I was interested to read earlier this week, Audi’s
announcement of their driver line-up for the forthcoming WEC season. Two cars
for the full season, and a third entry for Spa and Le Mans. When Allan McNish
announced his retirement from prototype racing, my mind immediately
went to Oliver Jarvis, and the thought that here was the chance that the
under-rated Briton needed. I had first met Olly properly in 2008, at which time
he was driving an Audi in the DTM, and it was clear where Jarvis’s ambitions
lay. We spoke about his single-seater career (which had included a win at the
Macau GP – a fine indication of form if ever there was one) and whether he saw
himself one day driving prototypes.
It was very much on his agenda, he said, and when in 2010, he
was given his chance at the wheel of a Le Mans prototype driving for Colin
Kolles I took particular note of his lap times, and was impressed as he easily out-paced his team-mates. His times were also noticed at Audi Sport, and two years later, he was included in the works squad,
driving a non-hybrid R18 with Mike Rockenfeller and Marco Bonanomi, as Audi took
the opportunity to run four cars - both at Spa and at Le Mans.
Spa | Le Mans | |
---|---|---|
Jarvis | 2m 04.346s | 3m 28.068s |
Bonanomi | 2m 04.492s | 3m 28.845s |
Rockenfeller | 3m 28.818s |
My non-statistically minded readers can take a break now (until the next paragraph!),
for I want to spend a moment explaining how my opinion has changed recently on
the calculation of average lap times. The problem comes when one compares, say,
the average of the best 25 laps done by different drivers. Whilst clearly
weather conditions, tyre choice and fuel strategy all have a part to play in
giving drivers a reason to go quicker or more slowly, I have found that it is
the number of laps that they do altogether that can skew the average somewhat
more. For example, if one driver’s 25 lap average is from a stint of 75 laps,
it is unlikely that he will be able to record the same average as a driver
whose best 25 is taken from a total of 250.
So, what I have done above is to use what I call the 20%
average – in other words, to take the average of the best 20% of green laps
completed.
In any event, the table does seem to imply that Jarvis was
quicker than either of his co-drivers in 2012. Now, just for interest, I
thought we should perhaps also look at Loïc Duval and Marc Gené, both of whom
drove the non-hybrid Audi R18 at Spa and at Le Mans in 2012.
Spa | Le Mans | |
---|---|---|
Duval | 2m 03.680s | 3m 26.887s |
Gené | 2m 07.761s | 3m 28.168s |
Here’s where the theory breaks down somewhat, as at Spa, Gené was only out at the start of the race on a damp track, but at Le Mans, Jarvis’s times compare very well against those of the Spaniard. Duval, meanwhile is in a different league (still is, in my view).
Lucas di Grassi appeared in the Audi family in the Brazilian
WEC round at Interlagos, in August 2012. He landed a drive alongside McNish and
Kristensen - replacing Dindo Capello, who had retired after Le Mans. Interestingly,
di Grassi is also a Macau GP winner, two years before Oliver Jarvis.
At a track that he knew well, he was immediately quick. In
the race, his average lap times were quicker than any of the other Audi drivers
(in a race that was dominated by Toyota).
Interlagos was the Brazilian’s only race of 2012, but in
2013, he and Jarvis were to share a car – along with Marc Gené, as Audi focussed
purely on the hybrid R18 e-tron quattro. Jarvis, Gené and di Grassi would only
get two races though – at Spa and Le Mans, as the Ingolstadt manufacturer could
commit to full season entries for only two cars.
At Spa, it did seem as though di Grassi held the upper hand:
his average lap time was 2m 01.834s, compared to Jarvis’s 2m 01.945s, although Olly’s
stint was only 40 laps, compared to di Grassi who did 67. Gené’s average, by
comparison, was 2m 02.610s.
Le Mans was a different matter, with Jarvis doing more laps
than either Gené or di Grassi. The weather was probably kinder to him as well: Jarvis’s
average lap time was 3m 26.481s, di Grassi’s 3m 27.378s and Gené 3m 27.583s.
Jarvis must have been hoping for a season-long seat at Audi
this year. He’s the same age as the Brazilian, but somehow (a) his career has been
a bit slower getting going and (b) di Grassi seems to have made a ‘bigger’ name
for himself at Audi Sport than the modest Englishman, not least in terms of his ability to set the car up, which I suspect will be all-important with the new regulations
arriving for this year.
At Audi, it is all about the chemistry. I would like to
think that Jarvis’s day will come, and I would expect him to shine alongside
Bonanomi and Alburquerque. The tricky thing for the drivers of the number 3 car,
of course is that they will only get two races in which to prove themselves,
whereas the drivers of numbers 1 and 2 will have the full season. Not that
there was ever any question of breaking up the Tréluyer / Lotterer / Fässler
combination of course.
But one wonders how many more seasons Kristensen will drive,
and then, surely, Jarvis’s time will come?
Interesting analysis. I know that your are the data guru. I wonder what other ways you have looked at lap time consistency and race outcome. I would be interested in exploring the effect of drivers' median lap time relative to the rest of the class. That is I would write a hypothesis to test something like: quick pit stops (say 67th percentile), good fuel economy, and median lap time greater than xx% of the field will result in a podium finish 95+% of the time. Fuel economy may be redundant when we have BoP adjustments for fuel economy and it is indirectly captured by lap times (faster laps are generally correlated with higher fuel consumption).
ReplyDeletePhilip