
I don't even play this game
Developers must be geniuses
you realize you didn't know anything about the game until you're grown up haha
“Our first guest is having a terrible time” made me laugh for no reason XD
For me it was "I am slightly disappointed 🙂"
The first guest is having a Six Flags day
Glad i'm not the only one
The mental image of being pissed off while riding a roller coaster is hilarious to me
2:08 I love how you thought about this video. Constructing a rollercoaster with the part of the photo section in frame, while using the nice green tiles as background for text. And while you explain, the rollercoaster passe by from time to time. It's a small detail, but a really nice one and very appreciated.
Wow, I did not know there were 4 types of on ride photos. I had intuitively or accidentally stumbled onto the 3 dollar price through my years of playing, but noticed alot of guests saying I already have one. Thanks!!
I quickly whipped up some code to simulate a few situations. I don't think it's simulating 100% correctly the guest behavior as prices above 6.1 were never considered by any guest. It also doesn't consider duplicate photos, it only considers a guest entering a ride once and deciding if they want a photo. In any case, here's what I found:
For 100000 guests with evenly distributed happinesses, 3 euros was the best, and it went downhill from there.
For 500000 guests with normally distributed happinesses of mean 249 and stdev 50, 3.4 euros was ideal.
For 500000 guests with normally distributed happinesses of mean 180 and stdev 50, 3.1 euros was ideal, but 3.0 was very close.
For 500000 guests with normally distributed happinesses of mean 128 and stdev 50, 3.0 euros was the ideal.
For 500000 guests with normally distributed happinesses of mean 100 and stdev 50, 3.0 euros was the ideal.
So I guess that if you have a park with a lot of happy guests, you should charge 3.4 euros. If the guests are more gloomy, you should charge less.
Even if all the guests in your park are at maximum happiness, your expected profit from charging more than 3.30 won't exceed 3.30
3.70 * .875 = 3.2375
4.10 * .750 = 3.0750
4.50 * .625 = 2.8125
etc.
So there's no reason to charge more than 3.30 unless you're feeling really lucky.
The optimal pricing would be between 3.00 and 3.30 depending on how happy your guests are.
Although, since guests can only have 1 photo per ride type, its possible if they ride enough times, they'll eventually buy a photo at the max price their happiness allows. If they ride a ride 8 times with a 3.00 photo, they will buy one 3.00 photo. If they ride 8 times with a 6.10 photo, there's a high chance they'll buy the photo on one of those rides.
Yeah i would assume that 4.5€ or 4.9€ might be the most optimal price in a park with happy guests since they have multiple opportunities to buy a photo
I did the math in my comment as well but here's what I calculated it out to be...
At an average peep happiness >180
peep will 1 ride in each category before exiting the park - $3.30
2 rides - $4.50 (expected return is $3.87)
3/4 rides - $4.90 ($4.29 - $4.59)
5-8 rides - $5.30 ($4.79 - $5.18)
8-20 rides - $5.70 ($5.27 - $5.68)
21+ rides - $6.10 ($5.73+) aka peep has to fail a 12.5% roll 21 times for you not to profit here
Assumes peep will remain at >180 happiness when riding each ride.
@@michaelt5188 That's a good calculation, but tbh I think whether it applies would depend on how the game calculates if a guest wants a photo in the first place, which the game does before it rolls the price range they'll buy it in.
Once again, this info brings me one conclusion.
The heartline twister is the worst coaster.
It always surprises the child inside me how much you can charge for things in this game. I only recently started charging $2.50. I also assumed ride stats affected it (and not just indirectly by making people happier).
They definitely should. Obviously this game isn't real life, but you would be way more willing to buy an on-ride photo if you had fun on the ride.
The combination of Marcel and Duerklinks contribution to the breakdown and mechanics of this amazing game is astounding. I learn so much from you guys! I dont want to minimize what anyone else has done, I know Bassie making plugins and X7123M3-256 making ride types!
deurklink
I hope you like the plugins! =)
No love for the OpenRCT2 devs? :-P
@@gymnasiast90 I love you ❤️
Where can you find these plug-ins? I've tried to find internet resources for new rides and scenery but a lot of the websites are either down or filled with malware
Its amazing how complex such game truly is. It looks so simple from the outside perspective. Keep up the good work Marcel, the video was great!
