The Best Driver of Each Decade
A look at each decade through the lenses of mathematical models.
The debate over who is the Greatest of All Time is arguably the most contentious topic among fans of pretty much every sport, and Formula 1 is no exception. This discussion has been fervently ongoing across the internet since the early days of online forums and has now expanded to social media platforms.
Frankly, I’m not interested in it; therefore, I usually ignore it. However, there are times when certain arguments really boil my blood. The primary offender is the tendency to venerate the present while degrading the past.
It’s an incredibly shortsighted perspective on the sport and on matters in general. It completely overlooks a fundamental concept that is well-known to anyone who has taken the time to learn about human history.
We Are Standing on the Shoulders of Giants
The first step in achieving something complex and advanced is often very simple and basic.
Take external data storage devices, for example. In the 1980s and 1990s, people used floppy disks for this purpose. Nowadays, we have USB drives that can store more data than half a million floppy disks. The technology of that era was quite primitive compared to today, of course. However, floppy disks were the leading technology at the time, no matter how amusing that may seem. More importantly, they represented a crucial step in the evolution that led to the development of USB drives.
You cannot skip a step and jump straight to the complex stage. To illustrate this with a different analogy, you cannot bypass the soldiers of the First World War, the Second World War, the Korean War, and so on, and expect to arrive at modern servicemen. Each of these conflicts represents a stage, with one leading to another, much like the steps of a staircase. Modern soldiers are no different; they too represent a stage that will evolve into even more advanced servicemen.
So, in Formula 1 terms, you can’t have Max Verstappen without Juan Manuel Fangio, Jim Clark, Niki Lauda etc. coming before him. This principle also applies to the next all-time great talent to whom the Dutchman will “pass the torch,” as he himself is a vital link in the chain of the sport's evolution. Interestingly, thirty to forty years from now, there will likely be many short-sighted individuals who will trash what we currently regard as modern Formula 1.
I hope I have expressed myself clearly and that it is easy to understand where I’m coming from and why I ignore the GOAT debates. I simply don’t believe that there is such a thing. I find it more appropriate to evaluate drivers' achievements within the context of their respective eras.
I’m an ardent agoatist. However, I’m not trying to convert anyone to my views; that is not the purpose of this post. I simply wanted to clarify my perspective.
My primary inspiration came from the extensive article written by the creator of the F1Metrics model, which ranks the top 100 drivers in the history of the sport according to this model. While the overall all-time rankings do not particularly interest me, the article features a section that highlights the best driver of each decade from 1950 to 2019, based on a two-year peak performance.
I appreciate mathematical models, because they strive to objectively evaluate drivers, which is undoubtedly challenging in Formula 1, where success largely depends on the quality of the car being driven. In essence, mathematical models aim to isolate and assess performance independent of the machinery.
Obviously, no model is perfect, and it will spit out an eyebrow raising results here and there, and there is no independent or objective method to confirm or deny the representativeness of these results. The same, if not more so, applies to the human eye which people often overlook.
What sells F1metrics to me is the observable trend of performance improvement over the decades. Since its inception in 1950, the sport has become increasingly complex and advanced. Innovations in technology, enhanced safety and reliability, the introduction of run-off areas, increased levels of downforce and grip, improved talent selection through driver academies, and a greater emphasis on physical fitness and diet have all contributed to the overall rise in drivers' performance. In my opinion, any model that aims to be taken seriously should replicate this trend.
I contacted a few authors of publicly available fan-made models and asked them, “According to your model, who is the best driver of each decade based on a two-year peak?” My goal was to gather data from various methodologies while maintaining a similar time frame to that of F1Metrics. Additionally, I requested the results for the current decade. I received only two responses.
The first kind soul I want to thank is the author of F1 Analysis. In fact, the author of the model has written several articles about the best drivers of the 1980s, 1990s, 2000s, and 2010s, free from the constraints I requested, as well as providing a much more detailed breakdown of each era.
The results he sent me look like this.
The next two best drivers of each decade are Alberto Ascari (1952-53) and Stirling Moss (1957-58), followed by Moss again (1960-61) and Dan Gurney (1964-65), James Hunt (1975-76) and Alan Jones (1978-79), the Brazilian duo of Ayrton Senna (1986-87) and Nelson Piquet (1983-84), Senna and Alain Prost (both 1990-91), Michael Schumacher (2001-02) and Lewis Hamilton (2008-09), Fernando Alonso (2013-2014) and Daniel Ricciardo (2016-2017), and finally Lando Norris (2022-23) and Charles Leclerc (2021-22).
The most notable point of contention here is the 2010s. I believe it relates to the two-year peak framework, as the author's own analysis of the past decade yields different results.
Other than the 2010s, the results aren’t controversial – meaning that one can easily construct an argument to defend them. Although the model largely aligns with F1Metrics, the primary difference is the lower ranking of Jackie Stewart. He is the runner-up in the 1960s and occupies the number one position in the following decade according to F1Metrics ratings, but he does not make the top 3 in either decade according to F1 Analysis.
The other interesting difference is the various peaks, which I find most exciting. Like many other fans, I've always considered Alonso’s 2012 title campaign a prime candidate for the best season of the 2010s; however, F1 Analysis does not include it in the Spaniard’s two-year peak. Perhaps he was even better in 2014, but his performance went under the radar due to the atrocious Ferrari he drove?
I would like to express my gratitude to the author of the F1 Mathematical Model for his cooperation. The dataset from his model dates back to 1968; therefore, only the results from 1970 onward are presented. The author has also wrote a more detailed analysis of the best drivers of 2017-21 and 2014-16. I anticipate that the next analysis will cover the period from 2010 to 2013.
The next best drivers, according to the F1 Mathematical Model, are Stewart (1970-71) and Ronnie Petersen (1974-75), Prost (1985-86) and Gilles Villeneuve (1980-81), Senna and Prost (both 1990-91), Alonso (2008-09) and Kimi Raikkonen (2004-05), Hamilton (2014-15), Verstappen (2018-19), Leclerc and Norris (both 2023-24).
It’s worth noting that the top 3 of the 1970s and the top 2 on the 2010s are very closely matched. Based on a two-year peak, there is not much difference between Lauda, Stewart, and Petersen, as well as Alonso and Hamilton.
This model also does not include 2012 in Alonso's peak performance. Furthermore, it treats Schumacher's 2002 season, in which the German stood on the podium in every race, in the same manner. This represents the most significant difference compared to the previously mentioned models.
The three models unanimously agree on the peaks of arch-rivals Prost and Senna during the 1980s and the subsequent decade, as well as puts Schumacher's first two years at Ferrari as his best two seasons in the 1990s.
Frankly, I am a bit disappointed, as I had hoped for a much greater variety in the results, considering that they are outcomes of different methods. However, the three models agree with each other more than they disagree when it comes to identifying the best driver of a particular decade.