Forza 4 vs GT5 physics (read the first post before contributing)

[Scaff]

That's not true dude and if you played GT5P you would know this. The manual gave you the tires to correspond to stock tires on each specific vehicle @ Zero agree with you 100%

That was GT5 not GT5.

GT5 has had two physics versions and a number of updates that have affected tyres, knowing anything about what PD intent tyres -stock or not - to represent is just about impossible. The only things we do know is that the tyres have no real world match and that they seem to act as grip multipliers as grades increase (and the later supports the former).

As for lap times being a good indicator or sim accuracy and tyre modelling accuracy in particular, I am yet to see a single bit of evidence that would back that up at all. Simply because its possible to match lap times in a sim does not mean the sim is accurate, to claim other wise when so many variable are involved is a logical failure of huge proportions.

It assumes that every part of the sim engine is perfect for a start (which alone is patently ridiculous - unless you wish to claim GT5 is perfect and then I would be quite interested in how both GT5 pre and post 2.0 are perfect), not to mention the variables involved that will cause differences in real world lap comparisons alone:

• Ambient temperature
• Humidity
• Wind direction and speed
• Has the track been resurfaced
• What else raced on the track recently
• Time of day

Take a track the size of the 'ring and you can have totally different weather on different parts of the track, yet your quite happy to say that if you can match the lap times with the same car then the sim is accurate?

What if the sim has a temp of 30C and a humidity of 80% and the real track had a temp of 12C and 0% humidity? What if the cars running a high boost turbo with those differing factors? Still think the comparison would be valid?

I don't.

Lets throw in another (as it relates to tyres so fits quite well), what if the tyre model in the sim has too much long-g and too little lat-g, you can accelerate and brake more effectively than the car should yet not corner as well as it should. However what if these two balanced each other out over the duration of a specific lap? How accurate is you sim now?

Its not.

All of which is also ignoring the 🤬 great elephant in the room, which is people engineering lap times to match. Would you be so bold as to claim no one has ever done it? That they haven't recorded lap after lap until they got the result they wanted and then stopped?


Too much can and does vary around an entire lap, and the long the track the more it comes into play.

 

[SlipZtrEm]

Not this again.

Lap time comparisons between real life and any video game is nearly useless, as has been covered pretty exhaustively in this thread. Is it handy, and a cool comparison? Absolutely. But it doesn't somehow make a game more realistic just because you can match some lap times. There are far too many variables, and unknown quantities in either game's physics engines, to say that one is more realistic than the other based on lap times.

This is, of course, conveniently skimming over two other far more important issues that people who seem to prefer GT are all too quick to dismiss: the whole "Mini and Vette showcase the same lateral-g figure" issue, and which mode are we saying is the realistic one in GT again, offline or online?

(Edit) Woops, the app didn't show Scaff's post, in fact, the last one I had seen before commenting was King's!

 

[dylansan]

There are different aspects of the physics model which affect the performance of a car in different ways. As such, lap time comparisons are the worst way to analyze the physics because you're combining all the individual aspects into one result. There could be some problems that make the cars faster than real life, and some that make them slower which, when combined, may result in close to accurate lap times for all the wrong reasons.

For example, if cars had excessive understeer but also excessive acceleration, a car's average speed around a track may be reasonable even though the dynamics of the car would be completely wrong.

The only way to effectively analyze the physics is to focus on individual aspects, the reactions of the cars to specific situations like understeer and oversteer and weight transfer, etc.

 

[Spagetti69]

Only on. SPIKE.

 

[Connoissuer]

I say that both games are equal. What makes me say it is the ABS system. GT5 can't get its ABS system right for ****. Forza has it down properly.

In GT5 the ABS doesn't even lock the ****ing wheels for a ****ing second. Its a damn If there were brakes such as that in real life, I think the closest system would be on a F1 car.

If GT5 could actually get its ABS system right, I think I'd have a chance at the GT Academy for once, since I don't use ABS.

