Blog post

Man versus Machine: Formula 1 Racing

By Andrew White | October 26, 2020 | 0 Comments

Formula 1Data or AnalyticsData and Analytics

If you have been following my blog you would know I am an F1 follower. Over the years I have blogged about how computing power is less useful than asking smart quarantine of data. This was demonstrated by Mclaren, an F1 team that gets the same voluminous data from its cars during a race as every other team, contracts with SAP (some years ago now) to use its then new in-memory analytics capability Sap Hana. The use of SAP software did not materially change the winning fortunes of Mclaren, even though the other teams didn’t have SAP software and all team’s had the same data streams.

More recently and my favorite F1-related blog touched in strategy. Toto Wolff, current chief executive for Mercedes, another F1 team, had just seen one of his two drivers win a race. It was a most unlikely win, given race conditions, use of safety cars, and a tire puncture on the last lap. He was asked after the race how his strategy had paid off. He said to the camera, with a wry German smile, “Strategy is learning by doing.” In other words, there was a strategy of sorts at the start of the race, but what was effective is that the strategy adjusts quickly and was flexible and adaptable, not perfect and so brittle. Brilliant.

Last weeks edition of the Economist included an excellent article (see Man v Machine) that explored the contribution of driver and machine toward a winning record. The article was triggered by an historic feat: Lewis Hamilton, a Brit driving a Mercedes car, had just won his 91st F1 race. This equaled the record by Michael Schumacher, a German, who had driven a Ferrari for most of his 90 wins. I remember the day seeing that 92st win and I said to myself I would never see that bettered. How wrong I was.

The article develops a model that tries to take account of driver, number of cars and different teams they raced for, and the car themselves, in order to rank the greatest F1 driver of all time. Cars are faster today; technology is more evolved; drivers are better, maybe…

The bottom line is that neither Hamilton or Schumacher quite get rated top. The both raves for fewer different teams than winning drivers of the past. This shows how the car provides a more consistent and reliably progressive platform for any driver: the modern drivers are lucky to enjoy them. History’s earlier winners had to win in many more different cars with different capabilities. Taking this into account with the assumptions made and the greatest driver of all time is Juan Manuel Fangio.

The point being that machines, more precisely the data and analytics used to design and engineer them, have a greater contribution to success than flesh and bone. Both man and machine are still needed. But the power balance continue to shift in different ways.

Of course, my long gone mother (bless her soul) told me who she thought was the best F1 driver. It wasn’t any driver mentioned in this blog. A good example of why no amount of analysis will resolve discussions of subjectivity. Then again, my mom was probably right.

Comments are closed