Q&A with Slightly Mad Studios: Ian Bell [Read OP]

[nyrmetros]

The track I most want to run my historic Indy cars on is Riverside International Raceway!

Agreed! Also Trenton, Ontario, and some Mid Ohio, Road America, etc...

 

[MroczMR]

Because they'd offer something unique to the game and create atmosphere. As long as it isn't like PD's fantasy really tight and short Pocono, we're fine.

Yeah... That track in GT Sport is Awful :/

 

[RacerDude08]

Fictional short tracks would please fans(at least American fans) and be a lot cheaper than getting a license for a real track.

On a side not. I hope PCARS 3 has dirt ovals... now that would be fun. Just look at the huge success iRacing just had with their dirt oval expansion

 

[MroczMR]

On a side not. I hope PCARS 3 has dirt ovals... now that would be fun. Just look at the huge success iRacing just had with their dirt oval expansion

Who knows... Maybe even DLC for PCars 2... They already have dirt physics.

 

[RacerDude08]

Who knows... Maybe even DLC for PCars 2... They already have dirt physics.

I'm interested to see how Live Track 3.0 works with dirt. Dirt tracks change a ton throughout a race, especially dirt ovals, so I would like to see how well that is simulated.

 

[jake2013guy]

Seeing as you can now officially talk about it, how does qualifying at Indy work? Will you be able to do the warm-up lap and then the 4 laps and have the opportunity to watch the other cars like F1 games of old?

 

[ozwheels]

If the in-game telemetry system from pCars 1 doesn't get much love. Please please please allow us to see damper shaft speeds so high speed damper tuning isn't such a pain in the 🤬.

 

[TheoSkullz12]

So agreed fictionals aren't a huge no no but realism is where it counts, i think if they could somehow get the additions of Pocono, Mid-Ohio, and Sebring this game would be solid. Sometimes i forget that i want them to hit majority of the disciplines. And i also forgot that we have the variations of those ovals such as the road courses and i didn't even know that daytona had a dirt variation.

 

[jake2013guy]

So agreed fictionals aren't a huge no no but realism is where it counts, i think if they could somehow get the additions of Pocono, Mid-Ohio, and Sebring this game would be solid. Sometimes i forget that i want them to hit majority of the disciplines. And i also forgot that we have the variations of those ovals such as the road courses and i didn't even know that daytona had a dirt variation.

I think we can agree on that fairly reasonably.

 

[TheoSkullz12]

Important questions regarding career mode. Will there be an actual point to choosing a car number this time. I feel like it was pointless to pick a number and be given a totally different one than what i chose. I would let it slide if the reason was because of your team mate had that number but idk.

When selecting your Name and Surname, i have encountered where (i guess depended on the length of name) the surname would be cut off or i couldn't put my last name (Richardson) like it would cut off the n. Has this been addressed and or looked at.

 

[HardRock]

In pCARS 1, the AI was using a simplified tire model IIRC, so they were affected by changing environmental conditions somewhat differently compared to the player. How was this issue tackled for pCARS 2? Since the tire model got even more advanced, I assume our CPU's still won't be enough to handle it for 40+ AI cars.

Another related question, if my assumption is right, and the AI still use a modified tire model, is there any chance for having an option that would allow them to use the full SETA model, with all the bells and whistles, when we are only racing against 1 or 2 of them? Or is that completely out of the question still, due to the resources needed to handle that or maybe the additional tuning that would come with allowing two tire models to be used by the AI?

 

[FuriousDemon]

I suspect AI will just need to be tuned on a per track basis, kind of how like PCARS 1 was. Since their tire model should be the same as in pCARS 1, they already have them dialed in for the most part, so I'm guessing more time to fine tune it under different weather conditions. Totally guessing here, it may not be as simple as it sounds.

 

[diesel97]

In pCARS 1, the AI was using a simplified tire model IIRC, so they were affected by changing environmental conditions somewhat differently compared to the player. How was this issue tackled for pCARS 2? Since the tire model got even more advanced, I assume our CPU's still won't be enough to handle it for 40+ AI cars.

Another related question, if my assumption is right, and the AI still use a modified tire model, is there any chance for having an option that would allow them to use the full SETA model, with all the bells and whistles, when we are only racing against 1 or 2 of them? Or is that completely out of the question still, due to the resources needed to handle that or maybe the additional tuning that would come with allowing two tire models to be used by the AI?

SMS is calling it a "simplified tire model" (just laymen terms) but if I had to guess they are completely different "tire models". I don't you could just take parts from "A and parts from "B" to make a "C" (SETA mode for 1 or 2 AI) it would mess everything up (completely different programs)so no IMO it can never happen.

