1. This site uses cookies. By continuing to use this site, you are agreeing to our use of cookies. Learn More.

racing simulators too fast

Discussion in 'Automobilista 2 - General Discussion' started by WhippyWhip, Nov 26, 2022.

  1. WhippyWhip

    WhippyWhip Active Member AMS2 Club Member

    Joined:
    Apr 4, 2020
    Messages:
    531
    Likes Received:
    178
    i tend to agree with this sentiment and have brought this up before (this is not my video), this is nothing agaisnt reiza as it's an industry wide problem
     
    • Agree Agree x 5
    • Disagree Disagree x 1
  2. Gevatter

    Gevatter The James May of Simracing AMS2 Club Member

    Joined:
    Mar 10, 2016
    Messages:
    1,483
    Likes Received:
    877
    I generally agree with the video that cars in sims are too fast compared to the real life cars, especially in the handy of Aliens. I saw that in ACC as well when the real life Nürburgring lap record was seemingly easily beaten by a few seconds in game. There's probably a bit of time that can be accounted for by the usual differences between real life and sim (no fear of death, no cost, no bodily excertion, endless training time, ...) but still, probably not a few seconds worth.

    Just to be thorough about the Corvette GTP, a very small detail is that the 1986 pole sitting Corvette GTP used a different engine compared to the AMS2 version. The detail is very small because according to this site it was only ~25HP less powerful than the V8 we have in AMS2 and it probably weighed less, so it shouldn't be much of a difference in laptimes. Road America also never changed its layout afaik, and the weather was good as well in 1986. Maybe the asphalt was way less grippy back then, who knows. The "bend" layout didn't even exist then it seems.

    Anyways, here's a link to the Corvette on the site Austin mentions. You can verify his data by clicking on one of the listed Chassis and click the "Full Archive Button" to see all recorded races and times.
    Chevrolet Corvette GTP - Racing Sports Cars

    One thing that is worth mentioning is that laptimes are a bed measure of "realism" in a sim. Simracers are on a very broad scale of ability. Alone in this forum I'm sure there are people who could match Austin's 1:48 but also some who might have trouble matching the cars real life 1:58, and many many in between. So the question is, where do you set the pace of the car if you want to go by lap time? Should you even do that? How do you compare yourself to the real life time if you're a developer?
     
    • Like Like x 1
    • Agree Agree x 1
    • Informative Informative x 1
  3. Marius H

    Marius H Internal Beta Tester Staff Member AMS2 Club Member

    Joined:
    Mar 24, 2016
    Messages:
    5,250
    Likes Received:
    2,010
    One of the reasons why sims are so fast is that we lack the fear of death. We never face consequences leading to injury or death. Only what we deal with is damage. I seen people go fullout in the qualifications because there's nothing that can stop them. 'Oh I slided in to the gravel. I will just go back to pits with a click'. Then during the race they either drive a little safer, or they can't drive at all, but that's another story. In RBR you had to really drive defensively without hurting your performance too much, if you hit something it was likely you were retired from the whole event. I think graveltraps, damge, and all that need a little more love, but on the other side I think it will frustrated the casual online crowd a lot and they will even ragequit more. It's a doublesided sword.
     
    • Agree Agree x 2
  4. WhippyWhip

    WhippyWhip Active Member AMS2 Club Member

    Joined:
    Apr 4, 2020
    Messages:
    531
    Likes Received:
    178
    any real life racing driver will tell you fear of death is not a thing for a professional, otherwise you'd get one psycho constantly 5 seconds faster than the rest
     
  5. bobbie424242

    bobbie424242 Well-Known Member AMS2 Club Member

    Joined:
    Jul 31, 2017
    Messages:
    1,011
    Likes Received:
    487
    Sure. But said professional still is not going to push as hard in a sim where nothing has consequences, crashed or damage cars a reset away, time is unlimited, cost like $0.
     
    • Agree Agree x 1
  6. Scar666

    Scar666 Zum Glück bin ich verrückt

    Joined:
    Jul 1, 2020
    Messages:
    1,232
    Likes Received:
    1,377
    This for me is the end to the complaints about cars being faster in sims than in real life...

    There are reasons why Fangio, Moss, Stewart and other greats from that era are so respected... They got closer to the limit than anyone dared...