My guess based on looking at the table is that 3.30 is the best price in a "typical" park, since the odds are negligibly different except for unhappy guests (which you shouldn't have anyway). But also that the revenue difference between this and a simple 3.00 is probably negligible anyway. Would be a little bit interesting to see some empirical results in some sample park, even if it couldn't be taken as representative of all parks. Or at least an expected value calculation based on SOME sample distribution of guest happiness, arbitrary or not.
3.30 would be 10% more income per photo for -12.5% as many photos sold for moderately happy people, so I'm still not sure it would be worth it. I'd only try it if at least half the guests were very happy
@@steepslopesmm2 Maybe 3.10 otherwise? I think that option, at least, is almost strictly better than 3.00 in practice.
Also many others in the comments are talking about how re-riding withing the same category causes the -12.5% to not actually be -12.5% in practice, which supports 3.30 or even higher.
guests won't buy on-ride photos at all below 118 happiness and their chance of buying one increases linearly from 118 happiness to 242, so higher prices are probably better to a weirdly high amount
@@steepslopesmm2 While that's true, the question becomes how many guests are only moderately happy (or less) when leaving a good roller coaster ride, though? How much would 12.5% *of* moderately happy guests represent as a percentage of the *total* guests leaving the ride?
That's a good explanation for how much you can charge. But it has a glaring omission: when does a guest even consider buying the photo?
I tried to research that, but I couldn't find it in the code.
@@MarcelVos Chris Sawyer's mystery 👀
@@MarcelVos Can't you watch the wallet of guests and check when the money is deducted?
I think its happyness/excitement as stats and calculations, more exciting rides tend to get more guest photos
I'd like to know that too. It might come in handy.
I often see guests riding the same ride over and over. This would also be important in determining the optimal price? Since each guest can have multiple chances to buy the photo this may favor putting the price at something below 100% chance.
Good stuff. I've always charged 2.50. Good to know I can up it some more.
Also, you can't just look at the distribution of guest happiness in the park, but rather their happiness when they come off the ride, which significantly boosts happiness when they have the opportunity to buy a photo. I think I'm gonna run with 3.30, since my rides are just awesome.
Woah those categories answered life long questions, I always wondered why guests sometimes said 'I already have a on-ride photo' even though I could see some guests have multiple on-ride photos in possesions.
Logic's telling me that a pretty good way to do this would be to charge $3 for wooden coasters and water rides since chances are guests won't ride them as much, but for steel coasters to bump them up by a bit, maybe to around the $4.10 mark since there's a far better chance for guests to roll multiples.
Of course, this depends on the odds of a guest even rolling a photo in the first place. If it's quite low it's probably better not to push to much and maybe go down to $3.30 or $3.50.
I am in endless awe that this one man put so much detail into his game. It is a true masterpiece
Today I learned that you can actually set the amount for the photos, I thought that it was a fixed amount before this
The fact that you not only explain the game's mechanics, but give examples as well is a _huge_ help to learning how RCT2 actually works. Thank you.
Guests are happier immediately after they come off a ride, so I would definitely charge more
I haven't searched TH-cam before commenting this, but with your encyclopedic knowledge of RCT2, have you ever made a fully optimized park? No other goal than to exploit the games program through standard play to have the mathematically best possible park.
Define "best"
He does beat many scenarios in record times/minimal clicks, if that's what you mean.
Thanks for the shoutout, glad my guide was useful for this video!
This gives me a weird idea for a park goal. Make one coaster from each category and charge 6.10 for all 4 on ride photos. After a while, start checking guests inventories to see if you can find one with all 4!
Sadly guests will only buy one on ride photo regardless if they ride a different ride that has a photo station
@@Josh-jw3go
That ist Not true, as mentioned in the video , there are 4 categories .
What I like to know is how much the excitement rating affects a guest’s happiness when they ride it. Maybe that’ll give me the edge to charge at least $3.40 per photo.
When it comes to pay-per-entry parks I charge $3.30. when it comes to pay-per-ride parks I don't really adjust it, if I do it's to $2.50, that way (particularly in parks lacking the Cash Machine) the guests spend their money where it really matters. In hybrid parks I typically change $3.00 for them, that way I'm compensating for the reduced ride prices a bit.