 

[LMSCorvetteGT2]

Uh...good points, but since there are two mods that post here very frequently, you shouldn't skip over the censor filter anywhere, but here you should make sure to watch out even more. Respect the AUP 👍

 

[Ealirendur]

Yes. But in a game it doesn't start at all - there is no fear when there is no damage, no injury and two button presses between anything and the beginning again.

There's no doubt that the visceral fear is entirely missing, but I don't think it's always quite that clear-cut - running in an online series with Heavy Damage on is a bit different for this as far as the sort of error margin you would allow without it; there is a certain element of actual fear, not so much of ruining your own race, perhaps, but definitely (for me, at least) someone else's. Even a minor rear tap received under braking cost me a lot of time last race, as heavy braking became quite entertaining.

You need no error margin at all when time trialling, clearly, as you say. Even when (sim) racing (with others, or perhaps even in a solo offline event you have some significant degree of personal time investment), I'd have a small margin, or try to prevent myself from further exploring the limits too much.

To the point - the safety margin left will vary by context and driver, both in real life and sim. It's perfectly valid to assume that for situations involving equal knowledge and confidence, the margins in real life will nearly always be larger for a given driver, probably significantly so; that doesn't mean the lesser one is zero.

I'd be curious to find out what sort of margins/constraints experienced real and sim racers set for themselves - I include myself in neither category, BTW.

As an aside, how much time is gained by the superior (posterior?) feedback sensors in real racing?

ObTopic: I have no position on Forza 4 either way, beyond some car lust (Dino, RX7 GSL...) - if somebody would like to donate the required hardware (including a CSR-E), I'd be happy to opine

 

[Zer0]

Why all the lap comparisson detractors resume the supportive argument with a: "lap matching makes a game more realistic because you can match some lap times" when it's only a proof of "some" realism and by itself without many other considerations first don't make one game more realistic than other?. I doubt that anyone is defending this with the first simplistic argument.

But surely the tyres PD fit to cars as standard are the tyres they believe are closest to the real world standard? How can you know that the simulation is perfect and the tyres are too grippy when it could be that the simulation isn't perfect but the tyres are just right?

Like I said, there are too many anomalies to draw anything at all from lap times.

First PD don't put the most realistic tyres by default, always have been with more grip since GT1. Same thing as why they put on all the driving aids by default in all cars. Just by accesibility. It's not a secret if you have tried before to play realistically a GT game.

In GT5P PD did put in the manual a list of recommended more realistic tyres for every car(always below the stock). In early GTs they have an specially dedicated low grip simulation tyres(control tyres).

As I said PD always oficially pick tyres with less grip for any real life test for a better handling accuracy, that also reflects more accurate lap times. That has been confirmed too by experienced drivers and there is plenty of practical track examples with opinions in youtube.

 

[Samus]

It's not proof of any realism though. I could put on Project Gotham Racing 4 and match a laptime around the Nurburgring, it proves nothing of the games realism. We know the game is arcade based but if you brought an alien down to play it and he did a lap time to match reality in PGR 4, would you tell him that means the game is realistic? No.

 

[JDMKING13]

There's no doubt that the visceral fear is entirely missing, but I don't think it's always quite that clear-cut - running in an online series with Heavy Damage on is a bit different for this as far as the sort of error margin you would allow without it; there is a certain element of actual fear, not so much of ruining your own race, perhaps, but definitely (for me, at least) someone else's. Even a minor rear tap received under braking cost me a lot of time last race, as heavy braking became quite entertaining.

You need no error margin at all when time trialling, clearly, as you say. Even when (sim) racing (with others, or perhaps even in a solo offline event you have some significant degree of personal time investment), I'd have a small margin, or try to prevent myself from further exploring the limits too much.

To the point - the safety margin left will vary by context and driver, both in real life and sim. It's perfectly valid to assume that for situations involving equal knowledge and confidence, the margins in real life will nearly always be larger for a given driver, probably significantly so; that doesn't mean the lesser one is zero.

I'd be curious to find out what sort of margins/constraints experienced real and sim racers set for themselves - I include myself in neither category, BTW.

