David Greco's Handling, Driving and Setup Q&A (?)

@TPMillo It seems there wouldn't be much point in changing it this year, now that it's siginicantly less powerful irl as well.

Back in 2019, the slipstream reduced drag by 13%, but everyone complained that it was too powerful, so it was changed to 10%. This is a consequence of not have powerful enough dirty air, so as long as you got in drs range, it was basically impossible to defend.

We're seeing something similar irl this year, the drs is very powerful, and the dirty air is much more manageable, so it's very hard to defend against drivers behind, unless you have a significant straight line speed advantage. To add on to this, the cars in-game don't typically have the same disparity in straight line speed as irl, with the engines being identical (aside from Merc), the only real difference is drag, they even get the same scaling for low, medium, and very high downforce configurations. Some of that seems to be fixed this year with the 1-50 wing system, I assume they'll just specify a downforce and drag range across the entire thing instead of scaling them.

On a tangent, the drs was changed from 12% to 15% in 2021, 12% actually being very close to real life, but I assume it was to compensate for the weak slipstream. The problem is the engines had too much power, each having 860 hp, and Merc 880 hp. For comparison, Red Bull and Mercedes were estimated to have between 850-855 hp, with Ferrari and Renault 15-20 hp behind. To compensate the cars were made very draggy. The problem then comes with drs, not only is it more powerful, but it's also reducing a higher base value, meaning it was far too powerful. Hopefully the engines are a bit more accurate this year.

If you play singleplayer there's actually another problem. The slipstream and dirty air is halved (or at least significaly reduced) for the ai. This means the slipstream is even more pointless, but also means they lose less downforce when following. I really can't think of a good reason for this, in my testing I found that the ai can handle higher dirty air values, so it almost seems like an accident. It means the ai is more reliant on drs than the player, but also that they can follow a bit easier through corners. Really hope this is changed, I don't see any reason for it to be this way.

Another problem is overtake, or rather how deployment works. As speed increases, deployment will start reducing, I assume in an effort simulate real deployment patterns without having to map it all out. Unfortuately it means overtake gets less effective at higher speeds, being reduced by 10% at 306 kph, 25% at 324 kph, 40% at 342 kph, and 75% at 375 (in-game theoretical top speed). It might not sound like much, but losing 40 hp at 324 kph is ridiculous. I hope this is changed.

Sorry about the long post, kinda lost focus. ('´ ˘ `)

@ActuallyImJulie could you expand on the ERS bit a little?

I thought Overtake mode in the game gave you the full might of the MGU-K's 160hp (164? 180?) regardless of wheel or engine speed.

Even thought it was cool that the game is represented the regs in disabling the MGU-K's output when below 100kph or thereabouts.

@mariohomoh It's actually fairly simple.

This tells the game what the maximum amount of deployment in Watts is (120000 is equivalent to 160 hp).

Then the mode you're in simply deploys a percentage of that number.

Then the game looks how fast you're going (in M/S), and deploys a percentage of your percentage. This is why Overtake is a problem, it exists under the same SpeedDeployLine as the other modes, it's not a seperate thing, even though it might look like it in-game.

Not sure what these percentages represent, I'm assuming loss of downforce/drag. Anyway, isn't 24-25% significantly more than 10% (potentially 5%) @Meza994?

And @ActuallyImJulie thanks for the ERS breakdown! Can't think of a reason for that.

@mariohomohYes downforce loss, base game downforce loss is 15% and 24% was the value i settled on to not have regular AI making big mistakes like crashing in the kinks on the straight in Baku (yes from 25% onwards they go wide when too close and just crash into the barrier).

Yes 24% is more than 15% but looking at IRL 47% last years its still not too significant IMO.. Especially since dirty air was always only starting at about 1s gap and by the way dirty air develops it would only reach the 24% downforce loss with 2-3 car lengths behind which you almost never are in the corners so the real difference at about 0.5s gap was more like ~4% less downforce if even that (in my mod)

Regarding ERS i think it was Codemasters way of doing a deployment mapping. So the MGU-K wouldnt deploy as much power, just like it does IRL, yes i know its not based on the speed they go at but on GPS but Codemasters probably didnt wanna go through the trouble of creating a ERS map for each track with a lot of points for deployment adjustment. Although that would be great for the sim guys as a setup option

---

@TPMillo wrote:
@DavidG53 Hi David, thanks for the super interesting comments.

Have there been any changes to slipstream in the new game? It feels quite weak in game compared to what we see in the real life data (the crazy Baku straight slip stream from super far infront etc) whereas the game feels like very little benefit unless you're very close to the car infront?