I think it's interesting just how much nuance is behind something as simple as charging for a photo. It shows just how intricate game design really is.
Interesting to find out that /2RD/2RD and /4RD doesn't show any difference at 1 decimal till €9+ where it doesn't matter.
A couple seconds after you reduced the Go Karts to 1 and things started going absolutely insane, my computer randomly got a Blue Screen of Death, and for a split second I legitimately thought it was part of the video!
The question now is, does riding the coaster changes the happiness? I mean when a very unhappy person gets on the ride and when he leaves the ride but just before buying the picture, does he get more happy and maybe so much more, that he would pay more?
See! This is what we need to know: How does guest happiness get increased?
If the stats of the coaster are good you can see their happiness gradually increase over the course of the ride, unless the ride is too long, at which time it will start to decrease instead.
Guest: I’m not paying £6.10 for a on ride photo of Roller Coaster 1!
**Rains**
Same Guest: Here is £20 for that umbrella I didn’t want!
Conclusion: Just charge what you like 👍
Your videos are the only time I've ever found math relatable. Thanks!
So what determines if a guest wants to buy a photo or not? is it exclusively based around the ride category grand slam or are there other factors at play?
It sounds like they always buy if they have the money AND get a hit on the random chance.
What I'm surprised about is that the position on the coaster doesn't matter. It sounds like you could put it right before the first lift hill, and they'll still get the picture.
@@Inuvash255 he shoewd in an earlier video that the photo section doesnt even need to be on a part of the track the guests go through and theyll still buy it. So theyll just pay $3 for some scenery photos
@@maxevans5611 not even that, they'd pay 3 for literally nothing as the photo section wouldnt even activate to take a photo ^^
That is a wise decision on keeping the guests happy.
I really enjoy your content! Can’t wait to catch one of your twitch streams!!
That explains why my one obligatory inverted makes bank with photos. Only knew 3 types. I'll try 4.90 up from 3.50 in future, my parks are particularly pleasing to guests. Also, rides give a big happiness boost anyway.
I'm always amazed that you squeeze out so much great content from this game
Another great video, Marcel! When I watch some of your videos on some of the strange algorithms behind every system in RCT2, I start to wonder why Chris Sawyer made some of these calculations so random and cryptic. Ultimately I think it comes down to the fact that he wanted to model behavior of guests, and given that human behavior is so random, doing some somewhat arbitrary calculations that vaguely mimic reality makes a more believable simulation while using little computation power.
I love how Marcel refers to guests with "it" like its some kind of object or animal
What a complex system! I am now quite curious what determines whether a guest wants a photo or not.
I appreciate that you said to divide by two again, rather than to divide by 4!
This on-ride photo from Compact Inverted Coaster 1 is really good value.
Its crazy how on such an old game something as simple as the optimal photo price can be considered through so many variables that its almost impossible to compute. One thing is for certain, it should never be lower than 3.
It's crazy how much thought is behind every little aspect of this game.
I do hope that open rct2 has/gets a feature where I can see the total number of guests for each mood: how many are unhappy, happy, very happy, etc.
Having a quick glance feature like that could really help inform what prices things should be at.
Not sure if I misunderstood but I believe you mentioned that all of this math aside, not all guests actually check if they want to buy a photo... Aside from owning photos from the same category of rides, is there any other math regarding what guests check if they want to buy a photo? Or do all guests run these numbers to check if they don't own a photo from the same ride category?
Wondered this aswell
He mentioned in another comment that he tried to research it but couldn’t find it in the code
This is so awesome
I ran the numbers for a single ride with photo booth.
- if less than 23/31 (75%ish) of guests are happy, then the ideal price is $3.
- Above this, the ideal price becomes $3.10, and as you gain more very happy guests it goes up further.
- If all guests are at least happy and more than 17/33 (52%ish) of them are very happy, then $3.30 is the ideal price.
- It never goes above $3.30 for just one photo opportunity.
The category segment of this video has given me a new money making strategy for when I'm starting a scenario. Try to have at least ONE of each ride category with photo section before building more rollercoasters, so that you may maximize photo income (as well as ride income) early and give yourself time to gain enough wealth to build a better park and more rides down the line.
