Assetto Corsa Alfa Romeo 4C vs. real life with Aristotelis (dev) as a pilot.

Johnnypenso

Well known double poster
Premium
28,470
Canada
Windsor, Ontario, Canada
Johnnypenso
Real life vs. game video comparisons are nothing new, but a dev driving a real car on a real track and then comparing it to the same car in game is something fairly unique. The driver in this case is Aristotelis Vasilakos, who is in charge of car handling and physics as well as UI Programming and Design. This video was more made to compare physics and car behaviour and not laptimes and you can watch the video and judge for yourself how closely they match up. A side mounted, external camera is used in both cases to monitor steering inputs and there are external shots as well.

Two things strike me about this video. Both cars behave very similarly. On top of that, there's this constant debate about whether real cars are more or less difficult to drive than we find them to be in [insert favourite game here]. AC sort of has the reputation of being more user friendly and on the more forgiving side compared to the likes of iRacing or rFactor but it seems in the video that the real car is much more stable and requires somewhat less correction than the virtual car does. In this one particular case at least, real car seems even more forgiving than the supposedly "forgiving" physics of AC. Kind of tells me that AC is on the right track and sims that feel like you are driving on iCE or are extremely difficult to control when in oversteer, are maybe just being difficult for the sake of being difficult.

Note: Real life conditions, 15C and green track, were matched as closely as possible in game. The track is a free mod and not included in the base game. The original post from the AC Support Forum is in the spoiler at the bottom of the post.



Hello everybody. Sorry for the shameless spam :)

My good friend Maurizio Spinali who is an automotive journalist but also great fan of AC, has been recently inviting me to test drive some cars that are also included in Assetto Corsa.So we had this idea to do some comparison laps between real and virtual.
The following video is our first try and we know that it's not perfect, but we had fun doing it and if you like it, we intend to make it better in the future with voice-over comments and more.

The car had street tyres and it was a rather fresh day with around 15°C ambient temperature and a green track. In the video you can see two laps, one is very "dirty" just trying to figure out how the car reacts when pushed to slide around. The second lap is a another go with a bit more smooth driving, although the shot rear tyres started to act on corner entry. Same approach on the virtual driving just trying to find out if the car reacts in the same way if given similar inputs.

I'm sure plenty of you will think I'm not driving good enough and you'll be able to point out various errors, but the focus was not on laptime. On the contrary what we wanted to do, was document the first couple of laps from my stint so that I wouldn't had the time to get accustomized to the car handling or the circuit layout, but rather to understand how the car reacts on various situations close to the grip limit and if I could use the same techniques the simulated car forced me to use.

The 4C... is not perfect when pushed hard. Understeer on power, oversteer on turn in and a difficult to understand power delivery... but brakes are excellent and the lateral grip is impressive. When you decide to go at 8/10 then it's perfect and the looks... oh the looks! This is a proper Italian diva.
You might also notice that we are not changing gears up to the limiter. We decided that way as the gearbox had plenty of journalistic kms on its life and occasionally was acting up strangely if pushed right to the limiter. 400rpms lower and it would stay perfect and make us happy all the same.

Hope you like it and if you do, subscribe to the channel and we'll bring you more.
Link
 
Compare that first clip of him sliding and the clip at 2:16 right from the beginning you can see He isn't in control as much as he is IRL, the game just has less grip but the car's reactions and body roll seem more severe, it looks like there is less dampening from the shocks also. You could improve your lap times in game from higher tire pressures quite a bit as the virtual car looks very loose otherwise. Bottom line - they still need to work on the tyre model quite a bit before they get the cars right.
 
Compare that first clip of him sliding and the clip at 2:16 right from the beginning you can see He isn't in control as much as he is IRL, the game just has less grip but the car's reactions and body roll seem more severe, it looks like there is less dampening from the shocks also. You could improve your lap times in game from higher tire pressures quite a bit as the virtual car looks very loose otherwise. Bottom line - they still need to work on the tyre model quite a bit before they get the cars right.
I don't see any issues at 2:16.
 
