Released GTE BOP - September 2020

Hello,

after the last BoP Update myself together with a team mate did some tests regarding acceleration and fuel consumption with the GTE cars. In order to help improve the BoP, I share the findings in this post. This mainly concerns long distance races like done in VEC, P1LMS and other leagues with several hours of race length.

How did we carry the tests out? (skip if you just want to see the results)


The results:
View attachment 34324

Observations:

So, let's start with the positives:

After the last BoP update the Aston Martin, BMW and Porsche are very well matched to each other.
The acceleration is pretty much identical. Short shifting or changing engine mapping have pretty similar consequences on acceleration and fuel usage.
Fuel tank is identical at 100 litres so refueling time is identical aswell.
I don't see anything that needs to be done to these 3 cars relative to each other!

Now let's look at Ferrari and Corvette:
Both are approx. 0,075 seconds faster than the other cars on the said 500 metres.
For the best acceleration you don't rev out the Ferrari completly, but short shift significantly (see cells marked red). In doing so, the Ferrari is not only faster than the other cars, but also has a lot better fuel range.
About 4% better than AM, BMW, Porsche, that are already slower, and about 5% better than the Corvette, that is equally fast.
On an average track (1:30 min to 2:00 min length in lap time), this means, the Ferrari can extend the stint length by 1 or 2 laps (while being quicker on the straights).
Additionally you can increase the fuel saving a lot (Race lean + shifting at 85% of max. rpm) and still be on par in straight line performance with the other cars (except Corvette).
This can save you easily a pit stop in a long distance race, which equates to 0,75 lap to 1 lap (that is a lot).

Another view to look at it: If the Ferrari is going as fast as it can and you want to reach the same range as the Ferrari you will lose 1,1 to 1,4 tenths in the 500 metre straight (see last 2 columns and cells marked yellow). I gather you have approx. 3 or 4 of these kind of accelerations zones in a lap. This means one loses about 5 tenths per lap with fuel saving compared to the Ferrari (that is a lot aswell).

What about the Corvette? Well, it is basically somewhere between the Ferrari and the other 3 cars. It is as fast as the Ferrari in a straight line while using 5% more fuel. Fuel saving can be done quite well (not as well as with Ferrari, but better than the others).

Due to the smaller fuel tank the Ferrari saves 3 seconds each stop and Corvette 1,5 seconds compared to the other cars.

Conclusion:

As already said, Aston Martin, BMW and Porsche are very well matched. There are slight differences, but I don't think any adjustments are necessary.

The Ferrari has a huge advantage, because of the fuel consumption. It doesn't matter which way you put it, but the longer the race, the greater the advantage the Ferrari has and it is significant. I would expect a further BoP update regarding this issue in the near future. (my advise would be somewhere around +6% with 90l fuel tank)

I think the Corvette should be adjusted in a similar manner, but the changes necessary are smaller compared to the Ferrari. (maybe around 4% with 95l fuel tank. Bit difficult to say as Corvette doesn't use less fuel, but it is faster on the straight and pretty easy to fuel save - maybe better to rob 5 hp or so)



General advise: (this has nothing to do with the rF2 GTE cars in specific)

As far as I know, the formula rFactor uses for calculating the fuel consumption is:
engine speed X filtered throttle position X car-dependent factor

The data suggests this assumption aswell (or at least it is not far off).
The reason Ferrari and Corvette can save fuel that well, lies probably in this formula in combination with a very torquey power curve (lots of power already in the low/middle rev range). You can short shift these cars, which will save you a lot of fuel, while not losing a lot of straight line speed.

In my eyes a better and maybe even simpler way to calculate fuel consumption would be using
engine power output X car-dependent factor

Fuel consumption in kg/h basically is proportional to the power output. The only influencing factor is the engine efficiency, which is dependent on a number of factors, but for a race car, which is mostly driven with full throttle and in a quite narrow engine speed range a simple car-dependent factor is a near enough approximation in my view.

Short shifting the Ferrari or Corvette would then mean, that you still wouldn't lose a lot of straight line speed, but you also wouldn't save as much fuel as is the case at the moment.



Finally a real virutal world example from VEC: There was a Ferrari team doing 38 laps on one stint of fuel at the preseason race at Imola. This was probably done with quite a bit of fuel saving. But several other Ferraris did 37 laps. We ran the Porsche and already had to do a bit of fuel saving to reach 35 laps, which were necessary for not having to do a second fuel stop in the 2h race. The fastest GTE of all in the 3 divisions was a Corvette doing 37 laps with one fuel tank.


