PC1 False advertising

I've seen many evangelists and perhaps even the developers themselves refer to Project C.A.R.S. as a "sim" or "simulator". Normally this term is used to describe software that tries to mimic reality. There is nothing wrong with arcade racers so why not just call the game by its real name and stop trying to advertise it as something it's not? There's no shame in admitting that Project C.A.R.S. doesn't try to be realistic. There's a big market for this type of game, even bigger so than for the niche sims, so please stop trying to fool people and aim for your target audience instead.
 
The FFB is what brings the car behaviour to life, and also acts as our early warning system regarding grip levels, as such, the more detailed the FFB, the more we can control the car as we do real cars.

Real life is 1:1, so any sim that gets close to giving us this level of feel is on the right track, and that must come thru the wheel and synchronize with the pedals....GSC does this extremely well, and so does Simbin in many of their cars.
And so does Forza 4. FM4 gives a ton of info through the wheel yet many here don't consider it a sim. Infact I'd said FM4 does a better job at it than the simbin titles as all cars in FM4 give info whereas in the Simbin titles some cars are great whereas some cars you can barely feel any info outside of track bumpiness.
 
And so does Forza 4. FM4 gives a ton of info through the wheel yet many here don't consider it a sim.

Evo's strong on simulated braking, weight transfer and feels extremely connected when you slide or completely get it wrong....as for FM/GT5 sim status, I don't own them, but people often start threads saying how difficult it is/was adapting to PC sims, and that just sounds to me like going from Shift1 on PC back to Evo where I spun for hrs as I could no longer mash the brakes.
 
Very interesting account of their move to Washington. Moving times are always hectic and full of tension, they're also great experiences. Good move ISR on sharing that.
I want to see the "uncut" version, or the "lost tapes", where they were at each others throats from time to time because you know they were. lol

Evo's strong on simulated braking, weight transfer and feels extremely connected when you slide or completely get it wrong....as for FM/GT5 sim status, I don't own them, but people often start threads saying how difficult it is/was adapting to PC sims, and that just sounds to me like going from Shift1 on PC back to Evo where I spun for hrs as I could no longer mash the brakes.
Don't know about Shift 1, I'm not going to knock it since I never played it, but I know with Forza I can go from iRacing to Forza to Race to rFactor and not skip a beat. Of course different cars and track combos will give me issues if I'm not familar with them but it isn't because one is "more sim" than the other. "simulated braking, weight transfer and feels extremely connected when you slide or completely get it wrong" you get all of that in FM4 and then some. All of that is there. But keep in mind EVERY sim I've played is a little bit different than the other. Each of them have something where you approach it a little different and it makes sense. Even with two identical cars on different sims something is going to be a little different and it should be.

Did both game makers model the EXACT same car, probably not. What tires were on it? DId they both have the same mileage? What kinds of miles are on it? The list goes on so these differences are to be expected. Turn10 uses Pirelli for their tire data. For all we know iRacing could be using Goodyear and Race 07 Falken. This would make a difference.

I'll tell you this, when it comes to simulating weight transfer in a game nothing beats what Forza is doing. Now of course this is all from a wheel users perspective. And what makes my observations somewhat credible is that I'm using the same , Fanatec GT2, wheel for all these games since it's 360/PS3/PC compatible. What I wish I could find is some absolute great wheel settings for each game. That's very hit or miss and even in the game it changes with the same car. Sometimes, in iRacing, I have to restart the session before it "kicks in". Don't know what that's about.

Like I said before if two guys play the same game, take the same car on the same track and one dude is going balls out while the other guy is using correct racing technique, if the guy using the correct racing technique is fast then I consider that game a sim. Or at least a great foundation for one.
 
but I know with Forza I can go from iRacing to Forza to Race to rFactor and not skip a beat.
Maybe i´m alone but the only games i can do that with is iRacing, NKPro and FVA.

The rest feels like a totally different set of physics.
The above to me feels like it has the same physics (i guess the most realistic) just different cars.

With iRacing you really have to discount some cars because of the tire model being in different stages etc but if you drive the HPD then jump into the Osella it will feel quite similar in physics.

Other sims doesn´t give me half the feedback that i get from the above.
It´s a different solar system in feel to me.
 
Like I said before if two guys play the same game, take the same car on the same track and one dude is going balls out while the other guy is using correct racing technique, if the guy using the correct racing technique is fast then I consider that game a sim..

