How realistic are Sim cars? This is an interesting question most people like to ask. There are tones of discussion on this subject online. Most of them turned out to be helpless. This question will live on as long as Sim products live. To answer this question, of course, objective studies need to be done by comparing to real cars and on real tracks in: • Grip state; and • Non-grip state (sliding state). 1) Grip state: a car dynamic behavior in a grip state has been extensively studied. The physics of a car in grip state is pretty known defined, and most models applied in simulations are very successful in the past years. Two online published results on rF1 with modified F3 cars showed rFactor indeed are very accurate, could be up to ~95% in accuracy. I could take this for grant that latest sims such rF2, AC, iRacing maybe even better. So we may not need to address grip state on this subject.2) 2) Non-grip state (sliding state): dynamic behavior is not that well studied and when and how a car starts to slide is pretty much “unknown”, as “black art”: So, judging if a sim is realistic or not is mainly based on comparing on real life on non-grip state (driving at and beyond the limit) sees more meaningful. Of course, we have no data be compared for non-grip state. I came a quasi-objective way to do this, by referencing Chris Harri techniques one real sliding. Chris Harris has been sliding many of cars on his videos. He taught us how exactly he did step-by-step. Here is his video shown how to step-by-step slide a BMW M235i: AC has the same BMW car. For rF2, one can expect to slide Nissan Z370 in the same way. One shall note that you have to turn off all assistance OFF as he did on real cars and with NO tuning (setup) --- shall use factory default! Assetto Corsa has wide range of selections road cars. The following cars in Assetto Corsa were found in Chris Harris videos (he slide all in exactly same way he taught us): Ferrari La Ferrari: https://www.youtube.com/watch?v=GaAQW8lVaRM Ferrari F40: https://www.youtube.com/watch?v=3MDTcXGsjuo McLaren MP4-12C https://www.youtube.com/watch?v=5mKB-8WUB5k Mercedes SLS AMG https://www.youtube.com/watch?v=UkjqQ1rXpRs Please do not reply “I can slide those cars”, this is not an answer! If both AC and rF2 implement physics on sliding state is correct, then one shall be able to slide those sim cars just as he did a way on real cars.--- applying max throttle on rear tires with reversed max slip angle of front tires. If you can do it, please provide video to prove if you can. Then I believe the physics is correct. (please no tuning -- setup. use Factory default).
read the post carefully and watch the videos first if your IQ is high enough. your first video is not the way Chris Harris sliding. the slip angle is wrong direction! rest videos you posted are not worth...
the listed cars Chris Harris did are all listed in AC. For rF2, you can try any rear-wheel-drive car. If rF2 physics correct, one shall be able to do so, because he did consistently on all real cars.
Except Chris Harris is on street tires rather than race tires, so the technique may not have similar timing. And I'm not sure if the 370Z in rF2 has street spring/shock rates available. Drifting consistently is not so easy as it looks for the unpracticed, as Randy Pobst showed in a video a couple months ago. Edit: there's a great rF2 video of someone (gui?) drifting out of Sao Paulo pit lane with the Cobra. Edit: Paul Loatman!
Please do not reply “I can slide those cars”, this is not an answer! Okay, whatever you say..... This is an interesting question most people like to ask. Funny, never asked it in my lifetime I KNOW what I feel, I KNOW what I like. Stop worrying or giving credence what others perceive and have fun !@!
The Chris Harris teaching on his BMW video is showing how to get into a sliding state, which is a deterministic state. In other words, it is can be applied to others as long as you master his techniques. He did just like that for all other cars. See at 6:1, 6:30, 10:27, 11:56, 12:50, 13:30, 13:50, and 14:15 in his Ferrari La Ferrari video: https://www.youtube.com/watch?v=GaAQW8lVaRM see at 0:40, 5:13, 8:30, 8:47, 9:04, and 9:20 in his Mercedes SLS AMG video: https://www.youtube.com/watch?v=UkjqQ1rXpRs This is only "quasi-objective" way I can think of, using his techniques to slide the sim cars. Without this reference, then we turn out to be subjective again. there are only two outcomes here: 1) you just cannot master his techniques; 2) if so, but you still could not slide the car as he did, then physics is not correct. Otherwise, it is correct.