    If you add in the hard but not impossible as Reiza continue to get closer to it, mechanical sympathy that drivers just had to have to reach the end of a race... Or simply avoid costly repairs in the pits over a race weekend... Drivers would routinely be driving well below their limit just to make sure there's no damage to any important components... Not every team had the budget to change an engine after every session...

    Every series used to be like Formula E where the ONLY time you saw the max performance of all the mechanicals was in qualifying when they had all night to do an engine swap... Of course the top end of motorsports is where you saw special 1 lap engines for qualy...

    The further you go back the more the fear factor and mechanical sympathy play a role in slowing down real world times...

    If you are still questioning this reality, just think how much more care you take heading through Monaco when you have damage on than when you have no damage on and are driving a track like Paul Ricard with large run off zones with no cut tracks to force you to be careful here and there...

    Plus as others have stated, even with all the bumps mapped perfectly thanks to laser scanning, there's no way of getting the grain in the surface of the road and how much grip there was with the track technologies of the past... Or for the safety work that goes on between sessions cleaning up the track... Or the red flags for dangerous conditions like oil on track... Or even how much dust was on the surface that year...

    So comparing times between sim vs real is a wasted exercise... It would be nicer if it was closer, but when simulating a car physics perfectly there's a lot of reasons as to why faster times can be achieved in the simulated world...
     
    Last edited: Nov 27, 2022
  7. Roar McRipHelmet

    Roar McRipHelmet Well-Known Member AMS2 Club Member

    Joined:
    Feb 25, 2020
    Messages:
    1,418
    Likes Received:
    1,139
    From the video:
    If a person was experiencing the behavior described above in real life and not a in a simulator, that's physics telling you to take a hint. Your body would tell you to stop, take a break, or at least take it a bit easier through the corners - either due to physical exhaustion or through fear alone. And I'd say that the same would happen even if you were a professional race driver from that era. It was the 1980s, and the practice of having professional race drivers do weight lifting as part of training for a race season is only a trend from more recent times.

    Sims are indeed faster than in real life, many of the reasons being what's been discussed in this thread so far. Even in an ideal world where a sim developer gets all the numbers right, that still doesn't account for non-quantifiable effects such as driver physique and fear that don't translate well into a simulation setting. Jumping across a sausage curb in a simulator risks underbody and suspension damage. Jumping across a sausage curb in real life risks spinal injury if you're unlucky.

    The takeaway discussion from this is not "why are sims faster than in real life", but rather "if sims get all the numbers right and are still too fast, should sims then add an artificial penalty factor in order to get lap times down to match real-life performance or should we just learn to accept the difference in ideal lap times?"

    I don't recall the scientific terms for this, but it's like the different approaches to tire models in sims: Having a tire model based on friction curves, vs. having a tire model based on physical tire properties. The simulator can either try to replicate each internal component as accurately as possible, at the cost of inaccuracy in the end result (i.e. lap times), or the simulator can try to replicate the end result, at the cost of inaccuracy of the internal components. It doesn't have to be either/or, it can and should be a compromise, but it will never be full accuracy in both domains.
     
    Last edited: Nov 27, 2022
    • Like Like x 2
    • Agree Agree x 2
  8. Scar666

    Scar666 Zum Glück bin ich verrückt

    Joined:
    Jul 1, 2020
    Messages:
    1,232
    Likes Received:
    1,377
    Well put...

    And far as the distilled question goes...

    No fudging for me... Put in the real numbers and I know the reasons why someone is faster than real life... They played a video game where the most deadly thing that can happen is to get bitten by a spider...
     
  9. CrimsonEminence

    CrimsonEminence Administrator Staff Member AMS2 Club Member

    Joined:
    Nov 4, 2018
    Messages:
    9,137
    Likes Received:
    8,205
    • Like Like x 4
  10. Roar McRipHelmet

    Roar McRipHelmet Well-Known Member AMS2 Club Member

    Joined:
    Feb 25, 2020
    Messages:
    1,418
    Likes Received:
    1,139
    One of the strongest negative motivators in racing is the fear of injury.
    One of the strongest negative motivators in gaming is fear of getting locked out of the game.
    I wonder how fast these Time Trial laps would be if:
    • 1 off-track or wall scrape, or restart/aborted lap: 1 hour locked out of Time Trial mode
    • 1 spin or damage-inducing crash: 1 day locked out of Time Trial mode
    I'm not suggesting that games should implement these policies, especially not as a money-making scheme where one can buy oneself out of the predicament, but it's a thought experiment that puts the situation in perspective.
     