Not a bad measure.
As far as which is the best most accurate sim, I don't have absolute certainty.....one thing we could do is combine our knowledge{the forum} and create a reality checklist, although even that can be tainted by some degree of subjectivity.

The biggest problem I have is when people say pcars is better than sim x, because not only do I get annoyed by what seems preposterous, I also worry that real sims can fade away if enough people get magnetized by GFX.
 
The biggest problem I have is when people say pcars is better than sim x, because not only do I get annoyed by what seems preposterous, I also worry that real sims can fade away if enough people get magnetized by GFX.
People have been known to be attracted to shiny things for sure. In the end tho, it'll all come around full circle. Eventually, if quality in simulation physics starts to ween out over the years, somebody will put more effort into physics to beat out the grafix guys. I doubt realism will ever completely go away, thats why we're all in this in the first place.

I agree with a lot of what t.o. says though. I can jump in and out of different sims no problem. If you have enough knowledge of racing and racecraft, you will have the advantage, no matter how easy or difficult it is to drive. Even a mid paced guy like myself would be a bit faster than some kid just starting out, regardless of wether or not the car rides on rails. If the cars in any sim, let alone pcars, drove on rails though, I would not drive them. There is no challenge.

I also agree with David on the braking aspect adding realism. In Iracing, if I miss a braking point, i'm dead in the water 9 out of 10 times. In pcars, if I miss a braking point, that number is more like 3 out of 10 times. While you cant "burnout paradise it" around the turns, it doesnt punish you quite as bad as some of todays sims. It will be interesting to see how the new tire model comes into play, not only in handling, but in braking too :)
 
The biggest problem I have is when people say pcars is better than sim x, because not only do I get annoyed by what seems preposterous, I also worry that real sims can fade away if enough people get magnetized by GFX.

I think this notion that sims and good graphics are somehow in conflict is a recent thing, perhaps because of the stagnation in racing sims since gMotor2 came out. In the past, sims pushed graphics hard. Crammonds GP series, Papy's Indycar, NASCAR and GP Legends all required enormous graphics power when they came out and looked amazing if your PC was up to it. RBR too had excellent graphics for its day and required a powereful PC. Simbin's GTR, GT Legends and GTR2 again all had jaw-dropping graphics for their day.

I would also argue that the "pretty" console driving titles like Gran Turismo and Forza are getting more sim-like with each release along with their improved graphics. For example they now feature damage and cockpit view as well as increasingly sophisticated physics engines such as modelling manual gearchange far better than rFactor for example.
 
I also agree with David on the braking aspect adding realism. In Iracing, if I miss a braking point, i'm dead in the water 9 out of 10 times. In pcars, if I miss a braking point, that number is more like 3 out of 10 times.

Perhaps I'll change my mind if I had CSP's, but the more I think about iracing, the more I think they've gone too far, and are a difficulty simulator.

In Evo and other sims that I have, if you overshoot, you usually lock up and run wide a bit, you then miss the apex and lose time trying to drag it around, and we can see this many times in real life racing, in fact Moto GP and their helicopter view does a good job of exposing that missing a braking point shouldn't automatically equal a huge off or snap spin.
 
. For example they now feature damage and cockpit view as well as increasingly sophisticated physics engines such as modelling manual gearchange far better than rFactor for example.

Good point, and these type of physics can certainly give one game/sim and edge over another.....but I still feel that Evo/Netkar/GSC are about right, mix the best of them into one sim and physics are close to sorted.
 
Perhaps I'll change my mind if I had CSP's, but the more I think about iracing, the more I think they've gone too far, and are a difficulty simulator.

In Evo and other sims that I have, if you overshoot, you usually lock up and run wide a bit, you then miss the apex and lose time trying to drag it around, and we can see this many times in real life racing, in fact Moto GP and their helicopter view does a good job of exposing that missing a braking point shouldn't automatically equal a huge off or snap spin.
Yeah I'm not a fan of the "tiny mistake and wind up in the stands" model either. When you think about iRacing everyone holds it up more because real drivers participate in it. However, from what I've seen, most of the real drivers are in oval racing not so much in the road aspect of it.
 
In fairness to Iracing here, I loved every minute of playing it, it had me utterly addicted and where it not for the failed physics of a couple of cars namely the Spec Racer and Solstice I could have paid up!