I asked same question a year ago on my youtube channel. Three guys replied and they tried on AC cars with Chris Harris' way: His reply: "Hey! I had roughly the same idea making this video http://youtu.be/wiDYj67ZvQI a couple of weeks ago. It shows Pagani Huayra and LaFerrari drifting including my german live-commentary. I can get the F40 and the SLS as well to drift fairly nice and also the MP4 is possible with a little bit of practice. I have driven with street tyres on a green track. At the end of the video you can find a whole lap witch both cars without my weird german voice xD If I had the time I would try to replicate every scene of the Chris Harris drfting video in AC, just for the sake of looks and comparison. Interesting: With the LaFerrari Chris says, that you can stay a gear higher as you think and I have tried that, too. Even the slowest ~50 km/h corner on Vallelunga is driftable in 3rd gear. Torque ftw " Other guy did too: The last guy replied: https://www.instagram.com/p/xuPJ0_vr9U/?modal=true How do you guys think the physics of AC cars? (PS: Chris Harris videos: See at 6:1, 6:30, 10:27, 11:56, 12:50, 13:30, 13:50, and 14:15 in his Ferrari La Ferrari video: https://www.youtube.com/watch?v=GaAQW8lVaRM see at 0:40, 5:13, 8:30, 8:47, 9:04, and 9:20 in his Mercedes SLS AMG video: https://www.youtube.com/watch?v=UkjqQ1rXpRs
"how to evalue realism of simulated vehicle" would be interesting subject to think/discuss of. Even test driving the real and the simulated car might not be good because driving experiences of those differ even if simulated and real car behaved in exactly same way. Maybe videos and telemetry data would be the best way then?
That's the part that matters and where almost every sim falls apart. In racing you push a car. You are constantly approaching and going over (regardless if it's small amounts, not-so-small amounts, or very large amounts) the limits of grip. Braking, turn-in. mid-corner, exit, corner after corner. This is where most sims fall apart. The way cars behave/act/react as you start to push and play with limits. Specifically in-terms of oversteer, the only sims that don't fall apart and don't go outside of their "sweet spot" are Netkar Pro, Live For Speed, and Driver's Republic Alpha (Assetto Corsa is part way there as the lower speed, less edgy cars, are quite good). They aren't perfectly accurate, of course not, no sim is, but the vehicle dynamics and kinematics go into and out of those states extremely "smoothly"; the "maths" of the physics go into and out of those states extremely "smoothly" regardless if it's a slow slide, a faster slide, a sudden snap-slide, large slip angle, small ones, high speed, low speed, a half-spin, a full spin, an F1 Car, an F2000 car. It doesn't feel like a different physics engine takes over, or something in the "maths" breaks when those situations occur in those 3 games (and partly in Assetto Corsa). Judging just from things like overall grip and corner speeds, well, that hardly means anything. You can technically have 100 different sims all nail down corner speeds and grip while each's physics drive completely differently. Some charts showing simplistic things like brake points, down changes, corner speeds, etc. don't mean much in my opinion. It's how the vehicle, the physics engine behaves/responds/acts/reacts that is important to nailing down vehicle dynamics and kinematics. That video looks wrong compared to drifting in real life, LFS, or even AC. It does all the typical weird ISI behavior. Massive issues with what I'm guessing is called lateral momentum and how the rear rotates relative to the original vehicle's direction of travel. That has been evident in any ISI physics engine from at-least F1 2002 to present day (RFactor 2), it's a far cry from Live For Speed, Assetto Corsa, Driver's Republic behavior, let alone real-life. The very first time I saw that NSX preview vid - in massive excitement if I may add - I could tell there were no revolutionary physics changes under the core but rather just another car added to the lineup (and a fun car at that, don't get me wrong).
This might be the dumbest physics talk I have seen on these forums. It's like reading a book review made by illiterate mouth-breather.
No, no intention to that, neither want/wish to do so. Honestly, I do not believe AC is better than rF2 on this regard. I did try the BMW 235i of AC. With "full throttle" as Chris taught to get into oversteer, I seem start to slide but no way I can manage stage 2 and 3 of which he was teaching us on his real BWM. Similarly, I tired some rF2 cars, stage 2 and stage 3 fails on me. The guys' videos on AC cars look like they managed both stages but I am not sure how much throttle they applied. I have a doubt on this. If applied that much throttle ("full") on both AC and rF2 cars, I got spin off right after encount the steering. I give up. Chris Harris made so easy on this video. there must be some truths. Hey, there are tons of hard core racers here. Some one got to try and let us know. We all need one and only one who is going to show us he can do that as Chris Harris did on real cars.
I personally think that the inherent lag in our sims are what make it tricky to drift well. I remember seeing a shot of kimi coming back on track, he cuts a big slide in his f1 car, but absolutely immediately you see his hands whip to correct, and as soon as it comes back out of it, his hands whip back to stop the overcorrection. When I've looked at slides that made me go "wtf", I always see my actual reactions are just ever so slightly behind the cars movements. The computer has to process the state, send it to the gpu and out to the monitor, which them has to display it there, and only then when I see it amd process it, can I respond, which has to go back in to the computer and be processed, etc. As far as I can see IRL, kimi would have had his whole body sensing the slide with just his uber quick reaction times as a delay. Not saying this is the answer, it's just something I always notice when something goes wrong for me.
+1 Golanv, I would say this isn't the most scientific look at the physics, that's all...it is a public forum though, so say what you want eh...just don't mind when people want to judge you as a "mouth breather"
