Why is AC more popular than rF2 in terms of modding?

I shouldn't have trusted my memory, I think what I recalled was a lack of (or insignificant) change of rolling resistance with differing tyre pressures, rather than no rolling resistance. Since others above have referred to "low" rolling resistance, I guess your comment here is directed at me, so consider that addressed.

 

Yes, that could be true. I've mostly just checked tyre rolling resistance with a single pressure. Something to test in the future. It only takes 5 min or so. Then again, any hard data on real-life response to compare with?

Well, not directly directed at you. I know that you know more than the usual person (most certainly more than me), so I assumed you were recollecting a read statement from one of the other 123 forum threads where everything within the physics is apparently broken.

 

I got the same "feeling" with lack of rolling resistance "relative to tyre pressure changes".

Speaking my own experience, me and several friends used to race in "live for speed" endurance leagues back in days. And in live for speed, there are immediate straight line speed reduction and increasing in tyre heat/wear when going from high tyre pressure to low pressure, which requires careful balancing to find sweet spot. But this behavior is missing almost entirely in RF2 when we initially started making setups in RF2.

In LFS, our racing team usually use a bit higher tyre pressure, especially in endurance race, to achieve less tyre heat and less wear, and gain a little more straight line speed. But in RF2, higher pressure would mostly only result less grip which in turn makes car more slippery and then more heat and more wear, and there is almost no change to straight line speed with different pressure.

In the end, we all end up going for the minimum tyre pressure in RF2 since there were no drawbacks for using minimum tyre pressure, as it was also mentioned by others in many old threads.

-------------

Lastly, also speaking from my own experience, we used to have a nice local sim-racing community for over 20 years, which we used to embraced all kinds sim games no matter pros or cons, including the most popular AC & and less popular RF2. However the community ended up split and broken apart about 2-3 years ago, because people enjoy debating which sim has the most realistic "tyre slip angle" or "tyre pressure" behavior (of course without real data to back up their points). And it did not take long before it all turns into "group fights" where fans from different sim games were starting to hate each other (including some of my old teammates because they could not convince me that AC is better than RF2).

So in the end, our community were split apart, and our local RF2 league was finally ceased all activities since a year ago due to group fights and lack of new comers, despite all the efforts for trying to save the league.

It would all end up differently if people were more open minded and stop arguing which game was best.

 

Are you referring to GT tires? I said (GT) tires wear too fast based on what Michelin said after one of the virtual LM:” Meanwhile, the wear rates factored in by the software were higher than in reality. The game effectively reproduced a cost of two seconds per lap (one stint equivalent to approximately 10 laps) due to wear, whereas – in reality – Michelin’s tyres deliver consistent lap times from the start to the end of runs of up to four stints at a time. Michelin hopes to be able to work with the rFactor2 platform’s designers so that competing teams have a broader spectrum of possible strategies in the future while demonstrating the consistency of its tyres’ performance more fully”

 

On more serious note, did you wach this:
?
I’m sure Lazza will quickly point out, that was oversimplification , but at least literally everyone can roughly understand it (without need for a mechanical engineering degree )

 

AC is all good. But can't have an experience quite like this in AC: . Not that there are a lot of guys in rF2 that cares about it and instead chooses something that can't do this.

 

Getting tyre wear rates noticeably wrong is disappointing, because it's a simple thing to adjust. This isn't a factor in physics engine discussions, or even in modding difficulty terms between games, because in most (all?) cases it's just a handful of numbers you can increase or decrease to adjust the overall wear rates.

Different models may allow for differences in the nature of that wear (how the quantity relates to temperature, angles, friction speed), but that's likely a level above what all players (and nearly all modders) could judge the accuracy of.

 

Yeah that seems like something that could be adjusted to be closer to real life by changing a couple of real time values for the GT tires but IDK. Sadly AFAIK it wasn’t corrected for 4 years since that virtual LM and I’m afraid never will - as most of physics issues in rF2 .
Since S397 took over development, physics were never high priority as they were already considered rf2 strength . They focused on graphics and UI instead which usually make more sense from business standpoint, especially for a PC video game. But rf2 had potential to be much more. I know some real teams used rf2 (yes I mean rf2 not only rf pro) to make (closed/ private ) mods. But I guess there was not enough potential income to justify focusing more on the physics.

 

ISI mentioned drivetrain in 2013 and nothing ever happened, so I'd be wary of putting this at S397s door (actually with the contact patch distribution fix and other more recent changes, I think they did ok, albeit still seeming to miss some of the simpler stuff - which always leads me to think there's a commercial reason you don't make an open game that really can do everything).

I'm more bothered by enhancements to LMU (whether actual physical model improvements, or realtime parameter style tweaks) that reportedly (from a dev) allow slip angles that aren't possible in rF2.