(can't upload .xlsx-file to the forum, if developers are interested in the raw data, I can supply the file however. But basically the screenshot shows the important bits anyways)

Hope this post helps in finetuning the GTE class, which since the last update already is on quite a high level regarding BoP in my view - with the exception of the fuel consumption of these 2 cars.

Cheers
Seb
Amazing test, Ferrari as i have heard even in ACC blog is very fuel efficient so it is a good point that it is same here in RF2. This helps thats developers have similar data to work with hence similar cars regardless of simulator.
Corvette Big V8 no big surprise here.
Adjusting their fuel tank capacity as you said should be fine.
 
Amazing test, Ferrari as i have heard even in ACC blog is very fuel efficient so it is a good point that it is same here in RF2. This helps thats developers have similar data to work with hence similar cars regardless of simulator.
Corvette Big V8 no big surprise here.
Adjusting their fuel tank capacity as you said should be fine.
Well, ACC doesn't do GTE cars. Plus I don't think Porsche will be that much worse (don't know if time stamp works, see 08:03:05, if it doesn't):

With that said, fuel economy of GT3 cars is way off in rFactor 2. They use a lot more fuel than GTE cars (at Sebring I used 22% more fuel with GT3 Porsche compared to GTE Porsche), even though they have a bit less power. It also depends quite a bit on BoP. If the car gets a reduced restrictor as a BoP measure, it will use less fuel. So even if the studio gets some numbers from the manufacturers, they very well might not be comparable.
GTE fuel consumption numbers in general are quite realistic, I think.

If one doesn't want to touch the fuel consumption numbers, you can also reduce the fuel tank, yes. I think that was already done to GTE Ferrari in a small update from 93l down to 90l or something like that, but it just isn't enough then. The fuel flow rate during the pit stop has to be reduced aswell then, because at the moment the Ferrari already saves 3 seconds per pit stop, despite the greater range, which doesn't sound that much, but it can be the difference between able to pass another car in the pits or not being able to (or being overtaken). The level in some leagues is that high, that this will make a difference.

I actually think that's the way real world WEC does handle it as part of their BoP measures.

But originally the problem is that short shifting saves too much fuel in cars with a torquey power curve (like Ferrari and Corvette). If that can't be fixed easily (which I think is the case in rF2), just increasing the fuel consumption somewhere in the range I posted in my first post, would be the second best (but easier and quicker to implement) measure, in my view.

Going to tag the devs here @Christopher Elliott

And I do want to repeat, that BoP in GTE is very good now. There isn't a dominant car anymore like BMW was half a year ago with almost 0,5 sec advantage in pace. ;-) This is complaining on a high level now or let's rather say providing data for finetuning the cars BoP regarding fuel range.
 
just increasing the fuel consumption somewhere in the range I posted in my first post, would be the second best (but easier and quicker to implement) measure, in my view.
How would you do this without changing car physics?
Are you willing to play with artificial physics just to have close racing?
 
You keep the dynamic tire physics... and you tweak other part that are less dynamic, pre-baked, spreadsheets, torque curves, aerodynamic.... etc.
Especially here if we talk about fuel consumption... pretty sure it doesn't involve a revolutionary model but merely what you would call 'artifical physics'...

Every realtime software has to have more or less 'artificial' physics, 'artificial' rendering or even simplyied logic... it's what game craft has been all about...
Visual or physics... it's basically the same concept... a mix of pre-baked and more dynamic systems... only recently we start to have more and more physical based rendering like raytracing in realtime... but even then you won't have fully raytraced game until 5years from now.
 
Rf1 had a rather simple fuel calculation. 100% Throttle x Time x trackfuelusage variable. So tracks like Le Mans had way to much fuel consumption due to the cars being at max throttle for so much of the lap. I haven't even dipped my toe into rF2 fizzies, but I'd bet they have a few more parameters in there now.
 
@Slip_Angel I dont think, a discussion about this is necessary, with the whole GT3 class having a way to high fuel consumption compared to real live or - what might be worse tbh - the GT3 Cars cars being as fast as GTE cars on some tracks.

In my first post I even stated how the modelling of fuel consumption can be improved (it's not perfect then, but you won't get accurate enough data from manufacturers anyway), but with how rF2 handles the calculation right now, I have no doubt that we would get closer to real world numbers on track with just increasing the fuel consumption.
 
