I see you play around with modding in rFactor and many of them can be found on youtube.
I just wonder - if you had proper physics simulation: you put real input data for tyres, wheels, all suspension settings, and so on and you get quite accurate output results, what minimum settings and what range of settings you need to make out of, lets say, XRG any frontengined, RWD 1T-1,5T mass available?
- mass (+-200kg), CoG, rotational (?) inertia (3 axis),
- types of suspensions?
- presets for engine types (torque, mass, inertia in transmission)
what else?
Play around!? A bit more than that! Or at least, my bills get paid now doing this. :P
You say ''if you had proper physics simulation'' .. Well in many ways current sim engines *are* proper physics simulations, and I do simply put in real settings and the results will be cars that stick to their real performance. Wether micro detials in handling are the same, thats not always the case, but its certainly not playing around, its fairly scientific. While rFactor currently models more than LFS, the physics engines are all.. physics based so enter the right weights etc etc and it should behave somewhat well, especially as these ISI guys and Scawen have their head screwed on decently well, no matter how much we complain about certain details.
While its all about entering the right numbers, you need to know what real numbers are, and especially regarding tires, you'll find a lot of conflicting and dubious data. Then you have to figure out how to translate the data so the sim works with it properly. Plus you can't isolate any part of the car. If you're an aero expect you could find lots of lift and drag data and copy it to the 0.0000x% accuracy, but if you don't know what 'tire stiffness' is, it won't bode well for mod.
In many ways I try to do what AndroidXP suggests a 'fool proof' tire model would do, enter size, and a type of compound and it will adjust the sim tire model parameters. I'm glad such a system isn't in rFactor as I like to 'invent the wheel' (at the risk of doing it wrong) and find out more and more about how they might work. If I could only enter size and compound softness, I would still be limited to what the programmer thought was realistic.
In many ways its quite amazing how decent LFS and rFactor can behave even when you know certain things are absent or poorly modelled. It bodes well for the future, provided Scawen and Terence remain functioning with their heads screwed on, which is fairly likely as I don't know many succesful decapitated programmers.
Btw, I (sadly) have no idea what direction rFactor 2 will take, but I do see they're having a real go at it, including the tire model. No matter how many people dislike rFactor, its rather impressive and even annoys Todd Wasson by being so damn clever with its suspension modelling.. There are just 1000 ways to screw it up, and thats where LFS wins, as Scawen has more knowledge than 990 of the 1000 modders so you get a far more coherent experience. Still, LFS would've been better if a true physics 'parameter' expert did the cars, as I'm sure even without the tire and suspenson physics updates, there is more to get from LFS as it is now!
First thing I found out feeding LFS with service manual suspension settings, was that it nose-dived as hell and suspension travel looked much too big as with current model most of shock absorbing is taken by tires. Therefore my idea of LFS' simulation is that relations between output and input data are not in fact correct but they are tweaked way around - to get roughly proper output data (for .raw files and forces). VWS made me quite sure that Scawen cannot leave it as it is because VWS is real and widely known car so it would be widely commented when real settings and simulated results would differ alot - which I gladly welcome due to better quality of the sim
as I know from my sources, in fact, tires are the part that is not completely modelled theoretically - in case of rain pattern cuts they are surveyed empirically, not designed from theory. But all the rest is very precisely modelled and surveyed.
You should be making second level mods tires types, suspension parts, engine types used then in mods
Well, I think a sim should have really good tested physics environment of the cars, presets for modding with restricted settings range (tires, suspension parts, engines, transmission parts... and certain calculations - ie. ccm vs. engine mass -> weight balance because of course everybody would make the car with biggest engine, smallest weight or tyre grip vs. durability and so on, so some realtions should also be included).
They could be distributed with game updates. The mods could be downloaded like skins - they shouldnt be much bigger than settings files. If someone disables CMX download (which is couple times bigger I think) he would see just game default body of the type.
Therefore cars could be fabricated at two levels - presets for parts and cars themselves - trying to get best results in class.
Mods are free indeed, but I am fortunate enough to work for a simulation developer (not retail) 2 days a week, and for a GP2 team, and for Reiza Studios. I'm not complaining
I know what you mean, you enter real springrates, hopefully corrected with motion ratio and the nose dive is too great. LFS devs recently found this out as well, its in the suspension 'anti'. Anti dive is a fairly tricky result of suspension geometry currently lacking in LFS. You can angle the wishbones (in a side view of the car) such that there is zero pitch during braking even with soft springs.. I believe LFS will have this in the future update.
Tires are of course understood well, but you would be amazed how many crap you can find on the internet about them. Typically in semi scientific books such as 'tune to win' or 'going faster' and worrying titles like 'make your car handle' . Lots of bogus, little real data. Its rare for example to find complete lateral and longitudinal slip tests for the same tire tested on the same rig. Raw test data also has a lot of noise in it, and the result of curve fitting models like Pacejka often gives a horrible view of the situation at greater amounts of slip.. Its a slippery slope those tires!
well, I derived my opinion from much more distinct suspension moves in nKP (try LFS on rumble strips or checkerbrick(?) - nearly no suspension moves while you can feel it big on wheel /rumble effect/ ). Therefore when I saw those nose-diving with normal settings I supposed that less suspension moves, bigger absorbing on tires is the way to set LFS cars for proper feel.
I got that my knowledge from my dad, retired university scientist specialized in suspensions and hydraulic transmissions, so I roughly know what surveys were and are taken. And simulated
I'm just an amateur (thats why it took me 2 years to design and make that green one and I am still just about to run the engine )
I've never played this game before, but most american road racing games are not as good as europeans games who have full support on there.
Our most realistic games are NASCAR type oval racing games which I love, but some times I wish we'd make both in one game (Grand Trismo might be 1... maybe)
Yes it's called ICantStandAnnoyingPeople-itus. It's an old guy trying to be 20 again and a pair of tits stumbling over every other sentence they manage to read out from the cue cards.
All I learned about rFactor 2 is that it's evolutionary except for the tire model, and it doesn't run on crappy laptops, which they apologized for about 17 times.
The interview is obviously there to create some anticipation, yet the "dude" doesn't want to say too much. He has a way of putting things that sounds not too far away from Scawen's when he's posting about LFS.
rFactor is probably the worst game i ever played. Even Need for Speed is easier and more intresting. Thats all there is to say. Comparing it to LFS is like comparing a rusty tractor to a brand new Ferrari.
I tried rFactor about a year ago, the biggest thing that put me off was trying to get my G force pro setup. Even with realfeel I couldn't get the same steering feel as in LFS.
As for everything else no complaints, plenty of variety, but the steering needed to be plug and play who wants to stuff around tweaking configs.
A bit more than that! Or at least, my bills get paid now doing this. :P
Therefore my idea of LFS' simulation is that relations between output and input data are not in fact correct but they are tweaked way around - to get roughly proper output data (for .raw files and forces). VWS made me quite sure that Scawen cannot leave it as it is because VWS is real and widely known car so it would be widely commented when real settings and simulated results would differ alot - which I gladly welcome due to better quality of the sim 
tires types, suspension parts, engine types used then in mods 
(thats why it took me 2 years to design and make that green one and I am still just about to run the engine 