Parks Performance Friendly Full Map in Progress: Safire Meadows

[edjenoh]

I've tried several times now, and failed every time, but I do think I should have the final formula for making a performance friendly park that fills the entire map.

I'm currently at six rides: Caroussel, Tri-Force, American Arrow Coaster, Victory Swinging Ship, Giga Dark Coaster and a Hammer Swing.
The key is to spread out the rides over the map. I will be aiming for 18 rides at first, one in approximately each 1/16 quad of the map + 1 Railroad + 1 Monorail.
For the remaining rides I will be going for 4 more coasters (Hydra, Dragon Kiddie Coaster, Boomerang inverted and a Woodie (maybe a hybrid, not sure yet), one river rapids, one karts and the rest will be flatrides.
I'm currently filling up the remaining open area's between the rides with scenery (much needed to fill up all them que lines, folks spend a lot more time walking with rides dispersed), after that the River Rapids is the next ride for construction in the area behind/next to the big modern entrance thingie (final picture).
I'm curious whether performance will tank before 18 rides, if it is still feasible I'll go up to 20, but we shall see. I'm playing in harder challenge mode with guests capped at 0.5 park ratio.

 

[Bullethead]

I'm curious whether performance will tank before 18 rides, if it is still feasible I'll go up to 20, but we shall see. I'm playing in harder challenge mode with guests capped at 0.5 park ratio.

A very interesting experiment in a noble cause. I wish you every success

FWIW, in my experience the main hit on FPS is the number of peeps, not the number of rides, scenery items, or buildings parts. The number of peeps is a set multiple or fraction of the overall park rating unless you manually set a lower limit. Overall park rating seems to be a combination mostly of the park's overall ride and scenery ratings, plus a dash of balance rating (whatever that means) and (optionally) marketing rating, all of which are factored by overall peep happiness, which seems to be measured by peep thoughts on leaving. That's my theory, anyway. No telling yet if it's correct, but I rather suspect it is.

So, to keep FPS up without setting an arbitrarily low population cap, yet still use all the available acreage, my theory predicts that you should build the park with as low a rating as possible, which means limiting either or both of ride and scenery ratings. Limiting scenery is the easiest---just have naked rides and shop boxes, nothing else. Then you can probably cram as many rides in there as you can fit, even if they have great stats. Limiting ride rating can be done by making rides with crappy stats and/or limiting the number of rides. But if the object is to use all available space, limiting the number of rides is something of a dodge, so that leaves just having boring rides, which is just as unsatisfactory as leaving all your stuff naked.

But in any case, limiting the number of peeps is the same as limiting your income, because by and large the only money you can make is what peeps bring in with them, given that only a few percent of them will ever use an ATM. Thus, to maximize income, you have to maximize the number of peeps in the park, at least up to what your computer can stand. If you have more rides than can be kept profitable by the number of peeps your computer can handle, then you're just throwing money away on the operating costs of excess capacity.

Thus, it seems to me that the goal of using all available acreage is counterproductive if your computer can't handle a fully developed, full-map park. It would probably be best to build smaller parks, with excellent rides and full scenery, right up to what your system can deal with. That way, you get the most income your system can generate.

But this is just a theory. It has made predictions. I would be very glad if you could prove it wrong, because we would all learn something in the process, and be able to make better theories in the future.

 

[edjenoh]

A very interesting experiment in a noble cause. I wish you every success

FWIW, in my experience the main hit on FPS is the number of peeps, not the number of rides, scenery items, or buildings parts. The number of peeps is a set multiple or fraction of the overall park rating unless you manually set a lower limit. Overall park rating seems to be a combination mostly of the park's overall ride and scenery ratings, plus a dash of balance rating (whatever that means) and (optionally) marketing rating, all of which are factored by overall peep happiness, which seems to be measured by peep thoughts on leaving. That's my theory, anyway. No telling yet if it's correct, but I rather suspect it is.

So, to keep FPS up without setting an arbitrarily low population cap, yet still use all the available acreage, my theory predicts that you should build the park with as low a rating as possible, which means limiting either or both of ride and scenery ratings. Limiting scenery is the easiest---just have naked rides and shop boxes, nothing else. Then you can probably cram as many rides in there as you can fit, even if they have great stats. Limiting ride rating can be done by making rides with crappy stats and/or limiting the number of rides. But if the object is to use all available space, limiting the number of rides is something of a dodge, so that leaves just having boring rides, which is just as unsatisfactory as leaving all your stuff naked.

