Kazunori Yamauchi: Driving Simulators Shouldn't be Difficult

Discussion in 'Gran Turismo Sport' started by GTPNewsWire, Jun 27, 2017.

  1. mrPetros

    mrPetros

    Messages:
    576
    and I used absolutely zero magic tricks in doing that :)
     
  2. Tassie_tiger

    Tassie_tiger

    Messages:
    1,641
    Location:
    Australia
    :rolleyes:

    Hyperbole much?
     
  3. XxHighWayStarXx

    XxHighWayStarXx

    Messages:
    59
    Location:
    Australia
    A ridiculous analogy. I wouldn't even rate it a one. For starters there is no "perfect" recipe for anything. We all have different tastes & preferences, so no one gets to dictate what is perfect or ultimate to anyone else. So what's your point ?
    How you even thought this had any relevance is pretty crazy indeed.
     
  4. 05XR8

    05XR8

    Messages:
    28,303
    Location:
    Australia
    It's true. The car doesn't look to be unsettled at any point. Bathurst is challenging. The driver only looks to be held up by AI(?). Unless that's a Player. Just plant your foot(or squeeze trigger), even over crests. The steering looks too smooth everywhere. You'll see corrections all the time, in real life and PC2. I think even Forza 6 simulated Bathursts imperfections in the road, that unsettle the car.

    Roll your eyes if you want. There just doesn't appear to be any communication with the road surface.
     
  5. Johnnypenso

    Johnnypenso Premium

    Messages:
    28,424
    Location:
    Canada
    I believe his "point" was comic relief and you missed it by a mile.:lol:
     
    Imari and Lukanyon like this.
  6. XxHighWayStarXx

    XxHighWayStarXx

    Messages:
    59
    Location:
    Australia
    I totally understand that Johnny.
    However, it takes test drivers to provide feedback for a car's handling and performance yes ?
    So my point is it is THEIR feedback which helps compile the data needed to tweak the car if and as needed. They then test drive the car again. On it goes. Rinse and repeat.
    Now when any car's stock handling is programmed & calibrated for a simulator, I have to rely on the fact that THAT handling simulation is supposed to be accurate for that car. But it's based on another driver's inputs, interactions & perceptions.
    Unless I have driven the exact same car, how do I know the simulation is accurate ?
    That's why I say no racing simulator can ever simulate reality. Because whose reality was experienced in test driving each of the cars? Certainly not mine, and it wouldn't be the same as the test drivers anyway even if I did drive them.

    No. I didn't. Seeing as he had nothing to say apart from a sly dig at me I treated his comments as I saw fit.
    And if you consider what he had to say as comic relief, well then you are easily entertained.
    Good for you.
     
    Last edited by a moderator: Jul 1, 2017
    marcvic likes this.
  7. PzR Slim

    PzR Slim Premium

    Messages:
    6,369
    Location:
    England
    The thing is though if you build an accurate physics simulation it's going to obey laws of physics of the real world. If you then make sure the cars are accurately modelled in terms of their chassis dimensions and flex, etc and that your tyre model is accurate. None of that needs subjective input. It only needs objective, measurable values to be plugged into the model. This model can then be tested against the real world and the model altered until the data coming out of it is comparable to the same data you can get from the real world. There is no subjective, perception involved in that. You then get a bunch of different people to drive the cars in that model and if the model is a match for the real world their perception of the model should mirror their perception of the real world, no matter how it differs from one another.
     
    Last edited: Jul 2, 2017
    Imari and weedybroz like this.
  8. XxHighWayStarXx

    XxHighWayStarXx

    Messages:
    59
    Location:
    Australia
    You mean just like you don't get to tell me what I do or don't understand ..?
    Anyone with decent reading comprehension would have found my meaning obvious. No mind reading necessary.
    Work on your comprehension.

    I get what you're saying, and of course it makes sense. However no 2 cars of the exact same model handle the exact same way. So they won't feel the same either.
    Compromises have to be made for a simulation, so it can never completely simulate reality.
    That's all I've been saying.
     
    Last edited by a moderator: Jul 1, 2017
  9. Johnnypenso

    Johnnypenso Premium

    Messages:
    28,424
    Location:
    Canada
    Already asked and answered. Telemetry. Taking measurements of known parameters. Anyway, you're simply creating a strawman in using perfection as the goal. We aren't at the point where we can create a perfect replica at home and may never be. That doesn't mean that you can't get to 30% or 50% or 70% and different games have achieved different levels of accuracy. Discussions revolve around which games are more accurate than others. None of them being perfect =/= all of them being equal. It's an irrelevant position to take really.
     
