The Guest Observer; get your crowd in control

  • 10 replies
  • 439 views
*

Offline Redhair

  • *
  • 178
  • Can't... Stop... Playing...
    • View Profile
The Guest Observer; get your crowd in control
« on: October 17, 2018, 01:59 PM »
I just posted this thread in the Planet Coaster forums.
Sorry to say, but I think it will get here more response when it comes to find opinions about something.
Please take a look at this very serious feature request and let me know what you think.
If you're very supportive, do not hesitate to support this idea on the official forum :)

Dear lovely people,

First of all, many thanks for the latest update and the stream tonight. It was wonderful to see all the creations and felt flattered also my Royal Diner has been shown.
Many suggestions came in the chat, and I've personally been dreaming since RCT3 for this suggestion and hope you all will take a serious look at it.

Whenever I create a park, the crowd control is a big part of the game.
How do you get more people move over to the next area;
Why is no one riding that particular ride;
Is this shop cost-effective, why (not);
etc.?

To gather this information (and more,) I do click on guests... I click on many guests... Also use the average statistics which can be found in the park management.
But... There's one thing to make crowd control even more manageable, a tool we do miss in our beloved game, a tool that could save us time but also gives nice to know information.

Ladies and gentlepeople, I proudly present: The Guest Observer.
Since words are hard to describe such an idea, I will give some visuals to explain what I'm thinking of, but in short it gathers information from selected groups of people on average:

(SPOILER, THIS IS NOT IN-GAME FOOTAGE)


^ This tool will be located on the multi selection mode, which can be found in the right bottom corner.
As you can see there is a new icon, the Guest Observer. This tool is the multi selection mode for guests.


^ Of course first select the tool by clicking it, and make a selection on screen. When you've done that the next screen will be the result of this wonderful tool.


^ You've selected a group of people as you see they light up.

Now let me talk you through the most important information that can be seen.
  • The amount of people selected.
    - This is giving a good indication of the popularity of certain area's and might even be the main use of this feature.
  • Most popular ride, based on the most visits of the selected group.
    - The most popular ride can change in certain area's in the park. People can stick around or will be pleased to move on quickly.
  • Favorite Coaster, based on the most visits of the selected group.
    - Idem for the ride.
  • Favorite shop, based on the most visits of the selected group.
    - Main streets do have (lots) of shops in the beginning of the park, but are there area's which are more profitable?
  • Average guest needs.
    -This is a very interesting option. Some coasters might produce a high nausea, is there an First Aid close enough? Are toilets needed in this area? Are guest hungry in this area... And so on ;)
This feature could be as comprehensive as you want, using the other tabs like thoughts and spend money.

Personally I do see lot's of benefits in this feature and desperately want it.
But that's just my opinion. What do you guys think?
Would this be lot's of programming, or is it achievable to implement this feature into the game?

Let me know if you agree or disagree and feel free to tweak this proposed feature!

Spoiler (hover to show)

*

Online Bullethead

  • *
  • 2108
  • shyguy's World Rocks!
    • View Profile
Re: The Guest Observer; get your crowd in control
« Reply #1 on: October 17, 2018, 08:01 PM »
Whenever I create a park, the crowd control is a big part of the game.
How do you get more people move over to the next area;
Why is no one riding that particular ride;
Is this shop cost-effective, why (not);
etc.?

I've studied peepology since PC was released, being as they said it was all about their AIs.  I've done quite a number of studies and can answer your questions.

1.  How do you get more people to move over to the next area?

The bottom  line is, there has to be something at least as attractive to peeps in the new area as there already is in the old area.  This brings up the whole peep decision-making system.  This is a very complex subject but I'll try to hit only the high points.

The main peep decision is where to go next.  They make a list of all their options, assign each a probability of being chosen based on the current perceived desirability of each option, and then roll d100 to determine which of the numerous options they pick.  It's thus impossible to state with certainty what any peep will decide to do next because things with low probabilities can happen, but in general they go with the high odds.

Peeps have 3 main categories of what to do next:  go on a ride, satisfy a physical or mental need, or leave the park.  The system is heavily biased in favor of peeps going on a ride next instead of satisfying a need or leaving the park.  Picking which ride to go on has its own convoluted decision tree.  The needs and leaving the park are much simpler:  they are pretty much driven by elapsed time.  All, however, are impacted by peep happiness.  The happier peeps are, the more they ignore needs, the less likely they are to leave the park before they run out of money, and the more likely they are to go on another ride, or try to even if they're out of money. 