That bothers me because rF2 development obviously came on the back of rF1, where one of the most impressive mods came from Michael Borda (who later joined, I'm not familiar with that timeline) based on F1 2006. The amazing new tyre model surely must have been intended to cover the gamut from modern narrow slip range tyres right back to bias plies, so to have an inherent inability to reproduce one end of that range seems a large oversight. Maybe that developed over time with changes to fix other issues, and maybe I'm just misinterpreting the context, who knows.

 

AC is not perfect, nor is rF2, we all know that the sensations of real life will never be surpassed by a simulator.

No offense intended, what should we see in that video that AC doesn't achieve and rf2 does?

 

The QSA fix by Michael Borda is the only major one I heard S397 did. IIRC Michael was tires “guru” at ISI and S397 , but actually didn’t create the tire model used in rF2. I think the guy that (co?)created it is now working for iRacing (since year or two). Anyway, later on there were only very small updates to the tire model.
Also the QSA fix was done because new, confirmed data came from tire manufacturer (Michelin IIRC).
Yes, sometimes real tire data is unreliable, or at least far from complete (e.g. only narrow slip angle measurements, same ambient temperature etc.) , but if empirical evidence consistently contradicts physical/theoretical model then it means something is wrong with the model. Anyway, It’s good the issue was found and fixed.

Regarding the part you said is bothering you: I’m not following LMU development, but I’m also surprised they say slip angles were too narrow in rF2 . I thought that at least GT tires in rF2 didn't punish sliding too much (so allowed high slip angles) . Also there are a few historic cars (made by ISI) that have tires that allow high slip angles (as it was IRL).

 

I was referring to physics generally, not just the tyre model. (Terence Groening is who you're thinking of btw, the TG part of TGM)

IMO what you're doing here is overstating that the model was incorrect, while understating the data was incorrect; exactly the same sort of data that empirical models are built exclusively on. Improving the modelling after that discovery benefits all tyres subsequently built on the physical model, and presuming all existing data was being used to make the tyres in the first place the rF2 can't be any worse than an empirical model. But people see "fix" and assume the tyres were more wrong than other games.

I didn't spell it out, but the slip angles I referred to are narrow ones. I.e. LMU allows narrower slip angles than rF2, or (perhaps) more flexibility on the dropoff rather than just the peak angle.

 

Yeah I just found his name in some other thread. And he’s been hired at iRacing since 2021. I wonder if he’s working on improving NTM on dry tarmac (I’m not fan of racing during rain).

My intention wasn’t to say rf2 tire model was generally wrong. It was still one of the best if not the best of all racing games/sims. But the QSA change managed to fix some behavior in the tire model.
Second part I’m not sure I get it , because it’s surely possible a physical tire model is less accurate than empirical data used to create it. I.e. tire measurements might be perfect , but a model still flawed in some aspects.

 

I think you should focus on what the QSA patch actually did.

 

It is the "je ne sais quoi" that is missing haha

All these multiple micro corrections. Neutral slides. The exact moments that tires either tunr into all slip, or bites in firmly. The subtle aspects of steering feel that also goes in the loop with all these things. Extra characteristics of car being on the limit. It is not something easily described, but after many hours of driving and being curious enough to examine closely real cars behavior it should become apparent.

AC is good, but it is simply smoothed out by its simplicity. The simplicity of AC is its superiority. But it also cuts down these special bits. For many simracers it is enough. This is why sim like rF2 is a level up. A level up that is has not revealed itself enough.

For example I don't think that it is even possible to simulate car like classic 911 down to its very subtle details of having very versatile response to driving style, the way it can be driven in such balanced way that corners would be taken 90% in neutral steer. AC doesn't seem to be quite able to do it. Yet it is doable in rF2. But we have lots of mods of classic 911 cars that barely does what they should do either., it is not a fault of sim though. You can throw and keep them sideways just fine though (as they should be able to), so I guess Nik Romano and most simracers finds them good enough.

 

There are many details there, dusty track, very bumpy, used tires, and you can also see that he is pushing the limit to control the situation with the sole objective of having fun and never to achieve a good time.

If you want to have fun with constant corrections you have many mainly classic cars to achieve this with, increase the wear and choose a dusty track to increase these corrections.

