2024 Italian Grand Prix - Monza, Aug 30 - Sep 1

[venkyhere]

About F1 driver consistency... what also seems to happen on average more than once per race (I haven't actually averaged it), is drivers setting the exact same lap time in two subsequent laps.

For example, this race:
Sainz 1:23.503 on Lap 26 and 27
Stroll 1:21.721 on Lap 21 and 22

That's just insane to me. I stumbled over this once when I just made the assumption that this rarely happens, because I thought the odds would be very low. Turns out, they aren't.

20 drivers, about 50 laps... that's a 1000 chances for it to happen. And we only have 3 decimals.

How ? Assume we are talking about 1.ab.xyz as the laptime. Assume that all 20 drivers are equally capable of driving strictly within 1.ab.000 to 1.ab.999, for each of their 50 laps.

Given a particular xyz combination time for a lap :
The probability for ANY driver to get ANY two laps with exact same laptime is 20 x (1/1000) x 49/50 = 1.96%
The probability for ANY driver to get CONSECUTIVE laps with exact same laptime is 20 x (1/1000) x 1/49 = 0.041%

So those 3 decimals are plenty enough to make the chances for such a thing, really low, mathematically speaking.

Now, in the real world, track temp, tyre temp, air temp, grip level, fuel load, wind level/direction etc are likely to remain almost same for two consecutive laps, so the chances that the driver inputs remain almost exactly same, are not that low. Hence, because of all these 'physical' conditions helping, consecutive laps producing the same laptime is more likely than ANY two laps with same laptime.

The point I want to make is, such a thing being not that unlikely, is not because of 'we only have 3 decimals'.

[woocasz]

Pretty average for a F1 driver to be fair. Just a case of driving to a set delta on the steering wheels and the times will be there. Max did the same in Mexico 2022. Impressive stuff, but for the level of these drivers - pretty easy.

you said Mexico 22.
https://i.postimg.cc/W3SB62Ck/screenshot-360-1.jpg

when you have a perfect car, most dominant in history of F1, he is a genius...
(lets see that kind of performance from Max in the current RB20...hahaha)

What Charles has done is more impressive, considering the kind of car he drives. Don't forget that at the weekend Carlos said the car had too much understeer (if you have been following F1 long enough you know that Sainz prefers understeer) so imagine what Leclerc had to deal with.


Qualifying showed how understeery the car is:
Q1 +4
Q2 +2
Q3 +2

+8 "clicks" on the front wing from Q1 to Q3. That's what Charles asked for in qualifying

I mean Max's 2nd stint on the Hard tyre in Monza was much more consistent than that of Charles.....

However, different strategy, so hard to say how long that Max would have been able to maintain that sort of time on the hard tyre.

Max drove within 3-4 tenths for 40 laps on the hard tyre in Zandvoort. Same with a couple of other drivers. Its just normal for top tier drivers.

Did you just compare 17laps vs 38 laps stint? lol
Besides fighting for victory and chasing by the fastest car on track vs cruising in 6th position is different story.

[hollus]

I think there are some errors there, either in that math or in the assumptions.
I can’t be bothered with the math right now (long work day over), so I’ll offer a different approach to the problem:
After any lap from any driver, there is a high probability that the next lap will be +\- half a second from that time, whatever that time was.
If we accept that, then there is about 1 chance in 1000 of the next lap nailing those 3 decimals.
So 1000 events with a chance of 1/1000 each, and your expectation will be for it to happen about once per race.

Of course the plus-minus half a second assumption is questionable, but arguably more likely to match in reality than with the random time in the plus-minus half a second window, as often times just tumble by 1 or 2 tenths lap to lap.


See for example here. Lap to lap times have a scatter of much less than half a second
https://www.reddit.com/r/formula1/comme ... /#lightbox

[dialtone]

I think there are some errors there, either in that math or in the assumptions.
I can’t be bothered with the math right now (long work day over), so I’ll offer a different approach to the problem:
After any lap from any driver, there is a high probability that the next lap will be +\- half a second from that time, whatever that time was.
If we accept that, then there is about 1 chance in 1000 of the next lap nailing those 3 decimals.
So 1000 events with a chance of 1/1000 each, and your expectation will be for it to happen about once per race.