Even non-artistic Neanderthals like me have no trouble keeping peeps in a state of bliss, so they'll die of starvation and UTIs before deciding to do something other than go on another ride.  Thus, for anybody's reasonably pimped-out park with a reasonable selection of rides, peeps are mostly deciding which ride to go on next.  Periodically they need to take a leak, consume something, or buy a trinket, and eventually they'll decide to go home as that becomes overwhelmingly probable over time, but for the most part it's all about rides.  Which segues neatly into your next question.

2.  Why is no one riding that particular ride?
The overwhelming majority of peep decisions are 2-fold.  First (because they're blissfully happy), they decide to go on a ride rather than piss, eat, drink, rest, buy a trinket, hit an ATM, or go home.  Having decided to go on a ride, they next have to pick which of your various rides to move towards.  This decision has 2 main factors.

#1 is the ride's fear rating.  All peeps have a max fear tolerance and will never, ever go on a ride with a fear rating greater than this.  Rides above 3.5 fear are too scary for about 80% of families (absolutely zero families tolerate more than 4.0) and rides above about 4.5 fear are too scary for 80% of ALL peeps.  As very few coasters allow kids to begin with, it does you well to keep the fear of kid-capable coasters (woodies, hybrids, etc.) no higher than 3.5.  And even large parks can't support more than a few rides with fear over 4.5 because only a small fraction of adults and teens will tolerate that much fear.

NOTE:  Peeps also have a max nausea tolerance that works the same way but for all peeps, their nausea tolerance is rather higher than their fear tolerance, and it's DAMNED hard to make a ride more nauseating than scary unless you do so deliberately.  So for practical purposes, you can ignore nausea except that rides with nausea above about 3.5 need a 1st aid station near their exit to prevent excessive puke on the paths.

Anyway, after excluding from consideration all the rides with too much fear, the peeps have to decide which of the remaining rides to go on.  In this decision, the overwhelming factor is ride prestige.  Other considerations are the travel time/distance to each ride and the number of times the peep has been on that ride already.  A short travel distance will trump a prestige difference of a few hundred points but a really high-prestige ride on the other side of the park will still be more attractive than nearby low-prestige rides.

NOTE:  Peeps only consider the cost of a ride (and the fullness of its queue) when they get there.  These factors can cause the peep not to go on the ride once he get there but have no effect on his decision to move there.  The only exception to this, since 1.7, is that peeps consider the cost of transport rides prior to deciding to use them or just walk to the other side of the park.  Peeps have a low tolerance for cost on transport rides so it's generally a good idea to keep them free or cheap if you want them to be used for transport.  Of course, peeps also now consider transport rides as entertainment so will have them as valid ride options regardless of cost.

But anyway, the main thing is ride prestige.  This is a weighted average of the ride's excitement rating, duration, scenery and, optionally, its age.  Assuming you have ride-aging turned off, then scenery will be a wash because you pimp everything to Very High.  Thus, the only significant factors are excitement and duration.  In these categories, all ride types are not treated equal.  Track rides have the easy win because they get WAY more excitement for their slow speeds than do coasters and flat rides, and because track rides are slow, they also have high durations if the track has any significant length.  And because track ride speed is so low, their fear rating is never much over 3.5 so they're attractive to nearly all peeps.  It takes a large, exciting coaster that still has a fear down about 3.5 to compete with a long, heavily themed track ride.  The tamer flat rides are sucking hind tit in that they're slow and short, and don't get the bonuses for slow speed that track rides do.

In terms of peep decision-making, ride prestige is broken into broad categories every 600 or so points.  IOW, 0-600, 600-1200, and 1200+.  Rides in the same category are about equally attractive to peeps (provided their fear is low enough) but are absolutely trumped by rides in higher categories.  Thus, peeps will flock to rides with prestige > 600 and largely ignore those with prestige < 600, which includes the tame flat rides and lame kiddie coasters.  It's really hard to make a ride with a prestige of 1200 and damn near impossible when keeping within realistic bounds to get significantly higher than that.  But it's extremely easy to make coasters and track rides with prestige of 1000-1100.  And it's damn hard to get the prestige of most flat rides and kiddie coasters above that 600-point threshold.  The new lapping of kiddie coasters helps as this greatly increases their duration, but the tamer flat rides are still hosed.