As an aside, how much time is gained by the superior (posterior?) feedback sensors in real racing?

ObTopic: I have no position on Forza 4 either way, beyond some car lust (Dino, RX7 GSL...) - if somebody would like to donate the required hardware (including a CSR-E), I'd be happy to opine

I agree with you 100% In my league when I play with my family damage is something to be scared of. It can totally ruin your car, and your driver moral etc. Just as you explain during races you don't want to be the one who crashes trust me. It will cost you big time in my league. I can faithfully say when driving during league play I am much slower, and careful then just joining in an online room. Driving my RX-8 is a real hassle for me in league play, I have a lot of money saved up and I am not trying to waste it on me doing some reckless driving I rather race clean and safe then try to finish the race 1st because majority of the times you just crash.

Why all the lap comparisson detractors resume the supportive argument with a: "lap matching makes a game more realistic because you can match some lap times" when it's only a proof of "some" realism and by itself without many other considerations first don't make one game more realistic than other?. I doubt that anyone is defending this with the first simplistic argument.


First PD don't put the most realistic tyres by default, always have been with more grip since GT1. Same thing as why they put on all the driving aids by default in all cars. Just by accesibility. It's not a secret if you have tried before to play realistically a GT game.

In GT5P PD did put in the manual a list of recommended more realistic tyres for every car(always below the stock). In early GTs they have an specially dedicated low grip simulation tyres(control tyres).

As I said PD always oficially pick tyres with less grip for any real life test for a better handling accuracy, that also reflects more accurate lap times. That has been confirmed too by experienced drivers and there is plenty of practical track examples with opinions in youtube.

I agree with you, and again The only reason I watch video's, read tech articles etc is to try to get the closest resemblance to my gaming style. With GT5 the ability to degrade tires is huge. I remember hosting a race with Integra Type R DC2 stock form and there was a debate to either Use CS or CM. I wanted the CM however the room wanted CS. I told them we would do two races CS the CM. In the end everybody felt that the CM mocked the stock characteristics of the cars tire the best. It was a great to see how the room responded to the CM. To take it even further even though we all had our cars in stock form none of the cars were the same, because each motor had different HP outputs. This can only happen in GT5 which Is something else I like. Depending on when you got the car, mileage on the car, maintenance on the car you will always get different HP output. This is big in my league when you get people with the same car. Its always cool to see which person car is better because of the many factors that effect HP output.

 

[jujunosuke]

Hi everyone!
Sadly, i do not own any of those game, so i cannot tell which one has better physics.
However, i am studying car physics and tire models since 2 years now, and i am about to release a beta of my car simulation very soon.

IMO, I think that people does not get the good approach for testing which game is better.

It require more effort for testing but a probably a better approach would be to pick up 3 similar cars on both game and test them on a similar condition and similar track.

take a basic FWD car, RWD car and 4WD car and do the same test for each of them on each game.
Test them without any assistance : no ABS and no Traction Control.

Here are some example :
Test Understeer/oversteer
Brake around 60m/100 kmh with wheel locked.
Try to make a burnout.
Try to make donuts.
Handbrake turns.
etc...

This is of course the longest approach, but, as a developer who is doing my own car physics, this is how i would test both game physics.

Btw, the laps time does not mean anything. You can not compare car physics with a lap time.

I hope that my input can help.
Sorry about my non native English.

bye !

 

[Goshin2568]

Hi everyone!
Sadly, i do not own any of those game, so i cannot tell which one has better physics.
However, i am studying car physics and tire models since 2 years now, and i am about to release a beta of my car simulation very soon.

IMO, I think that people does not get the good approach for testing which game is better.

It require more effort for testing but a probably a better approach would be to pick up 3 similar cars on both game and test them on a similar condition and similar track.

take a basic FWD car, RWD car and 4WD car and do the same test for each of them on each game.
Test them without any assistance : no ABS and no Traction Control.

Here are some example :
Test Understeer/oversteer
Brake around 60m/100 kmh with wheel locked.
Try to make a burnout.
Try to make donuts.
Handbrake turns.
etc...