 

[TexasTyme214]

In pCARS 1, the AI was using a simplified tire model IIRC, so they were affected by changing environmental conditions somewhat differently compared to the player. How was this issue tackled for pCARS 2? Since the tire model got even more advanced, I assume our CPU's still won't be enough to handle it for 40+ AI cars.

Somewhere in this thread Ian said that the AI use a modified brush model, which is used in many Sims

 

[ozwheels]

@HardRock

Just want to interject here to say that our AI are using a full brush model which as The American says, is simplified compared to our player car seta model, but it is also the same physics used by some of the best of our competition 'for' the player car...

4 AI cars running our Seta model would take up all of the power of a PS4 (and an Xbox). It's the first real time fully dynamic node based model used in a commercial game. Our AI run a brush model, like AC and Shift use for the player cars.

 

[nyrmetros]

So agreed fictionals aren't a huge no no but realism is where it counts, i think if they could somehow get the additions of Pocono, Mid-Ohio, and Sebring this game would be solid. Sometimes i forget that i want them to hit majority of the disciplines. And i also forgot that we have the variations of those ovals such as the road courses and i didn't even know that daytona had a dirt variation.

Those would be great track additions!!!

 

[RomKnight]

Somewhere in this thread Ian said that the AI use a modified brush model, which is used in many Sims

In pC1 it really was a simplified model but the optimisations and rewritings the magics and whatnots SMS pulled the AI evolved to the known since forever in simracing BTM.

Funny how simple it was and yet it was the best that we got for so long. Also means I'm getting old...

 

[FuriousDemon]

In pC1 it really was a simplified model but the optimisations and rewritings the magics and whatnots SMS pulled the AI evolved to the known since forever in simracing BTM.

Funny how simple it was and yet it was the best that we got for so long. Also means I'm getting old...

You're saying in pc1 it was not BTM but something simpler?

 

[HardRock]

I suspect AI will just need to be tuned on a per track basis, kind of how like PCARS 1 was. Since their tire model should be the same as in pCARS 1, they already have them dialed in for the most part, so I'm guessing more time to fine tune it under different weather conditions. Totally guessing here, it may not be as simple as it sounds.

Sure, more time tuning things will definitely help, but the issue is that different tire models won't react in exactly the same ways to driving conditions, which was partly why the AI's performance on track varied wildly relative to the player's in pCARS 1.

To use a really simple example, let's say that with the SETA model an alien can go around a corner at 100 kph in the dry, while the best AI can do 105 kph with the brush model. Both are at the limits of their respective physics would allow, so you dial back the AI, right? Well, here's the problem. The AI could easily be slower already in the same corner in the wet, because their tire model could be more sensitive in wet conditions, or maybe at higher surface temperatures, or when more dirt is on the track, etc.

What I mainly wanted to know is how this problem is approached for pCARS 2. Is it even seen as a problem? Was the brush model of the AI maybe extended with new features to bring it closer to the SETA model in varying conditions? Am I asking too many questions?

 

[TexasTyme214]

You're saying in pc1 it was not BTM but something simpler?

A simpler BTM. You can do a lot with various tire models.

 

[ozwheels]

Sure, more time tuning things will definitely help, but the issue is that different tire models won't react in exactly the same ways to driving conditions, which was partly why the AI's performance on track varied wildly relative to the player's in pCARS 1.

To use a really simple example, let's say that with the SETA model an alien can go around a corner at 100 kph in the dry, while the best AI can do 105 kph with the brush model. Both are at the limits of their respective physics would allow, so you dial back the AI, right? Well, here's the problem. The AI could easily be slower already in the same corner in the wet, because their tire model could be more sensitive in wet conditions, or maybe at higher surface temperatures, or when more dirt is on the track, etc.

What I mainly wanted to know is how this problem is approached for pCARS 2. Is it even seen as a problem? Was the brush model of the AI maybe extended with new features to bring it closer to the SETA model in varying conditions? Am I asking too many questions?

I believe what your asking is how do the AI deal with changing track conditions. If that's the case Ian or someone has stated the AI see Live Track 3.0 just as the player does.

 

[David Brooks]

Sure, more time tuning things will definitely help, but the issue is that different tire models won't react in exactly the same ways to driving conditions, which was partly why the AI's performance on track varied wildly relative to the player's in pCARS 1.

To use a really simple example, let's say that with the SETA model an alien can go around a corner at 100 kph in the dry, while the best AI can do 105 kph with the brush model. Both are at the limits of their respective physics would allow, so you dial back the AI, right? Well, here's the problem. The AI could easily be slower already in the same corner in the wet, because their tire model could be more sensitive in wet conditions, or maybe at higher surface temperatures, or when more dirt is on the track, etc.