But like things like WoW and others, I simply can't justify paying that amount per month! I know you can get deals, but then I would feel like I had to play the damn thing, and to then be asked to buy cars, tracks and the like I just think is bit of a rich con when you are paying a fee far more than the RRP of most games per year anyway.

Americans will do it because, well theya re Americans and they love NASCAR which is what it really is all about!

But as a month free user a year or so ago I did very well, got up two licences in that time (oval and road) and was on the verge of doing well until the Spec Racer lulled me in and I lost points and ratings in one race, damn thing!

Regarding ovals, the Legend in particular is a very, very realistic car. I raced against quite a few guys who raced them for real and they said they could use IR for quite a lot of things, track guides, gearing, revs etc. I found those cars amazing fun, I won a lot and beat some pro guys. Even the street stock was fun, harder to master. But the close, amazing races I had in Legends have been unbeaten in any sim I played, some of my times were top 100 worldwide and I guess being good helps you love something! but even so it felt so real! Errors were punished and smoothness, learning, bravery were rewarded.

And this is the crux of PCARS, I just have never felt that immersed in the physics, it never pulls you in, it never feels real, and remember this was IR before the tyre revamp and stuff, with mainly default settings.

The satisfaciton of qualifying on pole at Laguna in a Skippy, having a duff start then passing a few and not touching a verge or car to finish 2nd and go up nearly half a mark was amazing!

And PCARS, it might prove me wrong, will never, ever be able to do this. It is nowhere near.

A proper simmer doesn't give a damn about pretty, yes its nice, but I have played some amazing mods on RF that feel amazing, allow great racing and look like poo!
 
A proper simmer doesn't give a damn about pretty, yes its nice, but I have played some amazing mods on RF that feel amazing, allow great racing and look like poo!

For me the graphics are one of the things that contributes to the immersiveness you talk about. As I've said already - sims have traditionally pushed graphics hard. You may not be interested but I think of myself as a "proper simmer" and I do give a damn. I want I want realistic physics AND realistic graphics :)
 
There's always different goals that different people have out of games or softwares.

One aspect of simulations is modelling a real world system on a computer, gaining some benefit in terms of cost or reproducibility as a result. The complexity will vary depending on what the goals are, as it's impossible to model the real world to an absolutely correct fidelity. In my job it takes weeks of compute time on large servers just to simulate a fraction of a second of something that's made up of only a few thousand atoms - and even then we have to swap out accurate systems for rigid assumptions and short cuts to get the compute time down to something possible. Move up to the size of a car and you have to take even more short cuts, move up to simulating in real time and you have even more.. to the point where you simply cannot concentrate on everything and have to focus on certain areas for the goals you are trying to achieve.

In F1, the simulators are primarily used to assess what gives the best lap time - be that parts, setup or driving line etc. so the fidelity is concentrated on getting that bit as accurate as possible, and everything else is only important insofar as it helps that goal (gives the driver enough feedback that they will respond in the same way they would on the track for example, but no more).

On the other hand, games have more widely ranging goals - primarily to be fun, but they also differ in how they achieve that. Some concentrate on trying to match lap times with reality along the lines of the real simulator, others on making you feel immersed in the racing experience so 'simulate' the driving experience, others on maximising adrenaline, or overcoming challenges, or just plain silliness and humour.

I like a range of games, but in general even in 'simulators' I tend towards enjoying simulating the driving experience over technical accuracy of the model - that means I expect the kinds of changes to improve lap time to be broadly consistent with the kinds of changes a racing driver would expect to do to improve lap time, but I also need the game to compensate for the fact I am not a racing driver in real life - I want the game to make me feel like one though, just like I do with games where I pilot a jet, or trek up a mountain to battle a skeleton hoard.. ok so there's not much of a real life reference in the last one, but you know what I mean.

Things that take away from the expected behaviour take away from that immersion, whether that's a setup change not having the expected result, or a handling model that rewards unexpected driving styles. But they don't need to be simulated perfectly accurately to the point where because I'm not a racing driver, I wouldn't be able to take advantage of them or even tell the difference.

Graphics can aid different goals of course - whether that's recreating the experience accurately, or enhancing it to compensate for the loss of information playing via a computer rather than being in the car, or going past that to forming pleasure in and of itself (beautiful sunsets and scenery for eg.). And to add to the mix, it's really subjective, and is highly influenced by our expectations which are altered by technology and other games we see around us - I remember thinking Micropose GP was a stunning advance over Revs, now I'd struggle to enjoy it.