I don't see any issues at 2:16.
not really an issue, but the grip levels aren't exactly up to my expectations from the "top" racing sim. I keep forgetting this is a WIP simulation but I still like to point it out.
 
not really an issue, but the grip levels aren't exactly up to my expectations from the "top" racing sim. I keep forgetting this is a WIP simulation but I still like to point it out.
You can tell the grip levels are incorrect by watching a YT video?
 
You can tell the grip levels are incorrect by watching a YT video?
if they are comparing the exact same technologies; track temperature is the same in game and tyre compound is the same, then yes, there is something a bit off from the game to the video. Could be anything, but I would say it is tyre grip and a little of suspension setup is different. It is more noticeable when he catches the car right after the drift. The best comparison is from 3:00 - 3:12 he has to over-correct the car much more to straighten out, you can see it is quite loose in the game. Nonetheless I am sure they Kunos be improving the game with every update, so my observations will be mute in a few months time.
 
Last edited:
if they are comparing the exact same technologies; track temperature is the same in game and tyre compound is the same, then yes, there is something a bit off from the game to the video. Could be anything, but I would say it is tyre grip and a little of suspension setup is different. It is more noticeable when he catches the car right after the drift. The best comparison is from 3:00 - 3:12 he has to over-correct the car much more to straighten out, you can see it is quite loose in the game. Nonetheless I am sure they Kunos be improving the game with every update, so my observations will be mute in a few months time.
Since we're microanalyzing a Youtube video, at 3:06 you can clearly see he's holding a larger steering angle entering the corner. When the car bites, he has to overcorrect. Looks like driver input error to me. I circled his left hand in each car:

upload_2016-2-22_9-53-44.png
 
No I do not believe it to be all driver input error, as you suggest Johnny. I have watched and re-watched this clip with your findings in mind already. Nice try though :sly:
 
No I do not believe it to be all driver input error, as you suggest Johnny. I have watched and re-watched this clip with your findings in mind already. Nice try though :sly:
Nice try. I already pointed out the exact moment of driver input error at the exact time that you said the "best comparison" was showing a physics issue causing overcorrection. It's as obvious as the nose on your face what happens and it's perfectly in line with what a simulated and accurate physics engine might do along with the real car. Try turning your front wheel another 75 degrees more than needed and have grip return to the front end and see what happens while you are pounding the throttle. Does that mean the physics are perfect? No. It just means that your argument that the physics are flawed based on your cursory analysis of a YT video is bollocks.
 
Nice try. I already pointed out the exact moment of driver input error at the exact time that you said the "best comparison" was showing a physics issue causing overcorrection. It's as obvious as the nose on your face what happens and it's perfectly in line with what a simulated and accurate physics engine might do along with the real car. Try turning your front wheel another 75 degrees more than needed and have grip return to the front end and see what happens while you are pounding the throttle. Does that mean the physics are perfect? No. It just means that your argument that the physics are flawed based on your cursory analysis of a YT video is bollocks.
No. I highlighted the whole area of time in my previous post, not a lifeless still image like you constantly think I am talking about, triple-take a look again at that moving video to comprehend the true meaning of my post. Hit that replay button and sit still, try not to fidget just Let it sink in. My statement might lack a little clarity or a more long, drawn-out dumbed-down elaborate explaining just for your unusually thick skull but seems as you have already "summed up my whole argument" and compart-mentalised said argument it is apparent all you care about is starting flame wars on this forums, so I am done here. The ship has sailed! Goodbye. I have said my peace, you can choose to think really out-of-your-comfort-zone harder to learn or keep on twisting my words like typical ungrateful manchild.
 