 

Checkout this post:

Honesty speaking rF2 seems a lost cause

Especially relevant to modding:
“Tires in rF2 are in some ways sublime. From tech side. What often lets down is how they are utilized. Code behind tires is a masterpiece on many fronts but for it to work tire has to be built with extreme care and has to have chassis that is flaw/bug free. You can have best possible tire on a car but if chassis/aero is screwed then no amount of greatness in them will save the car from being absolute shiet. This is a trade off that rF2 has. It can be truly amazing when you put enough time, care and science behind a project like car with lots of data, know-how and fine tuning. It is a sim that can't be rushed when making a car because if you rush your making process in others sims you risk making a average car instead of good or very good car, while in rF2 if you rush it you end up with absolute dog excrement of a car. Get one thing wrong like chassis flex/ultra flex and whole car will feel off, even tires won't behave correctly.”​

 

So M. Borda wrote on the blog:
“Some correlation issues crept up over time, and as we’ve collected more data, it became apparent that it wasn’t on the data side. You may think that this might be sloppy, however, the reality is the way data is measured, interpreted (smoothed / adjusted / fitted), scaled, or worst of all, even copied between tyres, makes trusting data a very difficult thing to do. For those interested, Niels Heusinkveld, had a more in-depth piece about this kind of data ‘fitting’ which I pointed to in a previous entry. Anyway, without derailing myself too much, it’s not unreasonable to suspect issues with the data itself, in this case we had strong suspicions that the data was measured under a single condition, simply offset and then applied to different data points. This doubt left plenty of room for us to believe that our model was probably correct, when considering the obvious short-cuts the manufacturer had taken in measuring the data. So everything was rosy, we thought, and then we finally obtained the same type of data from another tyre manufacturer. This time, they went to the extra step of measuring at multiple loads. Once we had this corroborating information, it became obvious there was a glaring issue with our tyre model. Of course, this was an original part of the tyre model that hadn’t been touched for years, taken for granted. Furthermore, this was essentially a non-issue before the introduction of the contact patch model. After a little thinking and investigating, it became obvious that our ‘harmless’ simplification, wasn’t so harmless after-all.”.

So if I understand correctly they had a simplification in the tyre model (QSA part) that seemed good/close enough, especially before contact patch model was implemented.

Later on have been receiving more data and some small issues occurred, but they ignored it because assumed it was issue with data they received (which is common), until got data from different manufacturer who provided same type of data but at multiple loads(which was missing previously) . With contact patch model the issue was obvious(especially at high loads) and Michael managed to fix it.

So to sum up: they thought the tire model was correct despite some minor issues with data correlation, until broader data (I.e. including previously missing data) was provided by different manufacturer.

 

I remember one guy in AC forums shared opinion that tires in AC doesn't wear enough, perhaps even correctly guessed the cause - don't remember precisely it was like 2016 or smth like that... Devs didn't quite take that seriously.

Some time later. A year maybe. Kunos discovered that something regarding tire loads was missing in tire wear computations and fixed that.
Shout out to Kunos for being open and transparent that time.

What this situation told me:
1. People, thousands of people, will use wrong stuff for a long time and will never complain/criticise as long as it is fun and has no experience breaking issues. At least when being in the community circle.
2. Devs believe what they do is right, because they try to make things right, and perceptions/opinions/observations of other people are either not paid attention to, or getting a slow reaction, or devs discovering issue themselves and that someone guessed the issue long time ago is just a coincidence.
3. According to "true" fans the realism of simulation was 100% before, and 100% after. Because the way drifting car works remained the same. Although I imagine drifting durations decreased. And things are still fun even when they aren't completely correct. Even more so - sometimes things are more fun when they aren't completely correct - sad face.
4. Simulation should never be taken for granted. Almost certainly there are things in every sim that doesn't work how they should. And either no one knows about it, or developers will ignore that till they will have same opinion and that is understandable. Sadly, I think, there is an issue in rF2 where there is not only lacking of time (a lot of side projects continuously leaving rF2 on hold), but there is also hardly anyone specialized enough to solve issues. I wouldn't be surprised if TG himself would struggle a lot, or maybe he wouldn't.

anyway on side note playing with bit advanced tire stiffness parameters in realtime sections, I think it is possible to simulate flat tire, possible to give more rolling resistance and make tire loose and springy, or just loose a lot could be done. So theoretically if it would be somehow linked to tire pressures the nit could be fixed AND THE REMAINING ISSUE THE MOST MADDENING ONE WOULD REMAIN THAT YOU DON"T EVEN HAVE TO USE CLUTCH !!!!!!1111

Anyway I think this AC and rF2 comparison is a bit flawed, because what is being compared is:


VS

 

One of the main problems I see with the AC tire is the impossibility of overturning a vehicle by climbing two wheels over a high curb, it doesn't seem very realistic in this video either but surely in rF2 it is possible to do it in a mini/abarth/clio with suspension rigid and slicks.



In AC, when you tilt the tire, it will simply slide, preventing overturning. I seem to remember that this is solved in ACC but it is a deficiency in AC.