Don't get me wrong, fair BoP is good, even though i barely play races.

My only point is that I would hate to have a good BoP at the cost of messing with official data thats all.
if this is not the case then its all fine.
For me i want the virtual cars HAS to match real life cars in terms of driving,physics etc i honestly don't care if it couple seconds slower.
If this is how it is IRL then i will be much happier than say making it artificially faster by changing the stuff that you can't change via BoP.
 
It would be a goo option, to have the BoP as upgrade (league edition)
Then we could choose between both options.
 
Last edited:
IMO, RF2 just need some simple slider for BOP options (such as intake restriction, weight, or even drag), which server admin can adjust/load easily on server/client side. This is one of the best feature I missed from live for speed.
 
IMO, RF2 just need some simple slider for BOP options (such as intake restriction, weight, or even drag), which server admin can adjust/load easily on server/client side. This is one of the best feature I missed from live for speed.
intake and weight would be enough. But for sure a slider from -5% to +5% would be great, dealing with upgrades is a pain
 
IMO, RF2 just need some simple slider for BOP options (such as intake restriction, weight, or even drag), which server admin can adjust/load easily on server/client side. This is one of the best feature I missed from live for speed.

Simple in an idea but no doubt very much harder to actually implement, nice thought though :)
 
Will there be some sort of feedback from the devs regarding the fuel consumption numbers and how this topic will be handled?

Otherwise probably the league organizers will have to have a look into this everyone on their own.
 
@Christopher Elliott

So .. it's been 2 years and despite numerous updates/BoP, these GTE cars still have weird torque cut/TC in 1st gear.

Even with autoclutch off, TC off, if you release the clutch in 1st gear while reving to redline - the cars just bog down and refuse to spin the tires. These are 500+ hp RWD GTE cars. They should easily be able to spin the rear tires with TC off.

What makes me think it's some weird torque cut logic you guys have programmed in is that you can't spin the wheels in a straight line.
But slow the car down to just above 30mph, crank the wheel full lock, and THEN dump the clutch while redlining it in 1st gear, and all of a sudden - boom, the rear tires will break traction and you can do a 360.
But try to do the exact same thing with the steering wheel completely straight, and the rear wheels just die like i'm driving a 80hp Suzuki Samurai from 1986. It's really ridiculous.

Apart from the obvious simulation aspect (as in these are real race cars that we are trying to re-create), there is a safety issue as well. right now, if you spin in a tight corner/apex, you can't just dump the clutch and spin the car around safely. you have to do a 15 point turn while to get the car turned around. It is incredibly dangerous and not all realistic. Just look at De Phillip last night at 12 hrs of sebring after his M8 got punted in turn 3/4. He was facing backwards. Looks like he hit the TC override button, and dumped the clutch and spun the car around and was able to re-join safely.

Just search for any of the official factory IMSA twitter accounts, and you can see they all easily spin the tires leaving the pitbox. That's while under pit speed limiter!

I'll just leave this here:
https://twitter.com/jameypricephoto/status/1327968566507950081
Try to do that in rf2 right now and you look like a newb 15 yr old learning how to drive a stick for the firs time. The car just bogs and dies.

Is this s397's way of implementing the WEC rules about not spinning tires in pit lane? Because if so, it's a horrible implementation and quite frankly a piss poor way of doing it.
 
@Christopher Elliott

So .. it's been 2 years and despite numerous updates/BoP, these GTE cars still have weird torque cut/TC in 1st gear.

Even with autoclutch off, TC off, if you release the clutch in 1st gear while reving to redline - the cars just bog down and refuse to spin the tires. These are 500+ hp RWD GTE cars. They should easily be able to spin the rear tires with TC off.

What makes me think it's some weird torque cut logic you guys have programmed in is that you can't spin the wheels in a straight line.
But slow the car down to just above 30mph, crank the wheel full lock, and THEN dump the clutch while redlining it in 1st gear, and all of a sudden - boom, the rear tires will break traction and you can do a 360.
But try to do the exact same thing with the steering wheel completely straight, and the rear wheels just die like i'm driving a 80hp Suzuki Samurai from 1986. It's really ridiculous.

Apart from the obvious simulation aspect (as in these are real race cars that we are trying to re-create), there is a safety issue as well. right now, if you spin in a tight corner/apex, you can't just dump the clutch and spin the car around safely. you have to do a 15 point turn while to get the car turned around. It is incredibly dangerous and not all realistic. Just look at De Phillip last night at 12 hrs of sebring after his M8 got punted in turn 3/4. He was facing backwards. Looks like he hit the TC override button, and dumped the clutch and spun the car around and was able to re-join safely.