Last edited:
thanks for shortening my response and editing out the good bits, way to go Brqandon! :yuck:

I had no comment on the other bits, claiming a video game is unrealistic because of a slight difference in a YouTube video really isn't worth commenting on since that would apply to every racing game ever created. But Mr. Penso is possible the most level-headed, wise, and civilized member roaming the AC forum, and he's a Canuck so he's friendly by default (until he's confronted by nonsense). Anybody who knows him would verify that your adjectives have no association with him. Also, hurling insults and name-calling is rarely the best way to win an argument, but it is a great way to get people to stop taking your position seriously.
 
Stop with the name calling and personal insults, the AUP is quite clear on this front and anyone not following the AUP in this thread from now on will be taking a little holiday from GT Planet.

Have some respect for each other even wen you are in disagreement.
 
It seems to me like the game car has less grip than the real car, so Aris is pushing the game car extra hard to make the camera match up. Which explains the steering corrections and in general more raggedy driving in game. It's not necessarily a physics issue mind, could just be real life track conditions being better, or the game track not having proper grip level (being a mod track and all).

I would like to see how AC fares with Gran Turismo's GPS app thingy. Take real world throttle/brake/steering inputs and run it with the game engine. See how close it matches up 👍
 
No. I highlighted the whole area of time in my previous post, not a lifeless still image like you constantly think I am talking about, triple-take a look again at that moving video to comprehend the true meaning of my post. Hit that replay button and sit still, try not to fidget just Let it sink in. My statement might lack a little clarity or a more long, drawn-out dumbed-down elaborate explaining just for your unusually thick skull but seems as you have already "summed up my whole argument" and compart-mentalised said argument it is apparent all you care about is starting flame wars on this forums, so I am done here. The ship has sailed! Goodbye. I have said my peace, you can choose to think really out-of-your-comfort-zone harder to learn or keep on twisting my words like typical ungrateful manchild.

Wow, that escalated quickly... :lol:

I had no comment on the other bits, claiming a video game is unrealistic because of a slight difference in a YouTube video really isn't worth commenting on since that would apply to every racing game ever created. But Mr. Penso is possible the most level-headed, wise, and civilized member roaming the AC forum, and he's a Canuck so he's friendly by default (until he's confronted by nonsense). Anybody who knows him would verify that your adjectives have no association with him. Also, hurling insults and name-calling is rarely the best way to win an argument, but it is a great way to get people to stop taking your position seriously.

I agree with every bit of this. :cheers:
 
How typical to walk that thin line between debate and insult and then exit stage left to attempt avoidance of any kind of response. Bad form and unacceptable, however desperate you are you have the last word.

how typical you pick on the new guy. way to go about modding. typical tyranny. despite the fact you totally miss johnnys insults. just typical bs
 
how typical you pick on the new guy. way to go about modding. typical tyranny. despite the fact you totally miss johnnys insults. just typical bs
He insulted your argument, not you. You on the other hand directly insulted him, you also seem to have forgotten that the AUP requires you to use good basic grammar, that includes using a capital letter at the start of each sentence.

This has nothing to do with you being new, and everything to do with you not following the AUP. I would strongly suggest that you do so from this point on, and this conversation is done.

Continue in this vein and the warnings will become formal.
 
Just amazes me technologically speaking how high the bar has become when evaluating sim racing games. 10 years ago we were lucky if we got car game that would run at 30 fps and now we're arguing about steering wheel position of a car when compared to its real life counterpart.

10 years from now - "Just before the pits on Vallelunga there's a tree IRL that has exactly 26 branches but in the game it has 25. How can they call this a sim with such obvious mistakes? Waste of money, better off buying pCars 9"
 
10 years from now - "Just before the pits on Vallelunga there's a tree IRL that has exactly 26 branches but in the game it has 25. How can they call this a sim with such obvious mistakes? Waste of money, better off buying pCars 9"

Im pretty sure I saw that argument last week on the official forum, except it was about a shrubbery and at which angle the leaves grew from the branch. :lol:
 
Back