Wednesday, 12 August 2026

Are the scrutineers picking on my driver? - Update after the Belgian Grand Prix

Well I don't know if they're picking on *your* driver but they are picking on mine. Bar chart showing how often cars were scrutinized compared to the expected number at the Belgian Grand Prix. The values are displayed as a standardised residual. Car 16 (Leclerc) was the most over-scrutinized compared to expected, with a value of 0.83, followed by cars 1 (Norris) and 81 (Piastri) on 0.50, then 12 (Antonelli), 30 (Lawson) and 55 (Sainz)on 0.33. Car 11 (Perez) is the most under-scrutinized compared to expected with a value of -0.67, followed by cars 18 (Stroll) and 63 (Russell) with a value of -0.50, then a whole bunch of cars on -0.33. One car has been tested exactly as much as expected, number 23 (Albon). Look how over-tested Leclerc is compared to the others! 

I am being a little ridiculous, but Leclerc's standardised residual of 0.83 over expected is the third highest single race value after Norris in Barcelona (+1.30), and Lawson at the Austrian Grand Prix (+1.06). 

Entirely as expected, Perez, Stroll and Russell who didn't finish are 3 of the four least tested compared to expected. 

The difference isn't statistically significant (χ² = 3.201), but this time, it's my driver. 

The Belgian Grand Prix checks were slightly different, and there were more of them than usual (36, which is the second highest of the season, the highest being at the Australian Grand Prix which had 37 checks as the expected number but that's also the first Grand Prix of the season). 

The season to date: Bar chart showing how often cars were scrutinized compared to the expected number for the season up to the Belgian Grand Prix. The values are displayed as a standardised residual. Car 1 (Norris) is the most over-scrutinized compared to expected, with a value of 0.61, followed by cars 5 (Bortoleto) and 44 (Hamilton), with a value of 0.55. Car 18 (Stroll) is the most under-scrutinized compared to expected with a value of -0.77, followed by cars 14 (Alonso) and 77 (Bottas) with a value of -0.55. No cars have been tested exactly as much as expected. 

The difference are not statistically significant, with an χ² = 3.746. 

Interestingly, the χ² is increasing, and has been for the last 4 races, when, at this point in the season, you'd be expecting it to start going down. 

Possibly this is because of changes in reliability as the teams improve their cars (or try to) during the season. The Aston Martins, Cadillacs, Williams and Audis keep being unreliable and not finishing, while the McLarens seem to have sorted themselves out so are being checked more often and that's disturbing the numbers. 

Norris is now the most over-checked car compared to expected. He's followed by Bortoleto, Hamilton and Russell. One of these drivers is not like the others. 

Not sure why Bortoleto is that high given he's 14th in the standings. On the other hand, he has one of the lowest DNF rates (having only failed to finish 1 race out of the 10 so far). 

Stroll, Alonso, Bottas, Albon and Hulkenberg are the least tested compared to expected. That makes sense given the number of DNFs between them. They are also the bottom 5 in the championship. 

I find it amusing the Audi have one driver in the most over-tested bucket and one in the most undertested. 

No drivers have been tested exactly as many times as expected, but Colapinto, Lindblad and Piastri are almost on 0, being at +0.056 or -0.056. 

The Hungarian Grand Prix results will be the midway point of the season. We'll see if the χ² starts to come back down.
Posted by at

No comments:

Post a Comment

Subscribe to: Post Comments (Atom)