This is of course the longest approach, but, as a developer who is doing my own car physics, this is how i would test both game physics.

Btw, the laps time does not mean anything. You can not compare car physics with a lap time.

I hope that my input can help.
Sorry about my non native English.

bye !

The problem would be that you would have to do those tests in real life in order to judge which sim is closer to reality. And the chances of someone being able to do that are slim.

 

[thelvynau]

Hi everyone!
Sadly, i do not own any of those game, so i cannot tell which one has better physics.
However, i am studying car physics and tire models since 2 years now, and i am about to release a beta of my car simulation very soon.

IMO, I think that people does not get the good approach for testing which game is better.

It require more effort for testing but a probably a better approach would be to pick up 3 similar cars on both game and test them on a similar condition and similar track.

take a basic FWD car, RWD car and 4WD car and do the same test for each of them on each game.
Test them without any assistance : no ABS and no Traction Control.

Here are some example :
Test Understeer/oversteer
Brake around 60m/100 kmh with wheel locked.
Try to make a burnout.
Try to make donuts.
Handbrake turns.
etc...

This is of course the longest approach, but, as a developer who is doing my own car physics, this is how i would test both game physics.

Btw, the laps time does not mean anything. You can not compare car physics with a lap time.

I hope that my input can help.
Sorry about my non native English.

bye !

Beta as in playable? Or a simulation?

 

[crazy_cj]

maybe I need to try gt5 with a wheel.

but can someone please explain why you can accelerate hard with a ferrari 458 from the get go and the car remains straight without losing any traction.

forza 4 replicates what a car would do in real life....and thats lose traction and side step to the side.

its been 10 years and pd still havent got it right. and I ask......why?

is it a flaw in the physics engine?.

its the small things man.

 

[NOSWaster]

Let me guess. You're using racing soft tires. Use the stock tires of the car and you'll find that it will easily side step.

 

[crazy_cj]

Let me guess. You're using racing soft tires. Use the stock tires of the car and you'll find that it will easily side step.

Nope. using sports hard. stock tires.

the cars do not side step. at all.

 

[Barra333]

Maybe turn traction control off? It is set to 5 by default for all cars
EDIT: Irrelevant now that the original question was clarified, and the discussion became GT5 v Forza & threads merged

 

[NOSWaster]

Nope. using sports hard. stock tires.

the cars do not side step. at all.

Then I'd advise turning the traction control off, as was already suggested. The cars DO side step. It's quite easy to over-drive them.

 

[crazy_cj]

no guys. I mean when your stationary. from a halt...not around corners.

 

[byegad]

And the two suggestions re traction control still hold, traction control controls traction. That is wheel spin under acceleration.

 

[Nismo34]

I get what the OP Means, how many of you play Forza 4? In your everyday car, if you're at a stand still and rev your car, it tends to roll due to the spin of the engine. This is replicated in Forza 4 and because of this, cars have a tendency to veer right immediately after take off. In GT5, the car does kick out, but that's after burning out for a few seconds. On Forza and RL, this is immediate under heavy acceleration.

Think of it as a FWD's torque steer. The engine spinning one direction pushes the left side of the car into the ground giving it more grip which in turn pushes the car to the right. Same logic with skate boards.... This doesnt apply to GT5 and its only the spinning of the tyres (Eventually at different rates) which cause the back end to step out..

 

[1241Penguin]

Do you guys not know what he's referring to?



(Sorry I couldn't find a better video, but note the "side-step" from the car.)

This has nothing to do with Racing Soft tires or traction control. It's just another problem in PD's so called driving "simulator."

 

[dbarrade]

Please explain why a car actually side steps, then try again somewhere that isn't perfectly flat with perfectly uniform tyre grip and perfectly central weight balance. Think about what GT5 is simulating, now explain why Forza does side step in that situation.

 

[crazy_cj]

Do you guys not know what he's referring to?



(Sorry I couldn't find a better video, but note the "side-step" from the car.)