But in any case, limiting the number of peeps is the same as limiting your income, because by and large the only money you can make is what peeps bring in with them, given that only a few percent of them will ever use an ATM. Thus, to maximize income, you have to maximize the number of peeps in the park, at least up to what your computer can stand. If you have more rides than can be kept profitable by the number of peeps your computer can handle, then you're just throwing money away on the operating costs of excess capacity.

Thus, it seems to me that the goal of using all available acreage is counterproductive if your computer can't handle a fully developed, full-map park. It would probably be best to build smaller parks, with excellent rides and full scenery, right up to what your system can deal with. That way, you get the most income your system can generate.

But this is just a theory. It has made predictions. I would be very glad if you could prove it wrong, because we would all learn something in the process, and be able to make better theories in the future.

I am aiming to go for a 1:1 ride:scenery rating, because I do want the park to look relatively nicely themed and filled, yet still be performance-friendly. This is a bit of a challenge because you just have to make certain themed hubs/nodes and fill a significant part of the terrain with sparce scenery (wide paths/squares, lakes/rivers, sparcely placed rocks/trees and elevation/hills/mountains work best for this). Also, it's probably a good idea to have all the track rides cover a larger surface as opposed to cluster the ride close together, I think that worked out quite well on the Arrow coaster, it covers almost a complete 1/16 quad of the map.

I just tried to disable the guest population to the default 1.5 on the 6 ride park to see what it would be like with that many guests that would be attracted to 18 rides. Performance dropped to around 20-25ish, so probably 18 will be a good estimate for the limit, if guest count determines performance and rides and scenery are neglible. I don't think they are completely neglible though, so perhaps I will have to settle at 16 rides, but again, we shall see at the end. I don't intend to limit guest count below 0.5, because I also want the park to be profitable at harder challenge mode, and with less than 0.5 it can become quite a challenge to stay profitable, because guests are spending much more of their time walking (also interesting to see how the transfer rides will improve this issue on a dispersed park like this when I do build them).
I will probably also try reducing ride sequences on all the flat rides at the end of the build to lower ride rating, perhaps that will help me squeeze in another ride [tongue]

 

[edjenoh]

Scenery coming along with river rapid ride.
I did another test on that BitterJeweler 20k+ park. At the park rating and amount of rides I'm aiming for (gradually deleted scenery and rides) I got 25-32 fps at normal speed with closed park. Around 19-22fps with a 1000 and 17-22fps with 2.5k guests. I suppose that is reasonable enough. The real culprit indeed seems to be guests, with rides and animatronics and special effects perhaps also having a very minor influence.

 

[breezerHOG1]

very impressive [up] especially considering all the terrain you did is in harder challenge mode!

 

[Bullethead]

Wow, that's looking quite nice. it shows that a little bit, if done well, can still go a long way . But what's the huge, arch-shaped building we've seen several times?

 

[edjenoh]

Wow, that's looking quite nice. it shows that a little bit, if done well, can still go a long way . But what's the huge, arch-shaped building we've seen several times?

I don't know, I got it from the workshop, I guess it's sort of like a museum center for the jungle area behind it. I've put some artefacts scattered through out the jungle for sightseeings.

The River Rapids ride scenery is finished now....

... touched up the lake area....

.. and I've added another flatride with spaced theming, still needs expansion...

and also build a transport ride... I believe transport rides actually tank performance a bit harder than other rides, perhaps because they involved peeps computing whether or not to take them for faster traversing times (I tested this in another park by removing the transport rides and it gave me 1-2fps in a very crowded park (going from like 14 to 15-16 fps). I guess I will stick to one transport ride for this park, instead of the planned two.

The downside is that I'm already hitting the mid twenties with 9 rides, which doesn't look promising, since I was aiming for 16-18. There is no way that will still be performing above 20 fps.
The problem is that I can't cap peeps below 2000 at this point (which is about 0.5 of park rating), because my park stops being profitable at harder mode. Perhaps for performance friendly park it's a better idea to go easy challenge mode or sandbox. It's kind of a bummer, because you do want to see all the rides being used, but the guests are simply destroying fps. I guess I could go for 10-12 rides at stay above 20 fps, but as you can see, only about 40-45% of the map has been covered so far.... :

I'll finish this park as planned and see were the performance will land in the end. It's fun so far [praise]

 

[Bullethead]

Sad to see you're hitting the FPS wall despite such a valiant effort.

