One of the things sim racers in all their talk about realism just never seem to understand is how much work goes into making a real car handle the way the driver wants it to. There are entire series of chapters in vehicle dynamics books dedicated to exactly that aspect of handling: Controllability. I almost never see this discussed in the sim racing community. In fact, most of the books aren't nearly as much about making cars faster as they are about getting the handling a certain way so the driver is happy and doesn't struggle with control. Fast, but as predictable as possible and easy to drive. The car does what you want it to do. That means a happy driver. A happy driver is a fast driver. Fast drivers win.
Sim racers should talk less about realism and more about good handling versus bad handling. You can make a real car that spins every time you look at the radio if you want. It's real so it's 100% realistic, but it's terrible handling and nobody wants a car that drives that way. A race car will not drive that way because a race car driver wouldn't tolerate it if it did. If it's unpredictable and doesn't do what the driver wants, it's going to be slow and a miserable experience to drive.
The driver will bring the car to the garage and say, "This sucks. Every time I breathe on the throttle in the hairpin, it wants to spin, and once it gets a little bit loose, it just loops around without any warning and I have no chance to catch it. This is awful. Please fix it."
Engineer says, "Ok," and fixes it.
Driver says, "yeah, I like this better."
There's no conversation about realism at all, it's just about good versus bad handling. The bad handling version of the race car is just as realistic as the good handling version. There's a big difference though: One sucks and the other doesn't. The real race car doesn't suck.
If a sim racer has the same complaint about a car, the "engineer" (forum reply) response is this:
"Well, that's how the real car must drive because this is the most realistic thing ever made and how would you know if you haven't driven the real car and a race driver on TV once said it was pretty realistic and the physics engine does this and the tire model does that and you wouldn't know anyway because you aren't a real driver anyway and your forcefeedback sucks and you need a better wheel and you can't feel the g-forces and your perspective is wrong and your field of view is wrong and I've been driving sims longer than you and you're using a computer desk instead of a racing rig and you need a motion platform and the fear factors isn't there and and..."
Umm.. No. Stop it. The handling is just awful.
I had a back and forth with Matt Sentell once about this (Papyrus, he worked on IndyCar Racing II and probably I, although this was while we were trying to help with the offending sim). He essentially said that he often wanted to intentionally give players an absolutely horrible car, one that 100% could be proven was completely wrong in every way just to see what the reaction would be. I remember him saying something like "the hardcore alien types would figure out how to drive it anyway, think it was brilliant, the most realistic car ever put in a sim, and rip on all the ninnies who couldn't drive it because they just didn't have the skills." I'm paraphrasing, but I clearly remember him using the word "ninnies."
Back to bad handling: When bad handling is found in a sim car, a lot of sim racers seem to say something like "how do you know the real car doesn't drive just like that? Have you ever driven it? HUH?? GOTCHA!!"
While they might be right on an especially lucky day, they're probably not. Here's how I know it doesn't drive that way even though I've never driven the real car:
Because no race driver on Earth would tolerate it if it did.
A real driver would demand the engineers do what they can to fix it. Failing that, the tire companies would give them different tires to help the handling problems. The tire companies have to pay as much attention to this as the chassis people. They're not just throwing any old set of tires off the shelf and saying "good luck with your setup." They're engineered specifically for the cars, and in some cases, the combination of the car together with a specific track. Nascar uses different tires for every track on their circuit that vary tremendously from each other. One of the most important goals is to not kill the drivers.
The trouble with sim racers is so many of them seem to think that if their sim car is really hard to control, it must be realistic. I think this is an old vestige from the days when us old folks all went from Pole Position type games to stuff that had decent vehicle models like Hard Drivin' and the PC sims that came after it that were running similar physics and tire models. Sure, they were harder to drive than Pole Position, but they were also more realistic in the sense that they did all the weight transfer effects, the tires had slip angles, and all the other things that were completely missing from Pole Position that make it feel like a car. Not a good car, but a car nonetheless.
So the "harder = more realistic" idea, which was true then, stuck. I've never seen it take a vacation since. If anything, it's gotten even more attention.
Most of the "arcade" games today have more in depth physics and tire models than Grand Prix Legends had. After having worked with a couple of Need For Speed guys on Grand Tour Game (which uses the rFactor physics engine if you can believe that, I was hired to dumb it down into an arcade game which was very interesting), I can say for certain this is true of the Need For Speed series over the past many years. What made a game "arcade" 20 or 30 years ago was the physics model. Today and for many years already, not so much. Now it's more about the parameterizations of those models and some differences in tire models.
The trouble is that back in those early PC sim racing days, nobody really knew what the heck they were doing. These were computer game devs at best, not vehicle dynamics experts. Hard Drivin' and Race Drivin' nailed the physics especially well because Doug Milliken himself tuned it (he's listed as "test driver" in the credits because Atari didn't want to tip off competitors as to how they managed to pull off that miracle). It was using one of Milliken's own models from the automotive industry: A classic bicycle model (later expanded to a 4 wheel model for Race Drivin').
While the PC sims that have come since are more sophisticated by far than Hard Drivin' and Race Drivin', I don't believe the rest of the PC world guys were anywhere close in handling knowledge in those early days, so while it was true that "hard = more realistic" when comparing Grand Prix Legends to Pole Position, they were largely still awful because nobody back then had much of an inkling about how to tune a car as well as the Doug Milliken types could. Remember that the internet wasn't really a very useful thing yet so you were really left figuring it all on your own back in those days. I think my first book on the subject was "How to Make Your Car Handle" by Fred Puhn which I picked up at an auto parts store one day. That book got a lot of things right, but a lot wrong too, especially about tires. Those were the kinds of books we were all looking at back then.
So realism or no realism, the reality is that most of those early sims still suffered from one thing: Terrible handling at a level no driver would put up with.
Today people talk about realism to mean "the car drives the same way that the real car does." I think that's the proper way to think about it. That means, however, that if the sim car doesn't drive like the real thing, it's not "realistic." This also means that if the car is simply horrible to drive to the point where a real racing driver would want to take it back to the garage and yell at his engineer or the tire people, while at the same time not having the same urge with his real car, you can be pretty sure that the sim car isn't "realistic." There are substantial differences between the sim car and the real one. That's what I'm seeing loud and clear in Austin's video.
I've always thought it unwise for sim racers and developers alike to battle drivers on this. I've never understood it, quite frankly, unless it's the driver's first three seconds on a sim of any kind. Sure, there might be a little adjustment period for your brain to toggle the right switch, but once that's past, they are interpreting the vehicle response to their control inputs about the same as they would be doing in reality, something they know better than anyone else. Love him or hate him, Austin's video is a brilliant example showing just how obvious the differences between the real cars and the simulated ones really are, at least in this particular sim's case.
Less talk about realism and more talk about good versus bad handling. When someone says "the handling sucks," people really need to stop chiming in with "yeah, but that's how the real car is."
No. It isn't.