This has nothing to do with Racing Soft tires or traction control. It's just another problem in PD's so called driving "simulator."

thanks bro. spot on.

 

[Custer85]

Porsche 911 GT2 Sledgehammer, no side steping
http://www.youtube.com/watch?v=r5xMyVaqtkU

Corvette ZR1 vs Porsche 911 GT2, no side steping
http://www.youtube.com/watch?v=vD9AW-EB3vo

4 Cars, Ferrari 599GTB, Nissan GTR, Porsche GT2 and Corvette ZR1, again no side steping
http://www.youtube.com/watch?v=OZlGMX8G3B4

Side steping does rarely occur in real life and most of the time its due to the surface conditions. Many roads tend to lean to one side and that causes a car to sidestep. Spinning the tires while accelerating might cause a car to sidestep, but mostly because the road is either banked or the front wheels not facing foreward perfectly.

 

[Nismo34]

Porsche 911 GT2 Sledgehammer, no side steping
http://www.youtube.com/watch?v=r5xMyVaqtkU

Corvette ZR1 vs Porsche 911 GT2, no side steping
http://www.youtube.com/watch?v=vD9AW-EB3vo

4 Cars, Ferrari 599GTB, Nissan GTR, Porsche GT2 and Corvette ZR1, again no side steping
http://www.youtube.com/watch?v=OZlGMX8G3B4

Side steping does rarely occur in real life and most of the time its due to the surface conditions. Many roads tend to lean to one side and that causes a car to sidestep.

1st, there is side stepping, it just isnt as noticeable, but it is there...

2nd one.. Didnt get a good camera angle...

3rd one...

 

[1241Penguin]

Please explain why a car actually side steps, then try again somewhere that isn't perfectly flat with perfectly uniform tyre grip and perfectly central weight balance. Think about what GT5 is simulating, now explain why Forza does side step in that situation.

Uneven torque distribution? Something GT5 can't simulate? Like FWD torque steer?

Porsche 911 GT2 Sledgehammer, no side steping
http://www.youtube.com/watch?v=r5xMyVaqtkU

Corvette ZR1 vs Porsche 911 GT2, no side steping
http://www.youtube.com/watch?v=vD9AW-EB3vo

4 Cars, Ferrari 599GTB, Nissan GTR, Porsche GT2 and Corvette ZR1, again no side steping
http://www.youtube.com/watch?v=OZlGMX8G3B4

Side steping does rarely occur in real life and most of the time its due to the surface conditions. Many roads tend to lean to one side and that causes a car to sidestep. Spinning the tires while accelerating might cause a car to sidestep, but mostly because the road is either banked or the front wheels not facing foreward perfectly.

You've just said it. It does occur in real life, so why doesn't it occur in GT5?

 

[Nismo34]

Note the ammount of steering angle applied on take off..




Again, steering not as much as the vette, but steering none-the-less

 

[Custer85]

You've just said it. It does occur in real life, so why doesn't it occur in GT5?

It does if you try to accelerate on a banked corner like on the oval tracks and on certain corners of the Nordschleife (not the Karussell) for example.

Then again the video i allready posted:

http://www.youtube.com/watch?v=vD9AW-EB3vo

The Porsche does not sidestep and the steering wheel is straight at take off. I think you can provoke sidestepping by spinning the wheels like in your Video with the Corvette. In the video you can also see that the wteering wheel is not straight at the start. The first thing he does is counter steer to the side you whould expect, the opposite site of the steering wheel angle.

 

[Spagetti69]

Not mine but here goes.

The most common mechanical cause of torque steering is uneven driveshaft lengths, and the uneven twisting of the shafts under acceleration. With this uneven length, engine power is transferred to the wheel with the shorter driveshaft a fraction of a second before it reaches the opposite wheel. This first wheel gains traction and pulls the vehicle toward that direction before the other wheel can gain enough traction to compensate. As acceleration is increased, traction and lag in the front wheels are also increased exponentially. Hard acceleration, therefore, leads to high torque steering.