As a kid, I always assumed that the true value of a photo was based on the speed of the vehicle going through the on-ride photo section, but here comes Marcel to prove me wrong again. :)
Super interesting explanation, as always. Good job, Marcel!
I'd love to see a park that implements all your knowledge of shortcuts and exploits. From the sky it'd look like a troll park, but it would likely be as lucrative as can be.
So something to consider is also your ride excitement and queue length. Rides generally leave guests happier when they leave-- I've never really seen a guest come off a ride particularly angry, but I have seen them become unhappy from a long queue and then only become marginally happy by the ride. Considering guests can't buy a ride photo every time, it seems to me it may be a good idea, in terms of maximizing profit, to raise the price to maybe something like $4.50? Because if you've got a lot of ride photos in your park, chances are they'll probably roll a number that lets them buy it at that price at least once. At the default price, which I think is lower than the $3.00 you list here, usually every single guest in my park ends up with a ride photo quite quickly...
I always watch Marcel's videos (even some of the stream vods xD) but I hardly ever play RTC. It's just relaxing content ^^
Well this is eerie... I was googling this very question earlier today....
I have been charging all my guests $3.00 for photos for years. It just felt right and you just confirmed it!
it looks like its almost never a good idea to raise the prices
oh but you make a great point. they arent gonna be able to get a picture on every ride, so it does make some sense to raise the price.
€3 seems cheaper than the on ride photos in most real life theme parks I've been too; but then, I'm not sure I've ever bought one, because they're so expensive. Do we know what the chance is a guest will want to buy a photo?
Remember these prices are over 20 years old now.
@@kotor610 Even then they were low. I remember on-ride photos costing € 8 at Efteling in 2003, only one year after RCT2 came out (and 4 years after RCT1).
Real film is "expensive"
I did some rough calculations, and it seems the best play is to price the photos at the highest value where everyone will buy them, i.e. if all or almost all of your guests are very happy, then price them at 3.30, if there are lots of guests that are somewhat happy price them at 3.10, and if there are guests that are unhappy then price them at 3.00. If even 12.5% of guests don't buy the photo then it's not worth the extra price
I didn't include guests visiting multiple rides where they could get a photo because it would require knowing how many rides are in your park and how long the guests stay there, but for a single ride in a vacuum my method is the correct one. Since we're only talking about a couple dozen cents here anyways, the sensible play is probably to just set the price at 3.00 so you don't need to worry about average happiness or stuff like that
Even with just a single ride, guests will usually ride that same ride multiple times. You are losing out on a lot of profit if you just take the guaranteed money.
"i payed 3.00€ on wooden coaster 1 for my unhappy face. totaly worth it"
I think (if I understand this right) the photos have the souvenir flag, so whether a guest buys one is decided on line 1510 (at the time of writing) of Guest.cpp, which says a guest will buy a souvenir/photo when they have been on at least 3 rides and are happier than a random number between 115-242.
Of course I am nowhere near as carefully researched in my approach, I ended up mostly charging £3.20 as that's the highest I could get away with before I start regularly seeing "I'm not paying that much for an on-ride photo" in the guest thoughts.
Video idea:
You could show, how all of the properties of coaster are counted and all of the tricks, to make it better.
I for example sometimes wonder, if putting breakes before the loop would help with the air time, if I manage the speed of cart correctly.
It would be good for you to one day go in depth on the best location for the onride photos, as there's definitely a lot of variables involved!
"It's nearly impossible to weigh all of these factors"
Every mathematician, programmer, statistician, and computer scientist: "Hold our beers"
*complex mathing noises*
I did the math in python, averaged over 10k tests.
Charging anything over $3.30 is guaranteed to be sub optimal under all conditions.
Price | Average profit based on happiness level
$3.00 | $3.00 | $3.00 | $3.00
$3.10 | $2.73 | $3.10 | $3.10
$3.20 | $2.41 | $2.79 | $3.20
$3.30 | $2.06 | $2.87 | $3.30
$3.40 | $1.72 | $2.54 | $2.95
Conclusion
Ty for your video. I took a great time watching it
A cursory number-crunching shows that $3.00 is the best bang for the buck.