---

From the videos? You have to think the AI deploys the whole overtake a few tenths before you even get alongside them. Can't really judge against AI. It will need to be players vs player and then maybe a patch, if possible. Remember AI need to be retrained every time, and it is not a day or 2. is a very long process. Is a lot of data they have to check against.

---

@Meza994 wrote:
@ActuallyImJulie Where was it confirmed by devs that for 2021 slipstream/dirty air is halved for AI?

Besides no the AI can not deal with really higher dirty air too well, limit is 24-25% from then on high speed corners like in Silverstone become a problem which is still only almost half of the real life dirty air besides the length of dirty air which was completely wrong..

Soo David if you do read this by chance, has the length of dirty air changed or is it still only ~1 second?

---

Halved? not is not halved, same as player.

Should be around 1.5 seconds.

---

@Meza994 wrote:
@mariohomoh As David revealed, AI is same dirty air as for us, so point would be right but just doesnt fit the situation :D

You dont wanna know how overtake mode was pre 2020.. didnt even come close to 100% of MGU-K sooo yeah they werent close to real life but with how they were interacting with the modes of MGU and ICE its understandable Codemasters choose an easier approach, especially as details of said modes were not too public knowledge

---

Yes I know ERS is not the best implementation, but we focused on more important aspects of the physics. I know players want everything, but I can assure that, time is never enough. As I explained a few days ago, at a point in the project I must stop developing, so that the AI can train and the whole game can go forward (for example, practice programs too).

With patches we can adjust a few things, but the ERS would be an important thing to change. I read a few answers before about having a separate speed deploy for overtake. That is a really good idea, thanks.

---

@mariohomoh wrote:
@DavidG53 since you're so kind, does camber and toe impact on tyre drag?

Will the different aero components (FW, RW, diffuser) have different lift-to-drag ratios?

---

the impact is very little. Should be more I think, but needs a further big improvement that has been on my list for quite some time.

Yes, each aero part of the car have their own efficiency. And is not a simple linear model as it was before. Rear wing is the least efficient, As you increase the rear wing, you are decreasing the efficiency, so drag to lift ratio gets bigger. Front wing has actually a sweet spot that isn't wing 0 nor is 50. And this was possible by not making it just a simple linear model.

@DavidG53Mr. Greco don't listen to him, man! He just asked for porpoising not long ago! Yes, we'd all love more control over ERS but not if it risks going back to that foolish state of flicking through different modes to every corner.

Now if you're still listening, have tons of free time on your hands and is looking for the next big project to work on in these games, can we pleeease get Inside/Middle/Outside (IMO) temperature readings for tyre surface and make camber all the more important to tyre wear and cornering? 😥

---

@mariohomoh wrote:

@DavidG53Mr. Greco don't listen to him, man! He just asked for porpoising not long ago! Yes, we'd all love more control over ERS but not if it risks going back to that foolish state of flicking through different modes to every corner.

Now if you're still listening, have tons of free time on your hands and is looking for the next big project to work on in these games, can we pleeease get Inside/Middle/Outside (IMO) temperature readings for tyre surface and make camber all the more important to tyre wear and cornering? 😥

---

This falls into the same category of things to improve when I answered your previous question about camber/toe vs tyre drag. 

@mariohomoh also please I am not so old, just call me David 🙂

@DavidG53  thank you for all your participation, each your contribution here is very valuable for the community that appreciates every aspect and knowledge of the game and wants to contribute to the improvement of the game with its ideas 🙂

Right, after retesting, I was indeed wrong about the dirty air and slipstream being halved, I did my original testing all the way back in 2020, and must have used a bad methodology then. Sorry about that.

I'm glad to hear Overtake might be looked at, it would be a good start toward improving the ers a bit, as that alone would at least let modders play with ers a bit more.

Different aero balance for the cars is very good to hear, the game is definitely moving in a direction I'm a fan of, I might disagree with a few choices, but it all sounds very good so far.

@DavidG53 If you still have time, is there any particular reason the gear ratios don't match very closely to the real cars? As an example this is the gear ratios for all cars in-game.

Whereas this is what Ferrari were actually using.

Or Alpine with a very unique gear setup.

It doesn't really impact anything aside from the theoretical top speed, but it does change the sound of the cars a bit, especially Alpine.

Does it just not matter enough to spend time on?