Of course the plus-minus half a second assumption is questionable, but arguably more likely to match in reality than with the random time in the plus-minus half a second window, as often times just tumble by 1 or 2 tenths lap to lap.
[snip]

That formula would really only work if the laptime is random and uniformly distributed but it really isn't, the driver is specifically acting to keep the same time and receives feedback on the wheel for it. The probability is much higher than 1 in 1000 to hit the same time, I'd be surprised if it even is +/- 0.5s and not +/- 0.2s or even less, especially for adjacent laps.

[AR3-GP]

Is it possible that because the average speed is so high in Monza, that this makes it easier? After all, 1 tenth of a second margin on a track with an average speed of 160mph (~71 m/s), is equivalent to 7 meters of variance across the finish line. That's a very large distance to fall within the same tenth. It's the highest of the season if you go by the stats below:

Fastest 2020 F1 tracks
Race Circuit Average Speed Race Lap Record
Italian Grand Prix Monza 159.892mph 1m 21.046s
British Grand Prix Silverstone 150.808mph 1m 27.369s
Belgian Grand Prix Spa-Francorchamps 147.406mph 1m 46.286s
Austrian Grand Prix Red Bull Ring 144.253mph 1m 06.957s
Japanese Grand Prix Suzuka 142.760mph 1m 30.983s
Australian Grand Prix Albert Park Circuit 141.004mph 1m 24.125s
French Grand Prix Circuit Paul Ricard 140.910mph 1m 32.740s
Brazilian Grand Prix Interlagos 136.620mph 1m 10.540s
Russian Grand Prix Sochi Autodrom 136.615mph 1m 35.761s
Bahrain Grand Prix Bahrain International Circuit 134.144mph 1m 30.252s
Canadian Grand Prix Circuit Gilles Villeneuve 133.451mph 1m 13.078s
Spanish Grand Prix Circuit de Catalunya-Barcelona 132.726mph 1m 18.441s
Chinese Grand Prix Shanghai International Circuit 132.192mph 1m 32.238s
Azerbaijan Grand Prix Baku City Circuit 130.357mph 1m 43.009s
United States Grand Prix Circuit of the Americas 128.249mph 1m 36.169s
Hungarian Grand Prix Hungaroring 127.092mph 1m 17.103s
Abu Dhabi Grand Prix Yas Marina Circuit 125.133mph 1m 39.283s
Mexican Grand Prix Autodromo Hermanos Rodriguez 122.253mph 1m 18.741s
Singapore Grand Prix Marina Bay Street Circuit 111.138mph 1m 41.905s
Monaco Grand Prix Circuit de Monaco 100.544mph 1m 14.260s

[PlatinumZealot]

It is not intuitive that something like this should happen. The car gets lighter with each lap (which should cause a natural laptime reduction), however some of that may balance with the lap by lap tire degradation.

Drivers setting constant laptimes is something strange that should be studied further. Even if a driver set out to maintain a target, it shouldn't be possible given that the fuel reduces. I would be interested in knowing how the driver compensates for this or if they are aware that they are compensating for it. You would have to consciously lift to slow down.

Hamiltons pace is more intuitive. He gets faster as the fuel load burns off. Even if you did not change your braking points, the mass will simply accelerate and decelerate more quickly because it is lighter with less fuel.

https://i.postimg.cc/3Jzc6ZQq/image.png

You're overthinking it though. In this formula you treat the tyre as a resource to be managed. Management of any resource is enhanced when the environment is stable. (Stable inflation, stable exchange rate for example). The conditions in the race vary - a lot. The race engineer will use past data to prescribe a target lap time that keeps tyre degradation stable. Usually from FP2. They simply cannot predict what the tyre is going to do without a solid data basis. They tell the driver, we know the resource and we know how long it takes to be used under these stable conditions; target this lap time and your tyres will last X-laps. They will then update the driver to lower the target based on previous laps data or form other cars. Even as fuel loads burn off the drivers in this formula tend to get very stable lap-times with the odd bursts in speed to undercut, overtake, defend etc. And they have the target delta on their steering wheels.

[Waz]

They're also multiple seconds slower than the ultimate pace of the car. Fuel weight isn't having the same effect.