A Certain Uncertainty

2011 has seen overtaking at levels not seen for decades, prompting many to hail this year’s changes. However, some observers have claimed that qualifying is now almost an irrelevance. Mark Webber’s recovery from a disastrous qualifying session to finish on the podium in China left some thinking that it was hardly worth bothering with qualifying and that drivers were better off saving tyres. However, has the increased ease with which passes can be made really produced surprise results, with no relation to the starting order? And which races produced the most mixed up results?

A simple way to look at this is the see how the position of the drivers finishing the race correlates with the order in which they started. Note, not their starting position – removing the non-finishers from the equation eliminates the artificial gains of drivers at the back of the grid benefiting simply from retirements up ahead of them. Cue, the Pearson product-moment correlation coefficient which for a perfectly correlation (in our case, the drivers finish in the order they started in) gives a value of one, and a value of zero if there is statistically speaking, no correlation (and a value of negative one in the unlikely case of them finishing in the reverse order to which they left the grid in.)

Looking at the season so far, it may not be surprising that the opening race of the year, in Melbourne, was completed in an order most similar to that in which they started in. Eight of the fourteen finishers completed the race in the exact same position they started in and the other six only made minor gains and losses – Vitaly Petrov’s climb from sixth to third being the biggest mover. Nor is it any shock to see that Canada and Britain produced the most mixed up finish as the influence of the weather came into play. However, even their correlations of 0.75 and 0.76 respectively are not the sign of a remarkably chaotic race, although it’s fair to admit they ignore the way Jenson Button went about winning in Montreal – there are easier ways of going from seventh to first. To put the values into context, the thrilling 2009 Brazilian Grand Prix, which saw a mixed up grid after a wet qualifying session, scored a 0.31 as the back of the grid tore through the field while many of the front-runners tripped up as Button claimed his World Drivers’ crown.

2011correlation

However, who would have thought that statistically speaking, the German Grand Prix was actually quite a dull affair? The front three may have been nose-to-tail for almost the entire race but that’s pretty much where they had started. Kamui Kobayashi’s climb up the field and Sebastien Buemi’s recovery from the back of the grid after being excluded from qualifying the only major changes.

Similarly, who had Spain down as the most surprising dry race – a circuit renowned for “always” being won from pole? Mark Webber proved that assumption wrong, of course, and Nick Heidfeld’s charge from the back of the field to eventually finish eighth wasn’t the only substantial change. That got me thinking as to how a circuit widely considered one of two most boring tracks in Spain Formula One has faired in recent years. Curiously, while 2008 was a statistical procession, the race has enjoyed a reasonably healthy degree of changes of position, even if drivers were reliant on the pitlane in the past. However, it’s also noteworthy that while 2011′s new rules have increased overtaking, that hasn’t necessarily impacted on the eventual result. However, is the pattern of the last three years repeated on other circuits, after all – Barcelona featured one of the more ineffectual DRS zones of the year so far?

SPNcorrelation

What about Istanbul Park, for example? The Turkish Grand Prix was criticised by many for having too long a DRS zone that was making overtaking too easy. Yet the race finish correlation is slightly above the average of 0.84 – although removing the wet races moves that average upwards. The reason is, that for all the incredible number of overtakes in Turkey, normal running order was eventually returned once each cycle of pitstops was completed. However, while it was still an improvement on the previous two years it didn’t compare to the 2008 running of the race although a first lap puncture for the front-row starting Heikki Kovalainen ruined his race and help skew the statistics.

TURcorrelation

But what about the race that almost every fan seems to hate? The European Grand Prix at the Valencia Street Circuit? Last year’s running turns out not to have been all that bad, at least from a statistical point of view but by-and-large, when it comes to accusations of “processional”, the stats agree with the fans for once.

EURcorrelation

Ultimately, all this really says is that the fastest car/driver combinations succeed, whether that means putting it on the front row of the grid or wrapping up a podium finish. For all the overtaking seen this year, it still doesn’t add a increased degree of unpredictability to the race. The only way that is achieved is by the ages old “method” of changing weather conditions. Perhaps Ecclestone was onto something with his sprinklers suggestion…

This post was written by Maverick as part of the VivaF1 Bloggers Swap Shop Series.  You can find the rest of the posts here.

About these ads

5 thoughts on “A Certain Uncertainty

  1. Very interesting idea. I wonder how different things would have been if Vettel had had a couple of retirements in the early races, does one driver making or losing a few places make much of a difference to the eventual correlation?

  2. Pingback: Viva F1 » Blog Archive » Summer Swap: Bloggers and Hosts

  3. Thanks for the great post Mav, impressive work here and a very interesting result! I must admit I had a hunch that with the rule changes this year I thought that wet races were still better, but it’s great to see some analysis to add to that argument.

    For those who see KERS and DRS as too gimmicky, this adds weight to their arguments too…

  4. I think rather than KERS and DRS, tyres have been the big influence on overtaking – but not so much in terms of pace as in as their lifetimes. Shorter lifetime means more tyre changes and it has been the races that have seen lots of tyre stops that have produced huge amounts of overtaking. I think the reason is because more tyre stops mean a greater the number of realistic tyre strategies – not only the number of stops but the order the two compounds are used – with no single stand-out optimum strategy. That means that while the subtly different strategies have produced changes of position in the middle of the race, the end result is more or less the same which ever way they went.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s