<GearRatio1st value="2.95" />
<GearRatio2nd value="2.29" />
<GearRatio3rd value="1.93" />
<GearRatio4th value="1.59" />
<GearRatio5th value="1.375" />
<GearRatio6th value="1.19" />
<GearRatio7th value="1.05" />
<GearRatio8th value="0.93" />

Those are the mode values for all the teams' different ratios, so it would closely represent the real ratios if the point was simplicity.

As for this years cars, I've estimated the current gear ratios.

Red Bull
    <GearRatio1st value="2.95" />
    <GearRatio2nd value="2.47" />
    <GearRatio3rd value="1.975" />
    <GearRatio4th value="1.66" />
    <GearRatio5th value="1.4125" />
    <GearRatio6th value="1.2525" />
    <GearRatio7th value="1.125" />
    <GearRatio8th value="0.9925" />

Mercedes
    <GearRatio1st value="3.1" />
    <GearRatio2nd value="2.47" />
    <GearRatio3rd value="1.975" />
    <GearRatio4th value="1.745" />
    <GearRatio5th value="1.475" />
    <GearRatio6th value="1.305" />
    <GearRatio7th value="1.145" />
    <GearRatio8th value="1.0075" />

Mclaren
    <GearRatio1st value="3.175" />
    <GearRatio2nd value="2.47" />
    <GearRatio3rd value="2.02" />
    <GearRatio4th value="1.705" />
    <GearRatio5th value="1.475" />
    <GearRatio6th value="1.255" />
    <GearRatio7th value="1.1095" />
    <GearRatio8th value="0.9775" />

Ferrari
    <GearRatio1st value="3.1" />
    <GearRatio2nd value="2.47" />
    <GearRatio3rd value="2.075" />
    <GearRatio4th value="1.745" />
    <GearRatio5th value="1.475" />
    <GearRatio6th value="1.28" />
    <GearRatio7th value="1.1095" />
    <GearRatio8th value="0.985" />

Alpha Tauri
    <GearRatio1st value="3.1" />
    <GearRatio2nd value="2.47" />
    <GearRatio3rd value="2.02" />
    <GearRatio4th value="1.685" />
    <GearRatio5th value="1.435" />
    <GearRatio6th value="1.23" />
    <GearRatio7th value="1.1095" />
    <GearRatio8th value="0.9925" />

Alpine
    <GearRatio1st value="3.175" />
    <GearRatio2nd value="2.47" />
    <GearRatio3rd value="1.975" />
    <GearRatio4th value="1.645" />
    <GearRatio5th value="1.385" />
    <GearRatio6th value="1.21" />
    <GearRatio7th value="1.056" />
    <GearRatio8th value="0.9625" />

Aston Martin
    <GearRatio1st value="3.175" />
    <GearRatio2nd value="2.37" />
    <GearRatio3rd value="1.975" />
    <GearRatio4th value="1.725" />
    <GearRatio5th value="1.51" />
    <GearRatio6th value="1.305" />
    <GearRatio7th value="1.155" />
    <GearRatio8th value="1.0225" />

Alfa Romeo
    <GearRatio1st value="3.1" />
    <GearRatio2nd value="2.47" />
    <GearRatio3rd value="2.075" />
    <GearRatio4th value="1.745" />
    <GearRatio5th value="1.475" />
    <GearRatio6th value="1.28" />
    <GearRatio7th value="1.1095" />
    <GearRatio8th value="0.985" />

Williams
    <GearRatio1st value="3.175" />
    <GearRatio2nd value="2.325" />
    <GearRatio3rd value="1.975" />
    <GearRatio4th value="1.725" />
    <GearRatio5th value="1.475" />
    <GearRatio6th value="1.305" />
    <GearRatio7th value="1.165" />
    <GearRatio8th value="1.0225" />

Haaaaas
    <GearRatio1st value="3.1" />
    <GearRatio2nd value="2.47" />
    <GearRatio3rd value="2.075" />
    <GearRatio4th value="1.725" />
    <GearRatio5th value="1.51" />
    <GearRatio6th value="1.28" />
    <GearRatio7th value="1.1095" />
    <GearRatio8th value="0.985" />

As for general purpose values for simplicity, it could be something like this.

    <GearRatio1st value="3.1" />
    <GearRatio2nd value="2.47" />
    <GearRatio3rd value="1.975" />
    <GearRatio4th value="1.705" />
    <GearRatio5th value="1.475" />
    <GearRatio6th value="1.28" />
    <GearRatio7th value="1.1095" />
    <GearRatio8th value="0.985" />

I know I'm the only weirdo who actually cares, but it's bothered me for years.

Sorry again for the essay. ('´ ˘ `)