So I guess I'm not really in the category of hard core sim racers, and I like games like GTR:Evo precisely because they are aware of compensating a little for the fact you are sitting at a computer, and are probably not a professional racing driver - as a result they make me feel more like one.
 
So I guess I'm not really in the category of hard core sim racers, and I like games like GTR:Evo precisely because they are aware of compensating a little for the fact you are sitting at a computer, and are probably not a professional racing driver - as a result they make me feel more like one.

Evo is very hardcore IMO, ie, it makes me believe I'm driving cars/race cars low speed, med and high speed.
It sounds like you got beaten up with either iracing/rf2 and think that extreme difficulty/botched physics equates to realism.
 
An interesting post, kalniel, if I may say so.

But...

There's always different goals that different people have out of games or softwares.

One aspect of simulations is modelling a real world system on a computer, gaining some benefit in terms of cost or reproducibility as a result. The complexity will vary depending on what the goals are, as it's impossible to model the real world to an absolutely correct fidelity. In my job it takes weeks of compute time on large servers just to simulate a fraction of a second of something that's made up of only a few thousand atoms - and even then we have to swap out accurate systems for rigid assumptions and short cuts to get the compute time down to something possible. Move up to the size of a car and you have to take even more short cuts, move up to simulating in real time and you have even more.. to the point where you simply cannot concentrate on everything and have to focus on certain areas for the goals you are trying to achieve.

Yes. But if you do an experiment whereby you accelerate particles with charge up to relativistic speeds and make them collide, you'd better be ready to have the network of detectors catch all the by-product particles created. Similarly, if you simulate that you have to take into account every single particle collided and all short-lived and long-lived particles that appear as a result of that. Detail is critical here, 100% detail is vital to gouge the accuracy of simulations just prior to real-life experiments.

This is not so with commercial simulators. There are, as you say, too many variables involved (some of which science itself is still not certain of) and to fully, accurately simulate them all is wholy impossible with present-day cpu+gpu technology.

Not only that, it is also unnecessary. You certainly don't need 100% accuracy (given by simulating everything up to thermodynamics level) and not 99% - not even for professional simulators such as Cruden or rF Pro.

But the fact that we cannot aim that high with our present technology does not mean we cannot have standards and/or expect high physics fidelity from commercial grade car simulations. We can and should.

[...] but in general even in 'simulators' I tend towards enjoying simulating the driving experience over technical accuracy of the model

I personally respect this. However, the reasoning you established for it obviously disregards what I said above - just because it is not possible to simulate every object/entity in the system (road+áir+car+tires+dirt) with our present tech, does not mean that it is not simulating it. It is. It may not simulate every bristle of the tire, it may not simulate every rubber molecule interacting with other rubber molecules and lubricants and road and dirt, but it is still attempting (quite well) to simulate this system.

Maybe the question we should be asking is this: does it matter? Put it another way: what does matter for a racer or a simracer?

Well, forces, car behaviour, tire behaviour, collisions, trajectories. If, at the end of the day, a FULL 100% representation of that system yields 99% accurate results, and a commercial grade simulation yields 94 to 95% fidelity (some things may be a bit lower, others may be slightly higher), does it really matter?

Is it really a fail if a commercial grade calculates for a certain car, at a certain speed, at a certain corner, 2.31g whereas a hyper-realistic, 100% representation of the system tells you that the lateral acceleration is actually 2.42? In my humble opinion, no. It would if there was 0.5g or 1.0g difference, but not 0.11g.

So I guess I'm not really in the category of hard core sim racers, and I like games like GTR:Evo precisely because they are aware of compensating a little for the fact you are sitting at a computer, and are probably not a professional racing driver - as a result they make me feel more like one.

Well, you assume GTR Evo is not hardcore, when in fact it is. SIMBIN reworked things from GTR2 to Race07 and further for GTR Evo.

If one knows exactly what to do (and we have a couple of known modders right here that know what the process entails), we can make GTR Evo even better - close, very close to FVA, which is considered by many the best thus far.

Obviously, this is only my opinion.
 
