#F1 Forensics: 2015 season overtaking analysis



Brought to you by TJ13 Technical Analyst Tourdog

Overtakes make for good racing and meet Bernie’s requirement for a “good show”. That appears to be a commonly held belief amongst most F1 fans. Conversely, a ‘parade’ where most of the grid follow each other around a track for 2 hours is not very exciting. Yet whilst the fans want to see overtaking action, the teams are much less concerned about this. What matters to them is scoring points and more points means more prize money. In fact, it could be argued that the further forward teams are on the grid – which is usually the big budget outfits – the less interested they are in seeing overtaking.

My data collection this season allows me to break things down by the numbers, and while numbers don’t tell us everything, they can at least give us something solid to point to when voicing our own opinions.

Clip the Apex has some very good information on the number of overtakes that have happened both this season and in the past. You can see right away that the number of overtakes has been in free fall since 2011. This years total number of dry race overtakes has fallen from 741, to 513, with an average number of overtakes per race at 30.18. On its own 30 overtakes a race doesn’t sound so bad, the problem is that a lot of these moves didn’t stick.

I grabbed all starting and finishing positions in my data acquisition this year, and the numbers are quite telling:


Click Here to view in a new window

In this table, you can see in what position each driver started and finished. This is not where they qualified, but what grid position they were in when the lights went out, so it takes into account penalties that were awarded for things such as PU and gearbox replacement.

Finish position is also after penalties were awarded, for example Filipe Massa was DQ’d in Brazil, so he is listed in the chart as having finished 20th, instead of 8th. In order to make the numbers work, any driver that failed to finish for whatever reason, has a finishing position number in the order they went out of the race. So if a driver went out on lap one, his finish position is 20th, the next driver out is 19th, and so on. It is obviously not ideal, but was to me, the most logical way to do it. If a finish position has a strikethrough, the driver was not classified at the end of the race.

At the bottom of the chart is the drivers average starting position, and average finishing position. The difference is a positive or a negative number, a positive means on average they gained positions over the course of the races, a negative, they lost positions. That is the number next to “AVERAGE”, and does not take into account a DNF. Just below that, you will see “Races completed”, and then “adjusted”. In the adjusted calculations, I throw out any race a driver Did Not Finish, this is to take into account events that were beyond the drivers control, like accidents and technical failures. This gives us a clearer view of a drivers ability to overtake and hold that position.


Vettel did quite well, on average gaining two positions a race, calculated over 18 races. Kimi was only able to average 1 position gain a race, but he only finished 14 races. Over the course of the season Sebastian gained 38 positions, Kimi 14.


By far the best of the back markers. Both of their drivers averaged 2 position gains per race, and gained a total of about 33 positions over the course of the season. They also took no grid penalties for going beyond their allowed PU usage, so these gains are due entirely to the attrition of other drivers and their ability to pass.


The Mclaren drivers are prime example of how the numbers don’t tell us everything. As a team, they are second to last in total races finished with 24, but they somehow gained the most positions, 78 between them. However, Mclaren were sacked with an unheard of amount of grid penalties this year due to their constant changing of PU and Gearbox components. This put them on the back row of the grid for most of the races, and more importantly, behind the Manors. Passing the Manor cars may have driven their numbers up, but it is a hollow victory, as my nephew on his tricycle could have passed them with a stiff headwind.

Force India

For me the individual  standout is Sergio Perez. He finished 18 of 19 races, and was on average, able to gain 1.6 grid positions a race, for a total gain of 28. Hulkenberg, who many expected to outperform his teammate, was able to gain a single grid position a race and only managed to finish 13 races.


The old saying goes, you can’t win if you don’t finish. Case in point Pastor Maldonado. Whether his failure to see the checked flag was his own fault or not, 10 finishes was not helping Lotus’s situation, though when he did finish, he was able to gain on average 3 grid positions. How much of this was due to his qualifying skills (or lack thereof), combined with attrition of the field, and how much was is skills as a race day driver, we can debate endlessly. Grosjean did slightly better in the Races finished column with 13, and when he did finish, he gained about 2 positions a race.