Just search for any of the official factory IMSA twitter accounts, and you can see they all easily spin the tires leaving the pitbox. That's while under pit speed limiter!

I'll just leave this here:
https://twitter.com/jameypricephoto/status/1327968566507950081
Try to do that in rf2 right now and you look like a newb 15 yr old learning how to drive a stick for the firs time. The car just bogs and dies.

Is this s397's way of implementing the WEC rules about not spinning tires in pit lane? Because if so, it's a horrible implementation and quite frankly a piss poor way of doing it.
I don't have your problem. The cars will wheel spin/burnout easily.

Make sure you've turned off both the car setup TC & the assists TC
 
I don't have your problem. The cars will wheel spin/burnout easily.

Make sure you've turned off both the car setup TC & the assists TC
I'm not a newb. I've been driving isi based sims since GTR1.
No, i don't have TC assist on and NO in car TC is not on.

And despite people claiming the same thing as you, no one has ever posted a video of them launching s397 GTE despite multiple people asking for same thing in the last 2 years. Do a launch in the s397 911 RSR, and record the replay (the replay will have the pedal position/movement).

If it was my pedal setup, then it would be in all mods. I can launch the car perfectly in EVERY SINGLE other mod just fine. The closes thing we have to s397 GTE cars are the URD cars. And i can launch that/spin the rear tires 100 times out of 100 times. The ONLY mod i've never been able to launch is s397. It's like there is something weird going on with the clutch engagement in 1st gear.
 
@Morbo66 i agree with you. I have observed the same thing. It is possible to do that in ferrari GTE though BUT 911 RSR completely bogs down in standstill acceleration.
And i have seen IRL RSR do a very big/long standstill burnout on goodwood festival of speed.
 
Apart from the obvious simulation aspect (as in these are real race cars that we are trying to re-create), there is a safety issue as well. right now, if you spin in a tight corner/apex, you can't just dump the clutch and spin the car around safely. you have to do a 15 point turn while to get the car turned around.

It is possible to spin turn the 911 GTE from a standing start, but you can't dump the clutch. You need to feed it in a bit and hold it there, and you can spin the rears up and turn around.

Producing skidmarks from a straight standing start eludes me though, at least in a couple of mins of testing.

This seems to point to the much referenced 'engine inertia bug', moreso than the rev up/down speed (which could be a number of factors). The grip of the rear tyres just completely kills the engine rotation, unless you keep slipping the clutch.
 
It is possible to spin turn the 911 GTE from a standing start, but you can't dump the clutch. You need to feed it in a bit and hold it there, and you can spin the rears up and turn around.

Producing skidmarks from a straight standing start eludes me though, at least in a couple of mins of testing.

This seems to point to the much referenced 'engine inertia bug', moreso than the rev up/down speed (which could be a number of factors). The grip of the rear tyres just completely kills the engine rotation, unless you keep slipping the clutch.

Yes, I've read about the engine inertia bug on previous GTE threads. But if that was the only issue, then why haven't they fixed it yet? Does s397 do any UAT at all? Seems like their testers don't actually know how to do proper negative testing.

As for the spin turn - I know what you are referring to, but that's not a flat spin turn. As I mentioned above, you have to be moving above about 30~40mph AND have the wheel cranked to get it to break traction. That doesn't help if you're stuck facing backwards at a tight apex like hairpin in Long Beach , or La Source hairpin at Spa.

The fact that it will break traction IF you crank the wheel while at speeds above 40mph makes me think it might also be partially related to the locked power diff. Like there are more than one issue at play here besides engine inertia bug.

It's obvious they have the power diff value in the 911 locked at some stupidly ridiculous low setting for BOP purposes. It feels like power diff is locked at like some really low unrealistic value like 5%. You can feel it with the default setup, the car understeers like the Titanic on power exiting medium speed corners where aero/downforce is not enough to help the nose rotate. You can dial it out with setup but it's very obvious they tried to nerf the car with the weird power diff value.
 
As for the spin turn - I know what you are referring to, but that's not a flat spin turn. As I mentioned above, you have to be moving above about 30~40mph AND have the wheel cranked to get it to break traction.

No, I mean from a standing start. Try it, it is possible.
 
Back
Top