Indian Grand Prix: The Story from the Data

Posted on October 30, 2011


While the Indian Grand Prix was not the most action packed race of the year, it gave a whole new challenge in data analysis. Or a whole old challenge. That the tyres were lasting well was clear to see, but as the stints got longer it became clear that the degradation was essentially zero and that there was a clear trend of laptimes reducing during a stint. The most clear example of this came from one Michael Schumacher, who enjoyed a huge amount of success (in his first career) on Bridgestone tyres which did not degrade – but more of that later.

The race at the front was fairly straightforward. The traces from the top three teams (Hamilton is shown up to his contretemps with Massa) are in the chart below. As usual, the dashed lines are the intelligentF1 model fits of the races. The fits jump at the pitstops as there is a fixed pitstop time penalty in the model – and of course pitstops take a varying amount of time. The key thing to note on the chart is that the lines are curved – and obviously so. The intelligentF1 model fits are usually slightly curved (a straight line means that the loss in laptime from degradation is equal to the gain from the fuel load lightening) – but it is often difficult to tell. Here it is clear as the degradation is so small.

The underlying pace in the model is the same for all stints on the same type of tyre, so the model has done well in this low-degradation scenario – the fits tell the story well. The suggestion is that the underlying pace difference between these cars is very small – about 0.1s covers all the cars, although Button is slower (by 0.2s) on the hard tyres at the end of the race. After lap 5 the gap between the first two changed very little for the rest of the race.  The reason that Webber worked his was backwards can now be seen. Firstly, the cars ahead got a little more from their tyres in the last laps before the first stops, and secondly his inlap and outlap were slow at his second stop. The comparative inlap from Alonso was 1.2s quicker, and the outlap 0.8s faster than the Australian. The laptimes on the new hard tyres were marginally faster than the used softs, so all things being equal Alonso should not have made it past. However, the laptimes on the used tyres were getting faster, so staying out the extra 10 laps or so resulted in a big gain for the first two. Either Webber had got himself into some tyre trouble, or Red Bull got this one wrong.

There was an intermediate race between the front and the midfield containing the Mercedes, the Toro Rossos (once they had cleared Senna) and Hamilton (post-Massa). The intelligentF1 model fits for the Mercedes and Alguersuari are shown and the low tyre degradation is again clear.

The thing to note here is that the increasing upward curve of Schumacher’s trace (grey without crosses) goes well beyond that predicted by the model, whereas Rosberg’s trace matches well. With the degradation essentially removed in the model, the only way this can be is that fuel-corrected Schumacher is going faster at the end of this stint than he is at the beginning. The statement from his race engineer at the end of the second stint that his pace was good sounded a little incredulous at the time. I’m pretty sure that Mercedes have no idea how he was going quite so fast at this point of the race – he gained about 8s over what should have been possible, and it gained him a place on his team-mate. Given that this car ate its tyres more quickly than any other car in Singpore, Mercedes have done wonders with their tyre usage. Toro Rosso are not a long way ahead of the midfield, but they were clearly the fifth fastest car in the race. A quick note on Hamilton – he was quicker than the Mercedes in underlying pace, although only by about 0.6s, but if anything he was slower on the hard tyre, and a lot slower than Button.

The midfield battle was spiced up by three cars taking the safety car gamble by starting on hard tyres. The Di Resta/Perez/Petrov train can be seen in the first stint, with the Scot holding up the cars behind until making an earlier than planned stop. The Force India did not have the longevity on the soft tyres that the other teams had, so Sutil was swapped to a long final stint on the hard tyres. It kept him ahead of the gamblers, but the opportunity was still there to them to beat him. Interestingly, Di Resta’s pace in the first stint is poor in comparison with his other stints on the soft tyre, and overall the Renault did not seem to have the pace of the Sauber or Force India. The chart for the midfield is shown below. IntelligentF1 model fits are plotted for Perez and for Kovalainen.

The pace in the first stint is very evenly matched (once the Toro Rossos clear Senna), with the Lotus keeping up very well. Sutil is just able to keep his pitstop advantage, but the lack of degradation meant that Perez and Petrov were able to stay with him even on their used tyres. The Renault got to the end of its tyres first, and lost time stopping, and the Sauber tried the undercut. However, Perez’ inlap was poor, and the second lap out was three seconds off the pace (the deficit is clear in the fit). Instead of being a close run thing, this left Force India with an opportunity to make Sutil’s final stop and stay ahead – but it seems unlikely that this was planned as it meant going on to the (slower) hard tyres with half the race to run. Had Sutil stayed out, the slightly better pace of Perez and the small effect of new tyres should have been enough for the Mexican to get past. As it happened, the pace of Sutil on the hard tyres was good – only about 0.6s down on the soft tyres, although the degradation was (surprisingly) higher. This was enough to stay ahead and claim ninth from the late-stopping Senna, who was never realistically on for a point. The disappointment of the race was the disappearing pace of the Lotus – after the first stint it looked like they could be in the battle for a point. The intelligentF1 fit shows that the second stint is a mess – from being right behind Senna, Kovalainen loses more than 20s to the Brazillian. The Lotus was also slower on the hard tyres than the rest of the midfield. So much promise…

Williams didn’t feature as Maldonado retired early and Barrichello was severely delayed on lap one. However, they had good pace. Barrichello was forced into a long stint on the hard tyres, but his pace on those tyres was solid (about 0.7s down on the softs) and his underlying pace was quicker than Sauber, Force India and Renault. A shame – it would have been good to see a Williams genuinely in the mix for the points.

At the back HRT failed to beat Virgin. After Glock retired, both the HRT has a clear pace advantage over D’Ambrosio. Ricciardo had a long stop in the middle of the race, and Karthikeyan lost his almost 20s advantage by hitting the cliff on the softs and having to stop early for the hard tyres. They will have to take comfort from being a full 0.5s quicker.

So a win for Vettel, Button and Alonso on the podium, and another race in which Webber, Hamilton and Massa somehow managed not to do as well as they might. Some mysterious pace from Schumacher and a great strategy call from Force India to claim some points that they probably didn’t merit on pace. Promise, yet disappointment from Lotus, HRT and Williams. And to think I had considered the race a little dull…