Evo is very hardcore IMO, ie, it makes me believe I'm driving cars/race cars low speed, med and high speed.
It sounds like you got beaten up with either iracing/rf2 and think that extreme difficulty/botched physics equates to realism.
No I don't think that at all - it's not a two-way relationship. A 'realistic' car driving model is very difficult to drive fast, that's why there are professionals. That doesn't mean that a difficult game is therefore realistic though. It does mean that an easy game is not 'realistic' unless it's also easy to drive those cars fast in real life. But shooting for that kind of realism is going down the route of simulation/modelling for a sake other than making you feel like you are a racing driver, and isn't of interest to me or some other people. But humans are multitudinous and there are no doubt some people who like that sort of thing, in fact some games have little point other than to present a hard challenge and people simply enjoy conquering that challenge for its own sake - it's not for me to judge how other people should enjoy their games.

Yes. But if you do an experiment whereby you accelerate particles with charge up to relativistic speeds and make them collide, you'd better be ready to have the network of detectors catch all the by-product particles created. Similarly, if you simulate that you have to take into account every single particle collided and all short-lived and long-lived particles that appear as a result of that. Detail is critical here, 100% detail is vital to gouge the accuracy of simulations just prior to real-life experiments.
Detail in validating a model is not the same thing has needing a detailed model, but yes, you need the detail in the critical areas.

This is not so with commercial simulators. There are, as you say, too many variables involved (some of which science itself is still not certain of) and to fully, accurately simulate them all is wholy impossible with present-day cpu+gpu technology.

Not only that, it is also unnecessary. You certainly don't need 100% accuracy (given by simulating everything up to thermodynamics level) and not 99% - not even for professional simulators such as Cruden or rF Pro.

But the fact that we cannot aim that high with our present technology does not mean we cannot have standards and/or expect high physics fidelity from commercial grade car simulations. We can and should.
To what end though? For the sake of it is well and good, but is not efficient use of money/time. I'm suggesting that my gaming style requires that you should have the physics fidelity needed to give you the impression of being a racing driver. Any other physics modelling is wasted effort that could be better spent on improving the physics/compensations/enhancements better for me.

Now to someone else that's not the case - they want to have the most realistic model on their computer possible, even if that means either a) they have to be a professional racing driver to drive fast. b) the actual experience of playing at the computer is less comparable to driving the car. Underneath the numbers are more right (though still wildly approximated of course) and that is what some people are after, which I completely respect.

I personally respect this. However, the reasoning you established for it obviously disregards what I said above - just because it is not possible to simulate every object/entity in the system (road+áir+car+tires+dirt) with our present tech, does not mean that it is not simulating it. It is. It may not simulate every bristle of the tire, it may not simulate every rubber molecule interacting with other rubber molecules and lubricants and road and dirt, but it is still attempting (quite well) to simulate this system.
I don't think we're in disagreement here - the question is more of intent: Is the best way of giving the driving experience to simulate as much as we can, or is it to simulate to a lesser extend and tweak things to compensate for the computational power and media the game is presented through?

With computational models you often find that if you can't simulate one bit properly, you have to adjust the rest of it arbitrarily to bring behaviour back to expectations - we add in coefficients, correction factors and all sorts.

Maybe the question we should be asking is this: does it matter? Put it another way: what does matter for a racer or a simracer?


Well, forces, car behaviour, tire behaviour, collisions, trajectories. If, at the end of the day, a FULL 100% representation of that system yields 99% accurate results, and a commercial grade simulation yields 94 to 95% fidelity (some things may be a bit lower, others may be slightly higher), does it really matter?

Is it really a fail if a commercial grade calculates for a certain car, at a certain speed, at a certain corner, 2.31g whereas a hyper-realistic, 100% representation of the system tells you that the lateral acceleration is actually 2.42? In my humble opinion, no. It would if there was 0.5g or 1.0g difference, but not 0.11g.
Yep, and I'm just taking that to its conclusion by saying is it really a fail if you, as the driver, feel no difference between the 95 and the 99% accuracy? Now move on a bit and ask: "If I can drop to that 95% level and it saves me some time/money, can I put that time/money towards enhancing another aspect of the game that makes the player feel more like a driver, like say FFB or screen representations of weight transfer via floating huds etc.?" And it goes on - maybe I can simulate brake fluid movement to a 10% accuracy and the player doesn't notice - now I've got even more money to spend on other areas they will notice.