Happiness Ratio [H] = (255 - Guest Happiness)
Profit Ratio [P] = (% to buy photo * Picture Cost)
Average Profit = (H * P)
At $3.00, you average $3.00 per picture, since everyone that can, and wants, to buy a picture will.
Even just going up to $3.10 drops profit down to $2.91, as the 87.5% chance to buy at low Happiness drops the average, as it's the widest "band" along the Happiness gauge.
This does assume that Happiness is evenly distributed across all values, so a park with a higher average Guest Happiness could get more money.
It seems that lower is, to a degree, optimal; as lower costs increase purchasing odds across all Happiness levels, meaning more pictures are purchased when the game decides to look at the price.
The question this video raised for me is, why isn't the mine train coaster in category 3? I remember it looking very wooden whenever I built it in my parks.
After watching this and weighing my own analysis into it I would say pricing at 3.30 is a plenty safe default for most parks of a decent player. Still get that 100% happiest, and still really solid for what are hopefully outliers of ':)' to 'meh' or less. Otherwise the math on chance starts getting REALLY over the top and 'if then' happy.
I'd argue that in a well functioning park you're ideal price would be $4.50 or $4.90, assuming you don't mind the happiness hit.
I've been charging $5.50 so I'll be making some adjustments from this video because of that multiple on-ride photos category allowance I did not know about!
Edit: I did some math and it looks like when you adjust for multiple rides in your park for each category, the ideal price will be around $4.90 - $5.30, assuming peeps will ride 3-8 rides in a category prior to leaving your park and remains above 180 happiness when riding your rides.
At an average peep happiness >180...
peep will 1 ride in each category before exiting the park - $3.30
2 rides - $4.50 (expected return is $3.87)
3/4 rides - $4.90 ($4.29 - $4.59)
5-8 rides - $5.30 ($4.79 - $5.18)
8-20 rides - $5.70 ($5.27 - $5.68)
21+ rides - $6.10 ($5.73+) aka peep has to fail a 12.5% roll 21 times for you not to profit here
Assumes peep will remain at >180 happiness when riding each ride.
@@TheWatchernator He said he'd argue, not that he'll argue!
@@TheWatchernator I got into a bit of a rabbit hole on it and quantified the expected returns using probability :) Essentially the argument to be made is you have the opportunity to adjust for # of expected rides a peep will ride prior to exiting your park. Because there are varying % of success, each failed on-ride photo purchase will get another reroll chance when the peep rides another ride in the same category. The current recommended price only assumes a success rate from the FIRST attempt at an on-ride photo purchase for any given category. We obviously can have multiple rides in each category and can expect peeps to ride them as they experience your park. This is akin to being able to flip multiple coins at a time, but in this case, all we're looking for is just one heads result. Intuitively, you can see that the chance of NOT flipping heads diminishes significantly the more chances you get to flip the coin. The same applies here and the high level math is below...
At an average peep happiness >180
peep will 1 ride in each category before exiting the park - $3.30
2 rides - $4.50 (expected return is $3.87)
3/4 rides - $4.90 ($4.29 - $4.59)
5-8 rides - $5.30 ($4.79 - $5.18)
8-20 rides - $5.70 ($5.27 - $5.68)
21+ rides - $6.10 ($5.73+) aka peep has to fail a 12.5% roll 21 times for you not to profit here
Assumes peep will remain at >180 happiness when riding each ride.
Once I clean up the file, I'll publish it for further scrutiny.
I like knowing this kind of insight into Rollercoaster Tycoon, and I agree with Marcel, and will price my pics at $/£/€3
nice work man! you are very smart.
Would it be possible to create a park that could be stable in a condition where everyone is either very happy or very unhappy? Someone with happiness below 128 would consistently be mad about $3.70 photos, but someone with happiness above 180 would mostly be happy. Are there enough mechanics like this that you could create a park that could sustain people as both very happy and very unhappy simultaneously?
There's a small amount of random chance that a guest will just wander for a long time, never going on rides. I think you'd need a very optimal park design to keep ALL guests happy. Even when they have a map, all that really does is make their next turn aim towards their selected ride, it doesn't ensure perfect pathfinding.