Numbers, specifically averages, are a funny thing. They tend to be more telling at the top the bell curve, and less useful at the edges, case in point Mercedes. The numbers tell us that on average, the Mercedes twins lost about 0.25 positions a race. Well, when you are always at the top, you have no where to go but down, so every finish position they had below 2nd tended to hurt their numbers more. Lewis finished 18 races to Nico’s 17, and lost 2 more grid positions over the season than Nico. You may remember that Lewis choice to pit in Monaco lost him 2 grid spots and the top spot on the podium. This single mistake is what makes his averages look slightly worse than his teammates.


Poor Williams. Their future looked so bright after last years leap in performance. Most everyone attributed the majority of this gain to the new Mercedes Power Unit, but their drivers obviously share some of the credit for that. This year, things seemed to be moving backwards for them. Massa pretty much just held position all year. He neither gained nor lost positions on average, so where he qualified is pretty much where he finished, Bottas on the other hand, had a much harder time. On average Bottas lost a full position between the start and finish at every race he ran in. Again, the numbers lie a little, because this was not entirely his fault. The teams pit stops probably hurt him more than any deficiency in his driving, famously putting the wrong tyre on in Belgium, which cost him at least 6 grid positions. If Williams don’t want to get passed in the constructors by Ferrari, Red Bull, and possibly FI next year, they are going to have to step up their game.

Red Bull

Riccardo was supposed to be RB savior after the loss of their premier driver to Ferrari this year. Well, it didn’t turn out that way. Their PU problems definitely hurt him, but they also hurt Kvyat equally as much. Kvyat seemed able to recover from this problem much more consistently however. Ricciardo, like Massa, pretty much just held position finishing nearly exactly where he started over the 17 races he saw the final flag. Kvyat on the other hand, was on average able to gain over two grid positions a race, for total of 37 positions gained. Yes, Ricciardo took a 10 place grid penalty in Brazil for an ICE change than Kvyat did not, but one could argue that this actually should have helped him, as it would have been easier for him to pass the back markers during the race than those further up the field.

Toro Rosso

The two rookie were unfortunately in the same position as their RB brethren this year, saddled with a less than competitive Power Unit, but they like Kvyat were better able to cope and made gains similar to him. Verstappen finished 15  14 races, and was able to gain a little over 1 nearly 2 positions a race. Sainz, driving in the shadow of the his well hyped teammate did better, gaining almost 3 positions a race, but only coming home 12 times. Next year will give us a unique perspective on both of these drivers as they will have a 2015 Ferrari strapped to their backs instead of this years Renault, the only team of consequence that will be changing Power Units.

The question is whether the big regulation changes currently being discussed for 2017 will see any change in this kind of pattern.



9 responses to “#F1 Forensics: 2015 season overtaking analysis

  1. Although Verstappen was classified at the Chinese GP, his engine blew 4 laps before the end of the race. The average finishing position without that one would be 9.3 instead of 9.8.

  2. On McLaren: “… but they somehow gained the most positions”
    When you always start in the back row, if you finish the race and other competitors fail to finish, you are going to end up with overtakes without actually overtaking many people.

    Hulk and Alonso made the most first lap over takes and the Mercedes team made the least: http://imgur.com/S1MM2uw

  3. Ideally you should correct the data:
    – DRS overtakes
    – overtakes of backmarkers
    – overtakes of cars retired /retiring
    – overtakes while overtaken is in pits

    Pandora, here’s your box. I opened it.

    @tourdog splendid job you did all year!

  4. A nice piece of number crunching, thanks.
    You’ve clearly done a lot of work, so I don’t want to be picky, but the analysis probably needs a few other parameters factored in to give a clearer picture, such as:
    – DRS operation periods
    – VSC and SC periods
    – Speed deficit on straights
    – Weather conditions
    – Pit stop data and strategy differences
    – Team instructions

    • You’re right. And another thing which happens, the lower you’re up the grid the more you will profit of DNF’s of other cars.

  5. On McLaren, if you’re always last, you can only go up even when it’s not through your own effort. Just like Mercedes can only go down if they’re always sitting at the top. May it stay that way for a looong time.

  6. Great data. No so sure on the analysis. Probably the average of start vs finish is not what we want. Positions gained for each race is probably more useful.

    Then perhaps average for all drivers starting position vs positions gained and difference from average for each driver to see who is better or worse from a position.

    Overtaking is track dependent. So we want to see number of overtakes per race.

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.