    Last edited: Jul 1, 2017
  10. Scaff

    Scaff Staff Emeritus

    Messages:
    23,177
    Location:
    United Kingdom
    I'm sorry but that simply a gross exaggeration, in the past I have driven quite literally dozens of the same car model back to back on the road and track (I used to run product launch training) and the differences are so small as to be meaningless (unless something is wrong with one of them).

    They certainly don't feel so different as to make valid comparisons become invalid.

    No one has said otherwise, however that doesn't then mean we are dealing with something that is not objectively measurable.

    If what you are claiming was true then sims at any commercial level (and plenty exist in both the automotive and motorsport sectors) would not exist, yet they do. Simulation is used for design, development and even structural modeling and crash testing.
     
    niky, ImaRobot, Quakebass and 5 others like this.
  11. Lukanyon

    Lukanyon

    Messages:
    602
    Location:
    Brazil
    Perfect recipe: real life physics
    Alternate recipe: coding in order to try to re-create real life physics
    Cooks: game developers
    Tasters: players
    Taste: how the player/driver feels the physics.
    Developers don't try to re-create the feel, but the physics. Physics can't be altered (Newton laws and stuff). If the physics are close to real life's, than the feeling each player will have will also be close to real life.
     
    Wolfe, Knukel, UKMikey and 5 others like this.
  12. clos2727

    clos2727

    Messages:
    182
    thats just it. all that you named above, pertains to skills acquired/learned/experienced. has nothing to do with actually driving a car. anyone can drive a car. not everyone can race one. physics is what makes the simulator unique. if you've driven in real life and been on the track as well it isnt much of a transition in just driving. the big difference is how the vehicle reacts and behaves, even when you are not pushing it. (just driving) and you can quickly gauge a good simulator from a bad one.

    perfect example is Pcars physics. when the tires lock up/understeer or lack thereof. the way the car behaves is unlike anything you will have experienced in real life. it so bad.
    now take Assetto corsa, is perfect imo. its easy to jump in and drive. just going around the track easily paced. but once you start pushing it you need every skill you named above to keep the beast of the road not to mention stay competitive

    and that the beauty of Assetto. it differentiates the good drivers from the subpar and doesnt handicap you because others do not attain the same skill set.
     
    Johnnypenso likes this.
  13. weedybroz

    weedybroz (Banned)

    Messages:
    33
    Funny thing for me is i seem to be able drive the same in ac and pcars with close same results. On gt sport not so much .
     
    clos2727, Scaff and PzR Slim like this.
  14. Imari

    Imari

    Messages:
    11,134
    Location:
    Australia
    Would you really call it fear? I work with explosives daily. The first week I was admittedly crapping myself, but these days I simply have a healthy respect. I'm aware that I could come home in pieces tomorrow, but I trust in my own skill and my knowledge of how explosives behave to give me what I deem to be an acceptable level of safety.

    I think driving a car fast is much the same. At first it's terrifying, but then you realise that through your skill and some appreciation of how a race works you're actually fairly unlikely to seriously injure yourself unless you do something colossally moronic.

    There's a remarkable number of lethal situations that humans put themselves in and yet become very comfortable. See the gentlemen below. Fear gets in the way of rational decision making. You can't be racing at your best and still be scared. Those aren't compatible. You can have an appreciation of the danger, but you can't fear it and be performing at a consistently high level.

    I mean, you think that professional race drivers are actually scared every time they step into a car? It's a day in the office for them. If they were terrified then they'd get another job.

    [​IMG]

    Ah yes, like all the other people who are taking you to task over what you said. All of us are simply lacking reading comprehension. Gotcha.
     
  15. CLowndes888

    CLowndes888

    Messages:
    7,848
    Location:
    Australia
    You can perceive the handling of a car any way you like, but physics are physics. It's not a disputable thing, you have to accept them for what they are.
     
    PzR Slim and Imari like this.
  16. TomBrady

    TomBrady

    Messages:
    1,513
    Location:
    United States
    Well that's not necessarily an issue of physics. It can also easily be due to controls and/or force feedback but you're still somewhat correct in that what matters most usually is how it feels in the hands of the players. More realistic physics will certainly be more unforgiving, but not necessarily any less enjoyable to even a casual player if it feels great in their hands. This is why I believe force feedback for wheels, and controller calibration for gamepads is absolutely crucial to every racing game from sim to arcade. Sometimes even more so than the physics
     
    CLowndes888 and Imari like this.
  17. Imari

    Imari

    Messages:
    11,134
    Location:
    Australia
    Technically no, although because of the way games work developers tend to err on the side of making it easier for the player. But we can all imagine a game where the physics were tuned to make it harder for the player. In that case, making it more realistic would be making it easier. Hell, arguably a lot of early sims were exactly that. GPL is a famous example of being hard to the point of possibly unrealistically so.

