#F1 Forensics: Data suggests Ferrari made little progress in 2015

untitled

Brought to you by TJ13 Editor in Chief Andrew Huntley-Jacobs

The big hope for all neutral F1 fans is that Ferrari will make further inroads into Mercedes advantage in 2016. This optimism is based on the repeatedly pronounced Mantra that Ferrari has been closing the gap on Mercedes during 2015.

Sebastian Vettel intimated as much following the final race of the year in Abu Dhabi, though his analysis was a little – apples and pears. “It’s been an upward curve all year long,” said the four-time F1 drivers champion. “At the start of the year we were probably something like 1.5s behind, and now we’re 10-15s behind over a GP distance”.

The problem with looking at time gaps between cars at the chequered flag, is that cars do not run flat out for the whole race. In the 2015 season finale, Lewis Hamilton clearly gave up the chase of his team mate with 3-4 laps to go. Similarly, Rosberg wasn’t pushing to the very end. Of course we all enjoy analysing the ‘race pace’ data from the FP2 sessions, but even then the cars have differing amounts of fuel and different levels of aggression set in their engine modes. The qualifying sessions are really the best measure we can get of the relative pace between two different F1 cars.

Ferrari did win three races in 2015, but notably only the Singapore GP saw them claim pole position. But does this mean Ferrari have made any progress in closing the gap to Mercedes over the course of 2015?

If we take the same Ferrari and the same Mercedes cars and run them flat out consecutively at two different circuits, the  gap will be different. This is because the characteristics of each car design and its interaction with the tyre vary from circuit to circuit. So to try and understand the true measure of Ferrari’s progress this season, we’ll look at the gap between Mercedes and Ferrari this year in qualifying and see if we can determine some kind of trend.

Here is the gap from the lead Ferrari in qualifying to the lead Mercedes at each of the 2015 race weekends.

 

  • -1.400 AUS
  • -0.074 MAL
  • -0.905 CHI
  • -0.411 BAH
  • -0.777 SPA
  • -0.751 MON
  • -0.621 CAN
  • -0.355 AUS
  • -1.131 GBR
  • -0.719 HUN
  • -1.628 BEL
  • -0.288 ITA
  • +1.415 SIN
  • -0.661 JAP
  • -0.852 RUS
  • -1.772 USA
  • -0.370 MEX
  • -0.522 BRA
  • -0.814 ADH

Sebastian Vettel has outqualified Kimi Raikkonen at all events except in Canada, Great Britain and Abu Dhabi. In Canada, Vettel had an MGU-K problem and in Abu Dhabi was hampered by the team. So to cut Ferrari some slack, for these two races we will substitute the Kimi’s time for one 0.463 quicker – this is the average qualifying pace advantage Vettel has had on Raikkonen all year – and this is what we end up with.

untitled

 

We can see the trend without further statistical analysis. The purple line is the linear trend and shows that in the earlier part of the season were on average around 0.6 seconds slower than Mercedes in qualifying and by Abu Dhabi this is closer to just a 0.5 second deficit. However, this is a fairly crude measure and affected by some big variations which could create a statistical anomaly. So if we remove the data from Australian GP, GB, Belgium, Singapore and the USA we get the following, which of course slightly flatters Ferrari in terms of the average gap.

untitled

Oh NO!!! This is too depressing to imagine. Now the actual trend of Ferrari’s performance is deteriorating over the course of the year 😦  There must be some tools in our stats bag to cheer us up a little.

It could be argued that linear analysis is too crude a measure given the length of time over which the data is collected (that’s bollocks really, but let’s roll with it). So here’s another way we can establish the trend that is the Ferrari gap to Mercedes in qualifying.

untitled

Here the trend line is the moving average for three races. So we’ve chopped down the total time span for measuring the Ferrari progress of lack of it into three race chunks of data. Over the season Ferrari now appear to have made a small gain on Mercedes of around half a second, but are still just over 0.4s slower in quail trim than the Silver Arrows.

Maybe by cutting the data set we have done Ferrari a favour in terms of the average differential to Mercedes over the year, but harmed the trend of their performance. So here is the full seasons times, with a rolling average trend line.

untitled

Sorry folks, I tried my best. This chart with a rolling 3 race average trend line merely demonstrates what we learned in the raw data. Ferrari have improved over the year, but not very much. In reality, the Reds were never as far behind as Sebastian would have us believe. Of course if our very own Tourdog can come up with some kind of 2015 event by even race pace analysis (good luck that’s really complicated) of the Red team, this will probably show a greater level of progress by Ferrari in closing down the lap time gap to Mercedes.

Yet with the 2016 F1 car design and aero regulations changing very little and Pirelli being given no time to deliver tyres that go off the cliff, qualifying is set again next year to be the most important part of the F1 weekend. If true, Ferrari may win just 3 or 4 races despite all their crafty Haas F1 associated activity.

Stats and damn lies – let’s hope 😉

 

14 responses to “#F1 Forensics: Data suggests Ferrari made little progress in 2015

  1. Just goes to show the bollox that statistics really is….. do analysis, dont like the data, analyse the analysis, re-run the analysis and repeat until it shows up more favourable results

    points and finishing positions are what counts here at the end of the day.
    2014 = no wins and no poles,
    2015 = 3 wins and 1 pole

    ps – not necessarily a a Ferrari fan btw

  2. I screwed around with the data on excel and found that if you drop all the outliers, change a bunch of negatives to positives, rearrange the sequence, use an exponential best fit, and apply a fudge factor Ferrari are faster than Mercedes. Yes, lies and statistics.

    • I have to disappoint you. I set all the data to 0, except the drivers’ names. What’s plain to see then, is that Lewis Hamilton comes first in alfabetical order. Kimi Has a good name but his surname fucked him.

      So it all will be the same next year.

  3. nice try AJ -would like to see if the gap as a function of lap length, (and possibly time throttle too) to see if that makes for more positive analysing. The gap at Monza is bound to be greater than that at Monaco, for what its worth…

      • @schumerak and @gomer A quick piece of mental arithmetic would reveal there will be little difference. But feel free to take the time and send us a ‘voice of the fans’ analysis article.

        • @ The judge.

          Could you make the excel file available? Would love to peruse the court documents…. 😉

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