So, the bottom line is as follows:

#1.  Don't fill your park with "all green" coasters because "green" fear is too scary for vast majority of peeps.  Such rides will never do much business.  It's OK to have 1 or 2 such rides in your park provided it's big enough and you can allow 10,000 or so peeps into it without FPS issues  But in general, I recommend keeping fear, especially for kid-capable coasters, at 3.5, and for non-kid coasters at no more than 4.0.

#2.  Don't have just 1 big ride with a prestige of 1000 or so, especially if it's kid-capable and has a fear of 3.5 or less.  If you only have 1 of those, it will be the 1st choice of essentially all peeps in the park and everything else will die.  So you need a bunch of these evenly distributed around the park so as to prevent all the peeps mobbiing 1 single area.

#3.  Resign yourself that your flats and kiddie rides with prestige < 600 are just NEVER going to do much business if there are any other rides in the park with reasonably low fear and prestige > 600.  The ONLY peeps who will ever go on such rides are the most timid, whose ultra-low fear tolerance precludes them from going on any other rides.

3.  is this shop cost-effective, why or why not?
Up until yesterday, I knew the answer to this.  But now we have an assload of new shops so I'll have to experiment again.  HOWEVER, the answer prior to yesterday was pretty simple....

Peeps had (and maybe still have) strong preferences for the type of food and drink they prefer.  Those used to be burgers and hotdogs, water and juice.  They'd only buy the other stuff if either there were none of their favorites at all in the park or they were so critically in need (due to being blissfully happy) that they'd die if they didn't go to some other shop instead of walking 50m farther to something they liked better.  And because peeps can recharge energy for free while sitting on a bench, if your park was full of benches, you'd never sell a single cup of coffee or Gulpee Energy.

In general, however, peeps consume about 2 drinks per 1 food.  So you want an average of 1 food, 2 drinks, and 1 restroom per park area.  If you do more than that, the excess will suffer.

When we got the 1st batch of new shops (stuff like Missy Good, Street Fox, etc.), I ran an experiment presenting peeps with all the options and added up the results after several months.  All they wanted were burgers and hotdogs, water and juice and they hated everything else.  All the new shops were useless.  I presented those results to Frontier and asked for an abolition of peep food/drink preferences but since then I haven't noticed any change.  But now we have a bunch more new shops, which will only be worth using if there are no longer peep preferences.  So I'll have to run that experiment again.

Peeps also have preferences when it comes to trinkets.  Everybody wants a balloon, many peeps will get hats, but only only the happiest will ever get a snowglobe. and that only after they've been happy for a while.  So a large park with several thousand peeps can support 3 balloon stores, 2 hat shops, and 1 memento.

As for ATMs, using them is entirely dependent on a random peep trait, which is generally confined to about 2-3% of the whole population.  It doesn't matter how happy they are, they won't use an ATM unless they're predisposed to do so.  Thus, you only need a very few widely spaced ATMs.

1st aid's purpose is to reduce nausea and this is only needed at the exits of high-nausea rides like the Kickflip and Cube.  In such places, they do good business and prevent puke accumulation.  Everywhere else, they get zero business.
-Bullethead
NIHIL INIQVIVS QVAM ÆQVITATEM NIMIS INTENDERE
My Steam Workshop page

*

Offline Redhair

  • *
  • 178
  • Can't... Stop... Playing...
    • View Profile
Re: The Guest Observer; get your crowd in control
« Reply #2 on: October 18, 2018, 05:05 AM »
Bullethead, you should write a book about this game.
Not because this is a long answer, but because of your knowledge.
You just taught me lot's of stuff.

Your answer has proven that there is way more simulation involved than I ever thought.
Almost all of your answer was not in my knowledge (until now).

Thank you so much for sharing this, this will really help. Thanks man!

Might change my park for what I've created so far  :-[


*

Offline cap396

  • *
  • 459
    • View Profile
Re: The Guest Observer; get your crowd in control
« Reply #3 on: October 18, 2018, 08:24 AM »
Wow, Bullethead, you've got it all figured out.  Those tips are very helpful for understanding the peep decision making process.

I could never understand why my coffee shop was getting ignored even though the guests always seem to be thirsty.  I thought it was because of the location of the shop, but thanks to Bullethead's explanation, now I know that the guests prefer water and juice.

I'll add just one thing that seems to help get the peeps to spread out to more rides and helps prevent some rides getting ignored.  In the Scenario Editor, there is the option to give the guests the "Day Tripper" trait.  The guests that have this trait usually have a wider fear tolerance, usually from 0.0 to 10.0, so they are attracted to a wider range of rides.  But I haven't done the extensive testing that Bullethead has done, so I could be wrong, but it seems to be true for the park I'm working on right now.