What I mainly wanted to know is how this problem is approached for pCARS 2. Is it even seen as a problem? Was the brush model of the AI maybe extended with new features to bring it closer to the SETA model in varying conditions? Am I asking too many questions?

I have exactly these same concerns. In pcars 1 career I generally find I am competitive in the dry bit can't keep up in the wet. This means I have to dial back the AI in wet races and end up pulling out a huge lead in the dry only to have it disappear in the wet. It makes it less about wheel to wheel racing and more like a handicap reverse grid race.

It could just be that I am just not as good in the wet as I should be, but I think it is more than that. AI shouldn't be able to pass me on a wet track around the outside when I feel I am on the line and already at the limit of grip, especially when I can kick them in the dry.

Having said that, perhaps the physics changes around the limit of grip being more gradual and controllable will help me push harder in the wet and address my parity issue with the AI?

 

[HardRock]

I believe what your asking is how do the AI deal with changing track conditions. If that's the case Ian or someone has stated the AI see Live Track 3.0 just as the player does.

That's not quite what I mean. I'm sure the AI is affected by the parameters of Live Track 3.0. What I'm curious about is how the different physical limits between player and AI, due to the separate tire models, is being reconciled and if the process is any different than it was for pCARS 1.

 

[RomKnight]

Well, I tried (and failed) to find if in pC1 simplified tyre model was even based on BTM. It was worse than it is in pC2.

In anycase, if you took the late braking/aggressiveness at specific turns (depending on the track/car combo) AI wasn't as bad as one may think.

But, this is about pC2 and as everything the AI was also (much) improved.

 

[FuriousDemon]

Well, I tried (and failed) to find if in pC1 simplified tyre model was even based on BTM. It was worse than it is in pC2.

In anycase, if you took the late braking/aggressiveness at specific turns (depending on the track/car combo) AI wasn't as bad as one may think.

But, this is about pC2 and as everything the AI was also (much) improved.

For me AI in pCARS was good, other than their behavior in the rain. Good to hear it's much improved. I wonder how it compares to the latest AI patch in AC. Looks like it's really good now there.

 

[bleeder]

For me AI in pCARS was good, other than their behavior in the rain. Good to hear it's much improved. I wonder how it compares to the latest AI patch in AC. Looks like it's really good now there.

To me the one standout issue with the pCARS AI was the open wheelers' behaviour. They drove almost like they were in GT cars. If that's sorted then I'm happy. They need some self-preservation built in.

Also IMO the AI in AC on console isn't that good or particularly well-balanced because they're using Easy/Med/Hard etc presets, and I know they might get updated to PC behaviour levels, but without the difficulty percentage slider I don't think it'll ever match the tight races we are able to have in pCARS on console.

 

[Kennytomson]

So agreed fictionals aren't a huge no no but realism is where it counts, i think if they could somehow get the additions of Pocono, Mid-Ohio, and Sebring this game would be solid. Sometimes i forget that i want them to hit majority of the disciplines. And i also forgot that we have the variations of those ovals such as the road courses and i didn't even know that daytona had a dirt variation.

another vote for Pocono and Sebring.

 

[FuriousDemon]

To me the one standout issue with the pCARS AI was the open wheelers' behaviour. They drove almost like they were in GT cars. If that's sorted then I'm happy. They need some self-preservation built in.

Also IMO the AI in AC on console isn't that good or particularly well-balanced because they're using Easy/Med/Hard etc presets, and I know they might get updated to PC behaviour levels, but without the difficulty percentage slider I don't think it'll ever match the tight races we are able to have in pCARS on console.

Agree about open wheelers. I have AC on PC, didn't know it's different on console in terms of the difficulty.

 

[RomKnight]

've uninstalled AC from my PC. Have no use for it so I can't compare the AI.

For me AI in pCARS was good, other than their behavior in the rain. Good to hear it's much improved. I wonder how it compares to the latest AI patch in AC. Looks like it's really good now there.

I've uninstalled AC from my PC. No use for it nor the time anyway

 

[bleeder]

've uninstalled AC from my PC. Have no use for it so I can't compare the AI.

I've uninstalled AC from my PC. No use for it nor the time anyway

Yeah I don't play it anymore myself after they introduced a nasty bug where the brakes didn't work on any of your old setups. Can't be bothered to go back and input all the changes I'd made for all the cars, and coupled with the low track and AI count I've pretty much given up on it. At least for the time being.

The great physics and handling aren't the be all and end all for me, I kinda need more.