    Really, a simulator is trying to hit a sweet spot. Not to hard, not too easy. We have more examples of too easy than too hard, but it can and has been done.

    Totally agree. If the controls are hard to use, it doesn't matter how sophisticated the physics behind them are. It also applies to real cars. One can have a fantastically engineered car, but if the steering is loose and vague and the throttle and brakes have slop in them then it's going to be a dog to drive regardless.
     
  18. XxHighWayStarXx

    XxHighWayStarXx

    Messages:
    59
    Location:
    Australia
    In your opinion.
    And I never used the word "perfection" - you did.

    There's the "perfect" word again.
    Nothing is perfect.
    And now you're definitively saying that other people's feelings of a simulation will be close to real life.
    How mighty presumptuous of you.
    Thanks for speaking on all our behalves.

    No, only you it appears. And do NOT put words in my mouth.
    The passage you quoted resulting in my comprehension comment was only misunderstood by you.
    Not all other people.
    Yet again you fail to comprehend
    written language. It's not my fault if you cannot understand the gist of a simple analogy & require every single word to be spelled out just so it passes your personal validation.
    Perhaps I should've drawn some pictures (using stick figures & bright colours, of course) so as not to potentially confuse you.
    Silly me.
    Oh, and numbers of people who disagree doesn't mean diddly-squat. You like your safety in numbers do you? (rhetorical question)

    I believe you, and I don't doubt your personal experiences, but they are your personal experiences.
    But I think unless one drives every single creation of every single model of every car ever built then one will never be certain if a model's handling simulation is actually accurate. I know it's impossible and sounds crazy but, there you have it.
    In the end simulators, as far as I'm concerned, may be at best,comparable to real life, but that's all.
     
    Last edited by a moderator: Jul 2, 2017
  19. Greek

    Greek (Banned)

    Messages:
    93
    We want driving simulators that simulate driving with the highest fidelity possible when a direct drive force feedback steering wheel controller is connected to the system.
     
  20. Animera

    Animera

    Messages:
    2,361
    Location:
    New Zealand
    No FFB wheel exists that will make you feel what you feel when sitting in and driving a real car.
     
    hasslemoff and Imari like this.
  21. Scaff

    Scaff Staff Emeritus

    Messages:
    23,177
    Location:
    United Kingdom
    No, objectively these cars all behaved the same, V-Box and telemetry software is a handy tool for that.

    I was also not alone, we would use teams of upto 30 professional drivers during product launches, not of them ever saying car 'X' drives totally different to the rest, don't use it; and given that we were legally responsible for the delegates it would be the kind of thing you would want to draw attention to.

    No you don't, and that quite frankly is an appeal to the absurd. You don;t need to use every copy of an operating system to know how it works or how you feel about it, you don't need to taste every one of your favorite chocolate bar to know what they taste like and how you feel about it.

    Your argument that driver 'x' may find a car to suit his/her style and driver 'y' may find the same car to be twitchy and unpleasant doesn't hold ground either. These preferences don't change what the car is doing or how physics operate. The car is doing the same thing, and that same thing can be modeled. If its modeled well then driver 'x' would find the same car in reality or a sim to suit his/her driving style and driver 'y' would still find it unpleasant and twitchy.


    And?

    Now I also note that you ignored the point about the use of sims commercially in both the motor industry and motorsport, why is that.

    Oh and stop with the multiple posts, the site has an easy to use multi-quote system, please use it.

    I agree, but that wasn't what was been said, The highest fidelity was what was mentioned, which would mean the best recreation of these feedback an actual car would give you.

    None are perfect, but many come close.

    The biggest issue is that often feedback you would not get directly through a wheel is modeled (as it would normally be coming through the chassis of the car via your body), I'm lucky in that I have a tactile set-up on my rig which better simulates that kind of feedback. Some sims then allow me to remove that from the steering force feedback (Pcars and AC), which results in a much higher fidelity experience.

    Unfortunately, on the current beta build, the limits of the wheel is not the issue, the poor and incorrect FFB is.
     
    Johnnypenso and Imari like this.
  22. Mr Tree

    Mr Tree

    Messages:
    952
    Note that I was editing my post to comply with forum rules. As that one was sent in anger yet I couldn't change it and now it's gone.

    @XxHighWayStarXx you really don't understa'd these things. I advice you to take some classes on the subject.

    I'd also like to kindly ask to stop spreading these factually incorrect idea's of you. They are a serious insult to people who actually took the time to educate themselfs on said subjects.

    So while you might not acvept accademic titles I advice you to get one yourself. Titles don't mean everything but when the entire physixs world disagrees with you it might just be you that's wrong. If you disagree with that premace well watch out for thrme lizzard people then.
     