    • Agree Agree x 1
    • Useful Useful x 1
  11. Marius H

    Marius H Internal Beta Tester Staff Member AMS2 Club Member

    Joined:
    Mar 24, 2016
    Messages:
    5,250
    Likes Received:
    2,010
    in AMS 1 you've towing being simulated. In short: the further you are from the pitstops, the longer you've to wait if you just press ESC > Pitstop. I think it's a great feature. So the player actually have to drive safe and not push too hard in P and Q to avoid these waitingtimes. Imagine if you crash during Q and the waitingtime it too long to do another lap or two. That would at a whole new strategy.
     
    • Like Like x 4
  12. Scar666

    Scar666 Zum Glück bin ich verrückt

    Joined:
    Jul 1, 2020
    Messages:
    1,232
    Likes Received:
    1,377
    That's another huge part of it...

    It's not just avoiding repairs from Mechanical Sympathy, the driver wants to be in the car for as long as possible... Damage it and you lose track time... They don't get to play with their favourite toy for as long as they wanted...

    It would be an interesting experiment to see how many sim racers are hardcore enough to leave their PC running all weekend just to do an engine change that went wrong and required another engine change...

    It would really make the public lobbies less of a crash fest...
     
  13. ramsay

    ramsay Active Member

    Joined:
    Jul 25, 2020
    Messages:
    116
    Likes Received:
    54
    For someone who criticises accuracy in sims he seems to not factor in a lot of variables which are almost impossible to gauge accurately especially from decades ago. Also TT is not the same as qualifying IRL, it's a 1 lap fun, but artificial game mode, no outlap, warming tyres/brakes, etc. and to compare the two is just bloody stupid.

    More importantly did they use all the aprons/runoffs/kerbs (if they existed then) and did they have the same grip as the track? What grip level did the track have and did it vary from corner to corner (which most tracks do)? What were the track conditions when the times were set? Did they push the car to the limit or not to prevent mechanical damage and if they did could they rebuild the cars for the race? How much has the track changed? For example, the kerbs, aprons & runoffs are completely different or do not even seem to exist in the race he linked. They do not ride them as I assume they would lose control, time or damage the car as they are raised, but you can in the modern version modelled in the sim. All these would make a considerable difference to lap times. Would that be 10 seconds? I've no idea, but he didn't even seem to take any of them into account.

    To me there are too many variables to compare like for like and anyway using lap times as a yardstick for accuracy is a bit daft to me. You could make a car like a rocket on the straights and slow in the corners to get similar lap times or vice versa and the conditions make a huge difference.

    I've come to the point in sims that the feel of the car is the fun part in a racing sim to me, and if it's in the ballpark and believable that's all I want as I'm certainly no expert or so self-absorbed to think I know better when I haven't even driven the actual car.

    There are too many narcissistic blowhards in simracing imho.
     
    • Agree Agree x 2
    • Like Like x 1
  14. ricxx

    ricxx Well-Known Member AMS2 Club Member

    Joined:
    Apr 30, 2022
    Messages:
    581
    Likes Received:
    330
    In Europe the chance of getting bitten by a spider is close to zero :D We have Renato's explanation now and even without the insight from a dev we must recognise that simulation games will never be 100% realistic because you can't code all the RL physics, so there'll always be some inaccuracies, physics will always be an approximation and the goal is to be as realistic as possible.
     
    • Agree Agree x 2
  15. Roar McRipHelmet

    Roar McRipHelmet Well-Known Member AMS2 Club Member

    Joined:
    Feb 25, 2020
    Messages:
    1,418
    Likes Received:
    1,139
    ...but the chance of getting pounced by a domestic cat is close to one, even if you don't personally own a cat :D

    Ask any VR gamer how they know...
     
    • Funny Funny x 2
    • Agree Agree x 1

Share This Page