*

Online Bullethead

  • *
  • 2108
  • shyguy's World Rocks!
    • View Profile
Re: The Guest Observer; get your crowd in control
« Reply #4 on: October 18, 2018, 10:05 AM »
@ Rehair:  Somewhere in the tutorials section is a piece I did called "Peep Demographics:  Meet Your Average Customer".  I think the obsolete one on food preferences is there somewhere, too.  Also over at the official forums.  You might take a look for them as they go into more details.

But yes, peeps are always making decisions, which is why PC is such a CPU-heavy game and why FPS plummets once you get a lot of peeps.  Each peep group has its own separate AI thread that needs CPU time.

@Cap396:
I don't have it all figured out :).  But I have studied it more than most folks.  I think it's fascinating to watch AIs do their thing and I enjoy experimenting with them.

The main thing I don't know much about is what all the food "extras" do.  They all actually have an effect on peeps, they're not just cosmetic, so you have to be careful with them.  In fact, different types of food and drink by themselves also have side-effects on peeps besides satisfying the main hunger or thirst need.  There's a really complex interplay with all these things.

The baseline system is that peeps have physical needs for food, drink, restrooms, energy, and health.  Food, drink, and restroom needs increase over time with thirst typically running twice as fast as hunger and bladder (although teen hunger increases faster than adult or family, and family bladder increases faster than adult and tean.  Energy need increases the more the peep walks but goes down if the peep is on a ride, standing in queue, or sitting on a bench.  Health need only increases when the peep is exposed to high-nausea rides.  It goes down by itself over time, or if the peep pukes, or when the peep goes in a 1st aid.

Some food and drink, and/or their "extras" will satisfy more than 1 need and/or impact the rate at which other needs accrue.  For instance, coffee (except decaf) and Gulpee Energy both satisfy the energy need a lot and the thirst a little, and also add to the bladder need so the peep will need to use the restroom sooner.  Because energy is available for free throughout the park from benches, peeps will use benches way more often than they'll buy coffee or energy drinks.  Those coffee and energy drinks never sell very much unless you don't have many benches in the area.

Other examples are if you put fruit chunks in juice or slush, these will satisfy all the thirst and also some of the hunger, so peeps are less likely to buy food for a while.  Spicy/gassy foods and/or spicy "extras" on bland food will increase the impact of ride nausea on peep health, making them more likely to puke.  All food and drink consumption increases the need for restrooms but water by itself does so especially.  Peeps even comment on "wow, that water went straight to my bladder" or some such.
-Bullethead
NIHIL INIQVIVS QVAM ÆQVITATEM NIMIS INTENDERE
My Steam Workshop page

*

Offline cap396

  • *
  • 459
    • View Profile
Re: The Guest Observer; get your crowd in control
« Reply #5 on: October 18, 2018, 01:57 PM »
Each peep group has its own separate AI thread that needs CPU time.

So if I have say 4000 individual single peeps that aren't part of a group, that will use more CPU power than if I have 4000 peeps that are in groups (say 1000 groups with 4 peeps in each group)?  In other words, the amount of CPU power is determined by the number of peep groups, not the number of total peeps?

*

Online Bullethead

  • *
  • 2108
  • shyguy's World Rocks!
    • View Profile
Re: The Guest Observer; get your crowd in control
« Reply #6 on: October 18, 2018, 03:37 PM »
So if I have say 4000 individual single peeps that aren't part of a group, that will use more CPU power than if I have 4000 peeps that are in groups (say 1000 groups with 4 peeps in each group)?  In other words, the amount of CPU power is determined by the number of peep groups, not the number of total peeps?

To some extent.  As I understand things, all peeps are in "groups" which can range in size from 1-6.  IOW, no matter how many peeps are in a group, all groups run the same basic AI routines.  However, if there are 2-6 peeps in the group, the AI routine has to poll all the group members to arrive at a consensus decision.  Thus, the AI workload still increases with the total number of peeps, although having multiple peeps in 1 group reduces the total number of separate AI threads running.

See, all peeps spawn with randomized starting physical needs, even if they're in the same group.  And in families, which always contain adults and kids and sometimes a teen or 2 as well, these needs accrue at different rates.  The bottom line is that within each multi-peep group, some peeps will be hungrier, thirstier, and more in need of a restroom than others.  So the AI has to average all that out to come up with the single probability of satisfying one of those needs instead of going on a ride.  The only thing that's uniform for all peeps in a group is their fear/nausea tolerance.