Well, you assume GTR Evo is not hardcore, when in fact it is. SIMBIN reworked things from GTR2 to Race07 and further for GTR Evo.
With GTR:Evo they worked on improving the driver experience - I don't think the sim is massively higher fidelity (might even be less in some areas) - but they enhanced some things to help compensate for the fact you're sitting behind a screen - sounds, exaggerated motion to reflect weight transfer etc.

If one knows exactly what to do (and we have a couple of known modders right here that know what the process entails), we can make GTR Evo even better - close, very close to FVA, which is considered by many the best thus far.
But how do you define 'better'? ;) To me, it's about recreating the racing driver experience, to which underlying physics etc. are a useful tool, over and above 'realism', which when experienced behind a screen might actually seem less like the driving experience :)
 
No I don't think that at all - it's not a two-way relationship. A 'realistic' car driving model is very difficult to drive fast, that's why there are professionals.

The above is in response to David, but if you don't mind I'd like to comment on it.

I have touched on that subject previously (on the difficulty and realism issue) and elsewhere, so no point in revisiting the whole matter.

But no. a realistic car driving model is difficult because it will allow you (or should) to do as they (pros) do. Actually, in order for you to be really fast in a realistic simulation, you have to work hard and follow the same steps/techniques real life pros do. To do that you have to max out every single time on a number of parameters - concentration being one of them. This is not easy.

Professional drivers discussed this issue with me, so I'm 100% behind their ideas and claims in this respect: a realistic physics model does not make driving difficult. The experience is made more difficult due to what I mention above. For me, that is all there is to this matter.


That doesn't mean that a difficult game is therefore realistic though. It does mean that an easy game is not 'realistic' unless it's also easy to drive those cars fast in real life. But shooting for that kind of realism is going down the route of simulation/modelling for a sake other than making you feel like you are a racing driver, and isn't of interest to me or some other people.

Indeed. Some people are content with pushing every single button in the start sequence of a F-16 C Block 50, and for them Microprose produced Falcon 4. Others can't stand that level of detail and seek the Ace Combat series produced by Namco.

Big difference indeed.

Going down the route of simulation for the sake of simulation is what pushes the envelopes for simracing development. High fidelity physics requires massive computational power, and on top of this we have AI, shiny graphics and great sound. Developers have to scratch their heads hard to make it all work. Having shiny graphics and merely acceptable physics (a la Novalogic flight simulation era) does not cut it in the long run; on the other hand, having all this on top of a good,solid physics engine changes everything and allows for a far better racing experience.

But humans are multitudinous and there are no doubt some people who like that sort of thing, in fact some games have little point other than to present a hard challenge and people simply enjoy conquering that challenge for its own sake - it's not for me to judge how other people should enjoy their games.

Kalniel, I see more people frustrated over unwinnable races against the AI in eye-candy racing games such as S2U than I see people frustrated over the likes of FVA or Race Room 2.

Curiously, the more developers build on on the gfx & sounds front, the more difficult they seem to forcefully make their racing games - probably to atone for the lack of a proper physics engine?

Detail in validating a model is not the same thing has needing a detailed model, but yes, you need the detail in the critical areas.

One cannot live without the other. I imagine you work in high energy particle software systems, I have also been working in an industry for a number of years for which the demand for detailed, adaptable systems is very high.

From my vantage point, for sure.

To what end though? For the sake of it is well and good, but is not efficient use of money/time. I'm suggesting that my gaming style requires that you should have the physics fidelity needed to give you the impression of being a racing driver. Any other physics modelling is wasted effort that could be better spent on improving the physics/compensations/enhancements better for me.

I have no doubt games development investors and publishers like EA/Virgin/Namco think exactly like that.

Your gaming style, as you call it, has different requirements from mine, and does not inevitable translate to being more viable from a commercial point of view or a financial point of view.

If we were talking about a simulation that took to the extreme of simulating how an isolated, out of sight fuel tube bounces and interacts with moving parts or walls, then I'd say you're right, no point in going that far.

In regards to tire physics, chassis physics and aerodynamics, I completely disagree. Any added detail is not wasted effort, it is however yet another step in the ladder towards greater realism. And the same can be said about the engine - even though it may, for instance, not necessary to fully model the temperature variations of a piston throughout a stint.

