So the basic concesus is that for it to be a sim it has to be hard to controll?
I don't know what is the "basic consensus", can only tell you a few things, if I may.
Beginning with something someone said above (I think Kevin Jay):
"
Physics are impossible for most of us to accurately assess because of the lack of firsthand knowledge of how the actual car handles."
Physics, whether it be that of flight simulations or that of racing sims, is a complex subject. As I said to jgf at NG recently, though, anyone who wills himself/herself to study the underlying physics can understand what goes on and easily perceive if a racing sim has it right or not - in terms of physical realism.
Kevin above
mixes two concepts which are different and should not be mixed: physics and "how the actual car handles".
How a car handles is the result of physics, yes, but to a large extent of setups. You can ruin a perfectly good car with a bad setup, and you can compensate (to some extent) design flaws with proper setups. No way around this: setups, being the result of the interaction of team engineers and pilot (with or without analysis of telemetry), are of vital importance in determining how a car handles.
Knowing how a car "handles" is, therefore, best conveyed by someone who actually has or has had the chance to drive it. I'm thinking about drivers such as Ben Collins or Tanner Foust who have had the opportunity of driving dozens of cars, some of them professionally, but any driver with experience should be able to elaborate on how the car "handles".
The Physics of a car (aerodynamics, chassis, tires) is assessed through telemetry and equations (models) with or without the help of specific software (think about CFD models or tire models run with a variety of software and in different platforms).
Knowing how a car handles has a human element to it, the engineering/physics of a car does not.
Now, people may say "realism encompasses physics and car handling", to which I say no, not true. Car Handling is variable, as it depends on setups; one team sets up the Ferrari 458 with a slight understeer, another team sets its 458 up a bit oversteery and yet another team sets its 458 up to "feel" neutral. Different teams, different setups, different handling. Which are you, as a developer going to model? Better yet, which is the realistic 458? Probably none, probably all, it depends.
With that in mind, we should be able to cast aside definitely the Big Myth ("if you haven't driven a car in real life you can't tell if the simulated car is realistic or not").
-
Addressing Klazerman's "basic concept" of "a sim [...] has to be hard to control"...
No. At all. I have been arguing against that for years now. While it is irrelevant who started this absurd notion, fact is a lot simracers think that way, the public at large thinks that way (remember Jeremy Clarkson's introduction to Greger Huttu's prowess with iRacing: "the fiendishly difficult simulator") , therefore many a developer went down that avenue, often "forcing" the simulation physics to be "difficult", "hard", because otherwise the target audience might reject the sim as being arcade.
No. My own limited experience with real life racing drivers gives me a few pointers about what racing is and driving a race car is.
They (pro drivers) make their living this way, they have to be competitive. They have to give it all and demand it all from a car. This alone means they will be frequently driving on the limits.
Driving on the limits, so they tell me, is not hard because the car will lose traction inexplicably at any time (low speed, high speed, low load, high load). It is hard because it is hard work the whole time, because it is physically exhausting, because it requires a lot of concentration to find that small piece of Heaven which is maximum grip and good, natural handling. It is hard because they know they are pushing, they know some components don't last the whole race (sometimes not even part of a stint) and they can (a) lose time in corners or pits, (b)lose the car (engine compromised, suspension damage, ECU probs), (c)seriously compromise budgets if something goes catastrophically wrong and they crash a car.
Race Cars, nowadays, often have good aerodynamics packages, which provide high downforce at a "minimum" cost of drag. They're also served with composite materials which make them lighter and more manoeverable; they're also fit with light and resistant engines; large tires made from rubber compounds which provide dramatic adhesion and good resistance to wear and temperatures.
There is a
MASSIVE AMOUNT OF GRIP available in these cars. Dave Basu, pro racing driver (still competes at Dutch Supercar Challenge, Bram probably knows if it is so) in early 2011 at iRacing forums said that the real Corvette C6r (which he used to drive) was "very stable", "very much planted", "drives as if on rails". That was a GT3 C6r, if I recall correctly.
Stable. Planted. Drives as if on rails.
I have been told the same over and over by these guys. I have read the same impressions by the likes of Oliver Gavin or Tommy Milner.
The key is then, and has always been, how that grip builds up, how it can be kept and how it is lost.
For that we have different models as most of you know, from the Fiala and Pacejka ones, to more physical, transient ones (what Stefano, Gjon and Dave call "more dynamic models").
Which model we use is irrelevant, though. The objective is the same: provide grip where grip is due. No more, no less.
Sorry guys for the long winded post.