BTW, the fact that all peeps start with different amounts of hunger, thirst, and bladder, and that these accrue at different rates even within the same peep (thirst is faster than hunger, for example) means that peeps almost never are simultaneously hungry, thirsty, and in need of a rest.  This means that concentrating all your food and drink stores in 1 big food court or restaurant setting is a bad idea in practice, even if it's realistic construction.  Such food courts can satisfy all needs but peeps will usually only have 1 need that's critical when they enter.  Say they're hungry.  So they go in, buy food, and leave, probably to go on a ride.  But before they get there, they get thirsty.  Now they have to turn around and walk back to the food court and buy a drink.  Then they leave again, only this time they have to take a leak, so back to the food court.  The result is peeps wasting a lot of time walking and not going on rides, so they aren't as happy as they should be and you don't make as much money.

I find it's generally better to spread things out in park "areas".  An "area" to me is a chunk of the park that has a major ride, several flats, a transport station, and the standard batch of physical needs shops (1 food, 2 drink, 1 restroom).  The entrances to all these attractions are reasonably close together, then there's a bit of path connecting to the adjacent area(s).   Within each area, you can have the needs shops all side-by-side, or spread them out, it doesn't really matter, but don't put in many more needs shops than this.  And you have 1 such cluster in each other "area".  This prevents the problems of centralized food courts.
« Last Edit: October 18, 2018, 06:41 PM by Bullethead »
-Bullethead
NIHIL INIQVIVS QVAM ÆQVITATEM NIMIS INTENDERE
My Steam Workshop page

*

Online andreizsmart

  • *
  • 818
  • 'supercoolsuper'
    • View Profile
Re: The Guest Observer; get your crowd in control
« Reply #7 on: October 18, 2018, 06:33 PM »
Some of you may or may not know this, but some theme park design firms (including the one i work for per me haha)  have literally used the game's guest flow simulation programming to estimate and study walkway capacities for internal proof of concept. While a real landscape architect is used to determine the actual requirements and such, its important early in conceptual master planning to really understand guest movement, especially when determining what attractions and shops to place where in a park/land. The game provides a pretty elaborate and, most importantly, QUICK way to study.

So yes, the complexity of guest fluid dynamics of Planet Coaster is pretty advanced.
Initiating power transfer...

*

Online Bullethead

  • *
  • 2108
  • shyguy's World Rocks!
    • View Profile
Re: The Guest Observer; get your crowd in control
« Reply #8 on: October 18, 2018, 07:33 PM »
So yes, the complexity of guest fluid dynamics of Planet Coaster is pretty advanced.

Wow, didn't know that ;)

So many folks complain that PC's peep-herding isn't "deep" enough and constantly request features they think will add "depth".  But the game is actually quite deep and what they're really looking for is added "width" which are all merely ways for the player to have less money without adding any gameplay value.  If folks don't want as much money, they can reduce prices and/or not have the park open 24-hours/day.  Same result in the end without Frontier wasting time on adding another useless "width" feature instead of, for example, making more rides.

The whole crime system is a case in point.  The ONLY result of turning on crime is that it increases your overhead.  You have to hire some guards and install a bunch of cameras.  But with enough cameras, you have no crime, so the net result is simply the player having less money than the same park would generate with crime turned off.  It adds not 1 iota to the player's experience. 




-Bullethead
NIHIL INIQVIVS QVAM ÆQVITATEM NIMIS INTENDERE
My Steam Workshop page

Re: The Guest Observer; get your crowd in control
« Reply #9 on: October 19, 2018, 04:09 PM »
this is a great thread...but...now I have to pee. :)

*

Online Bullethead

  • *
  • 2108
  • shyguy's World Rocks!
    • View Profile
Re: The Guest Observer; get your crowd in control
« Reply #10 on: October 19, 2018, 04:46 PM »
this is a great thread...but...now I have to pee. :)

Too bad!  The majority of peeps in your group have decided to walk (rather than take the $2/seat train) 3/4 mile to the opposite side of the park just to wait in queue 30 minutes for this awesome darkride that itself last 5 minutes.  After that, your group will have another vote, by which time more bladders will probably be full  :P
-Bullethead
NIHIL INIQVIVS QVAM ÆQVITATEM NIMIS INTENDERE
My Steam Workshop page