I don't think we're in disagreement here - the question is more of intent: Is the best way of giving the driving experience to simulate as much as we can, or is it to simulate to a lesser extend and tweak things to compensate for the computational power and media the game is presented through?

Ok.

What is the racing experience? If you (or anybody) is willing to live with a little bit less realism in order to increase detail in other areas, than you have to face the inevitable question: can we talk about racing experience as the most valuable trait in a racing sim when you are not able to mimic that experience fully? Are you able to, in pursuit of the "exciting racing experience", simulate most sounds (even those less frequent)? No. Are you able to simulate the temperature inside a cockpit or a cabin? No (unless you go to funny extremes like a guy that dresses up in sauna-like suits before each race). Are you able to replicate most events in a cabin (vibration, smells, heat, sound)? No.

So, what you call "racing experience" is much more compromised than what we hardcore want: physical realism.

Plus, we should not forget something: it is far less complicated to produce premium gfx and sounds than it is a first rate physics engine, it is relatively easy to hide a third rate physics engine behind a wall of hyped up absurd marketing claims and eye-candy gfx.


Yep, and I'm just taking that to its conclusion by saying is it really a fail if you, as the driver, feel no difference between the 95 and the 99% accuracy?

No, no fail. Double that difference and I am certain it is a fail and any competent, intelligent driver would know the difference.

Now move on a bit and ask: "If I can drop to that 95% level and it saves me some time/money, can I put that time/money towards enhancing another aspect of the game that makes the player feel more like a driver, like say FFB or screen representations of weight transfer via floating huds etc.?"

That is exactly what is happening nowadays due to limitations of current cpu-gpu abilities. So, in order for you to take that to the letter, we would be talking about differences in accuracy as big as 10 to 15%, which in my opinion is unacceptable.

With GTR:Evo they worked on improving the driver experience - I don't think the sim is massively higher fidelity (might even be less in some areas) - but they enhanced some things to help compensate for the fact you're sitting behind a screen - sounds, exaggerated motion to reflect weight transfer etc.

You may not see the improvements in the tire model and globally in the physics engine, but they are there and SIMBIN told us about it.

I disagree with you on your take of what SIMBIN did with GTR Evo.

But how do you define 'better'? ;) To me, it's about recreating the racing driver experience, to which underlying physics etc. are a useful tool, over and above 'realism', which when experienced behind a screen might actually seem less like the driving experience :)

The fact that we cannot get simracers killed or hurt (other than standing up, tripping and falling over gear and monitor) or sweaty does not mean physical fidelity is out of reach.

For you, the subjective "racing driver experience" is the most important. I respect that.

Accept that, for me, the very objective physics fidelity is the most important thing. :)
 
Again I think we're more in agreement than it sounds - your examples of Namco and EA are examples I'd point to as not ones interested in the driving experience - they clearly have some other goal - cheap thrills, wowy explosions or whatever, which impacts on the modelling in such a way that it does not behave as you would expect or respond to your inputs the way a racing car would respond to a racing driver, so you have neither accurate physics nor a racing driver experience.

I am fairly confident (though now lacking either a source or as accurate a memory as I used to) that the reason tyre physics were added to games like GTR:Evo is precisely because that helped create a better driver experience - things behave more like you expect, your inputs have more realistic outcomes, and the exaggerated movement helps make up for missing senses. I am sceptical that the tyre parameters were chosen a priori, or even with absolute reference to real data, but instead once the model was adapted they were adjusted until such a point that it 'felt' right and realistic. Something I presume is happening with pCars as well (ooh back on topic!). The other significant gain is for marketing - if you say your game has modelled tyre physics it's another check box in the feature comparison list that enthusiasts look for. Some people will like that for its own sake or the sake of making the game more realistic, others will like it for the hope that it helps recreate the driving experience better.

I do completely understand the enjoyment in having a game modelled as accurately as possible - modern machines are marvels and I think it's a great use of them to try to simulate something like a racing car. But I don't think I've ever played a game that's marketed to the public that follows that direction purely.
 

Latest News

Online or Offline racing?

  • 100% online racing

    Votes: 96 7.8%
  • 75% online 25% offline

    Votes: 130 10.5%
  • 50% online 50% offline

    Votes: 175 14.2%
  • 25% online 75% offline

    Votes: 348 28.2%
  • 100% offline racing

    Votes: 480 38.9%
  • Something else, explain in comment

    Votes: 5 0.4%
Back
Top