    Lukanyon and Imari like this.
  23. Johnnypenso

    Johnnypenso Premium

    Messages:
    28,424
    Location:
    Canada
    Yeah, you did. It's implied in this statement.
    Reality is perfection. If it can't simulate reality it's imperfect. And again, it's a strawman position. And nice job acknowledging that I answered your query about how to measure reality vs. simulation.:tup:
     
    Last edited: Jul 2, 2017
    weedybroz, Imari and Lukanyon like this.
  24. RACECAR

    RACECAR Premium

    Messages:
    31,502
    Location:
    United States
    I wonder how many arguing that Racing sims should be hard have experience both behind the wheel of an actual car (in actualy hotlapping or actually competing) and virtual. I don't say this to mock anyone, but I'm genuinely curious as to where the majority of this mindset comes from and whether its from those with experience in both real and virtual world or just those that use a wheel in the virtual world.
     
  25. Animera

    Animera

    Messages:
    2,361
    Location:
    New Zealand
    I would say that it isn't driving the actual car that is hard.
    What is hard is learning good racecraft and judgement.
     
  26. PzR Slim

    PzR Slim Premium

    Messages:
    6,369
    Location:
    England
    I'm not sure anyone is arguing they should be hard. I see people arguing they should be realistic. So pushing a car to 90% of it's potential should be straightforward but trying to drive them at their limit should be difficult, as it is in the real world.
     
    Imari, Northstar and Johnnypenso like this.
  27. Johnnypenso

    Johnnypenso Premium

    Messages:
    28,424
    Location:
    Canada
    Reaching the limit in any game is difficult no matter how realistic the physics. In spite of it's obvious shortcomings, the leaderboards in GT5/6 always had the same people at the top and they were separated from the rest of the pack by a substantial margin most of the time. IMO they are two separate things. Physics accuracy is really about how the car performs, especially at the limit, in comparison to real life. It might be difficult as it is in a Lotus 25 or 49, for example, or a little easier like it is in a GT3 car.
    As I said above, some cars are more difficult than others at the limit and an accurate physics engine will reflect this. The discussion I see mainly revolves around the accuracy of a given physics engine and not that driving in general should be difficult. But getting to the limit is more of a skill thing than anything else and you either have it or you don't in my experience. Some will grow and learn but the movement from Joe Average, 4 seconds off the pace, to uber alien, is pretty rare. I raced karts as a teenager and at a pretty high level for the day, in my last couple of years. I was very fast, usually on the front row, but I never once thought it was hard because it just came natural. I raced guys in the same equipment who had 5 or more years experience and the only thing that separated us on the track was our natural skill level. I'm sure they would have found it extremely difficult to balance the kart on the edge like the front runners did but for us it just came naturally.
     
  28. CLowndes888

    CLowndes888

    Messages:
    7,848
    Location:
    Australia
    So your suggestion is that some cars are more difficult to handle on the limit? If Project Cars is anything to go by, than I agree with this statement. GT3 cars are much easier to handle on the limit compared to a classic F1 car without all the modern day aero.
     
    Johnnypenso likes this.
  29. GumShoe

    GumShoe

    Messages:
    219
    I think Kaz made a statement that in no way advances the discussion or genre.

    Couple of thoughts:
    1. Of course driving difficulty depends on the car, it's setup, the track condition, the driver's experience and familiarity, the environment (race, time trial, importance of outcome, etc).
    To say driving is simply hard or easy is pointless without proper context.
    Driving any car below limits of the car, environment and driver will always be easy save unpredictable events.

    2. On simulation, there are objective and subjective elements. Objectively, Grant Turismo does not simulate things the competition does (tyre pressure, track temp, flat spots, etc- sorry if wrong but there are things here). But it's like that cake analogy except...

    Recreating car physics is like recreating a cake with 1million ingredients (some not known). You'll never make it perfectly and you hope that all the ingredients you do add are correct and the assumptions for the rest are that they don't change the outcome significantly. Then everyone applies their own taste eating with different tools (hands, spoon, fork, etc) and comes to a subjective conclusion:
    Some feel it's just like the original, some say it's terrible, some note that minor flavours are missing or taste funny and others think the tool used doesn't allow the best cake eating experience.
    Some have never eaten cake before, others have eaten the real one 100's of times and know the finest details.

    So you see, a blanket statement that driving is easy or hard, and how racing games address that seems rather useless.

    People need to accept that there are many valid opinions that are supported by very different contexts... And if it meets your expectations/assumptions, enjoy it!
     
    clos2727 and Johnnypenso like this.
  30. Greek

    Greek (Banned)

    Messages:
    93
    You lack knowledge about many things. You must study. There is a problem when opinions are presented as facts.
     
    Mr Tree likes this.