It's quite rare for me to drop below 999 happiness in parks these days, exception being when overcrowded so I just default to and have to a while to 4.50, although I'll probably go up to 6 now given most guest except in pay to entry parks tend to ride several rides or rides several times. Rather just tank the minor happiness debuff. If I'm not capped on park rating it's almost always because I haven't got enough drinks/toilets/food so it's better just to slap down a few more of them around.
park rating is not guest happiness
@@CouchPotator Guest happiness affects park rating. That is why they are talked about interchangably.
First words I hear "this video is based" very true marcel
Based on what?
Some good basic statistics, really nice
This video is BASED
Also, where do guests keep getting all this money we can exploit out of them? I'm not complaining, just curious.
For unhappy customers, the price with the best expected utility is €3.00.
For moderately happy ones, it is €3.10.
And for very happy ones, it is €3.30.
Basically, the base price of 3 euros is high enough that as soon as the guests needs to randomly determine if they want to pay, you're losing out.
I was kinda hoping there would be a variable of where the onride photo was actually placed. For example, guests would be more likely to buy an onride photo if it was placed at a more exciting part of the ride like the bottom of a big drop, rather than an onride photo that takes the picture at the brakes.
before i try it myself, assuming a park has very happy guests and the chart of chances guests will buy photos is correct, this is my prediction for an optimal photo price:
2.90 + 0.40 * min(x, 8)
x is the amount of rides in the park and the same category that have a photobooth (yes i call it that for some reason). for example if you have 7 wooden coasters in your park and 5 of them have photobooths, x is 5 when calculating the theoretical optimal photo price for said wooden coasters.
tried to link a screenshot to a quick table i made, but for some reason youtube lately has been automatically deleting all of my comments that contain a link, so, guess i'll go cry about it.
seriously why cant they just disable posting comments with links? why instead say "returned error" when i try to edit my comment only for it to be gone when i reload the page with no explanation? fucking wish this site had a competent competitor.
nobody:
the very happy guest: :D
I think the category could dictate how much you can charge. Lots of metal coasters means lots of opportunities for a guest to go and buy a photo, so the risk of a higher price is a little bit mitigated and they'll probably end up buying a picture eventually. Wooden coaster only having one ride in its category means you want to make sure anyone who rides it buys a photo since..
Actually, do guests go back to rides over and over? Id like to believe they just pick any random one and ideally you just want to have as many diverse rides as possible to avoid the duplicate ride halfing.
I've watched a guest go immediately to the opposite side of the part to ride the same coaster like 4 times in a row, I think I had an ad out for that one
0:14 this video is BASED
0:15 "This video is *BASED* "
True
What about the amount of hapiness that guests gain from rides? How is that calculated? Maybe for nausea aswell
Well done 👍
These videos exploit my competence kink
Can you do a video on prebuilt rides? For example a top 10 worst or best
Marcel is the man
The optimal price (Maximum of "{chance that someone buys it} * {Price}" ) is
€3.00 for unhappy guests
€3.10 for somewhat happy guests
€3.30 for very happy guests
Pick whichever the supermajority of your guests are (or go with €3.00.. it is only a 10% increase)
I think 3.30€ is best. A guest might not be happy on his first ride, but he might return later with a Happiness value of more than 180 and so the chance of him buying an onride photo (if he decides to) is 100% at that point again
The best kinds of roller coasters always had on-ride photos.
Are River Rapids and Log Flume in Category 4 as water rides?
They are, yes. And so are Splash Boats (which have an on-ride photo section in OpenRCT2 but not in RCT2 or RCT Classic).
Dammit you always asking questions I never thought to ask but now I must know the answer to 😂😂😂😂
All this time I've just been pricing the photos for half what the ride costs, up to a maximum of $4.50. That always seems to work out.
Nice video.
When collab with Deurklink in Dutch? Would love that
Would be lovely. Wouldn’t mind appearing in that one too if they have a good angle.
I always thought that by the end of the ride the guest would have maximum or close to maximum happiness anyway
So cool marcel
1:33 I've just now learned that it's when a guests exits the ride, not when they get on the ride
and for some reason I didn't question it. like why would they buy a photo of a ride they haven't even ridden yet?