It seems we're in agreement that peeps are the major cause of FPS loss. That being the case, it behooves folks to determine how many peeps their systems can tolerate. Then, keeping in mind that since 1.2, all parks seem to attract a number of peeps equal to 150% of the park rating on all difficulties, the only way to stay within your system's capabilities is to hold park rating down so your number of peeps is within your system's tolerances.

But here's the thing.... Because only a few percent of peeps will ever use an ATM, by and large the only money you can get is what peeps bring into the park to start with, which is about $90 per peep. This means that the maximum potential monthly income of a park is determined by the number of new peeps who enter the park per month. The only way to increase potential income is to increase the number of peeps entering the park per month, which means increasing park rating, but that's limited by your system.

All that aside, your park looks great even though you're pulling your punches.

 

[edjenoh]

Sad to see you're hitting the FPS wall despite such a valiant effort.

It seems we're in agreement that peeps are the major cause of FPS loss. That being the case, it behooves folks to determine how many peeps their systems can tolerate. Then, keeping in mind that since 1.2, all parks seem to attract a number of peeps equal to 150% of the park rating on all difficulties, the only way to stay within your system's capabilities is to hold park rating down so your number of peeps is within your system's tolerances.

But here's the thing.... Because only a few percent of peeps will ever use an ATM, by and large the only money you can get is what peeps bring into the park to start with, which is about $90 per peep. This means that the maximum potential monthly income of a park is determined by the number of new peeps who enter the park per month. The only way to increase potential income is to increase the number of peeps entering the park per month, which means increasing park rating, but that's limited by your system.

All that aside, your park looks great even though you're pulling your punches.

I think the limitations of the engine are bigger, than the limitations on my system (it's about 2.5 years old, but still decent). I think the game in it's current state can't really handle a full park, which makes you wonder why they added the bigger maps to begin with...

But at least we can always just lower the peeps to about 1k and enjoy the park in decent fps, even though it will be a bit sparcely populated and not financially viable.

I think there could be a few fixes to improve this though:
-Improve peep movement speed, so they are walking to the rides faster
-Lower ride running costs and staff wages (although this would make the early game more easy), which could be fixed with:
-Drastically lower the ease with which one attracts peeps to the park (maybe 0.1-0.2)
-Increase the amount of folks that actively use ATMs, and perhaps transport rides

 

[edjenoh]

Yup, performance is already dipping below 20s now with more scenery, one more ride and 750 more guests. Oh well, bummer, but I can't abandon now, need to fill that canvas.

The show must go on..

I added another flat one (Sky ace)...

and mainly worked on the space area with a nice new plaza....

Also a shot from the jungle, rapid area, looking out over the space area.


The Boomerang and the Woody are up next after I finish more scenery.

 

[edjenoh]

Bad news, folks, I've truly hit the performance wall and I'm still short a whopping seven of the planned rides [where is it]

I made some updates with a brand new entrance and parking lot and futuristic boomerang inverted coaster.

But I'm affraid I will start a new full map performance park limiting it to 10-12 rides, and also running it on an easy challenge park, that way I can limit guest count a bit more while still making profit, allowing me to perhaps squeeze in a ride or two. Must succeed [big grin]

edit: some more angles of the park to show

Park entrance and parking lot

Space launce area with boomerang coaster

Other angle space flatride

Other angle Jungle Fever Rapids Ride

Jura River with housing and Garden Swing flat ride

Anaconda Giga Coaster back side

Overview park east side

Boulevard strip along the lake

The seven rides that were planned, but performance is hitting 13-19 fps depending on the area of the park, so it really isn't feasible to make a performance friendly park with this amount of rides and guests and have it be profitable at very hard challenge mode...

EDIT:
Did some more testing. On an empty park (and low guest pop of <1k) the performance swings between 19 and 29 depending on the location. The scenery density does influence fps quite a bit and I also noticed that the river rapids has lower fps than the arrow coaster in the 'dirt' hilly area, I think it probably has to do with the water effects or something, as well as the proximity of surrounding rides and scenery, since the dirt coaster has an empty area next to it, while rapids is at the center of the park.

Take-away message for the next attempt:
-Probably limit to 10 rides per park
-Watch out with scenery density and not use any blueprints with a high amount of objects
-Limit guest count to 2k or perhaps even 1k, run on easy challenge mode, so it will be more managable to be profitable and performance friendly (not sure how feasible that will be with only 1k, though, if it doesn't work, than there really is no other option than to go sandbox)