Best-of-all-time

There are many Best-of-all-time lists in Formula 1, including scientific ones. 

But none of them solves the issue of comparing apples to apples i.e. differences in competition over 75 years in Formula 1. 

The OneFormula ranking does.

The ranking

Dynamics

The ranking is dynamic. Since the outcome is defined by the level of competition per season, it changes every season. This is explained in the chapter "Methodology"

The graphic below shows that this has an impact on the position of active and non-active drivers alike:

Top 10

• Hamilton plummets from his best 5th place to 9th place, as Mercedes loses its competitive edge in 2022. His move to Ferrari is starting to take on dramatic proportions, prompting Bernie Ecclestone to send him a retirement message. In 2025, he ends up last in the Top 10 

• Since 2021, Verstappen is skyrocketing towards the Top 10 with a superior car, but by mid-2024 Red Bull falters, car and team alike. Towards the end of 2025, team and car recover and Verstappen comes closer than ever to the Top 10. 

Active drivers

2025

Even during a single season positions in the ranking change as can be seen in the graphic below:

• McLaren pushes both Piastri and Norris up in the rankings. Piastri's run towards the Top 25 is slowing down after Hungary, while Norris overtakes Leclerc after Mexico.

• Alonso reaches his highest ranking -12th- with Renault in 2006. With Ferrari, he drops out of the Top 25; with McLaren, Alpine, and Aston Martin, he ends up in the midfield.

• Antonelli's meteoric start flattens out in mid-season.

Methodology

Statistics generate differences of opinion. These differences aren't about drivers, but about the criteria used. This is the subjective element of all so-called objective lists. Moreover, we tend to rate current drivers higher than those of the past. Food for thought for psychologists.

Drivers and experts have different favorites than fans. Alonso scores highly among drivers. Bernie Ecclestone puts Prost at the top of his list. He must have his reasons.

It all comes down to the criteria used. Is it fair that Fangio is at the top? He is the only one who won his five titles with four different constructors. Not because there was less competition back then: that myth is quickly dispelled when we look at the levels of competition over the past seven decades:

More on criteria: FIA uses one single criterion for the championship - points. But it has changed the points systems six times over the past 75 years, which makes it unfit for "Best-of-all-time" rankings.

OneFormula uses three different criteria for its ranking:

1. Wins - in sprint races and regular races alike.

2. Poles - in sprint shoutouts and qualifying. A driver who takes pole retains it, even if he drops back on the grid for technical reasons or infringements in a previous race.

3. Podiums - in sprint races and regular races, but only second and third place count, to prevent double counting of wins.

Weighing factors

The following weighting factors are applied to these criteria:

1. Wins factor 3

2. Poles factor 2

3. Podiums factor 1

This can lead to discussion, but the end result is only minimally influenced by these factors.

Percentages

Wins, poles and podiums are not counted in absolute numbers, but in percentages of all Grands Prix completed by a driver. After all, this is a "Best-of-all-time", not a "Longest-of-all-time" ranking.

Below is an example of the difference between absolute and relative numbers.

DNF

A second characteristic of the OneFormula ranking are technical DNFs. By "technical," we mean car- or team-related DNFs. If a driver doesn't finish a race for these reasons, that race doesn't count towards his score.

This is relevant in a driver's ranking. Jochen Rindt -undisputed champion of DNFs- did not make it to the finish in 53% of his races for technical issues. It boosts his percentages and position in the ranking. Rightly so, because this is a ranking of drivers, not constructors.

Hence, a driver's score is calculated based on the total number of GPs minus the races in which a driver did not finish for technical reasons.

Races in which a driver does not reach the finish, due to spins, collisions or other driver-related issues, obviously do count.

Competition

The points system is applied to the scores of the top six drivers in each season since 1950. This group includes practically all drivers who have won Grands Prix and have achieved pole or podium finishes. 

Below two examples:

The next step is to select a tool that measures the differences between points scored. This would indicate the level of competition between the six drivers. I'll mention just a few:

1. The Gini coefficient

2. Herfindahl-Hirschman index

3. Absolute mean deviation

4. Standard deviation

5. Points winner compared to total points of the top six

6. Difference between numbers one and two

7. Difference between numbers one and two and numbers one and three

Option 7 is added since the first two drivers in the Championship often are team mates. This may not matter in the case of Senna / Prost or Hamilton / Rosberg, but in the case of Schumacher / Barrichello, Hamilton / Bottas and the unscrutable Papaya rules, results are unreliable due to team orders.

After trial calculations with all 7 options, I have opted for the standard deviation. It yields a high level of goodness without dominating the outcome of the ranking. It must be said that the differences between the 7 options are minimal.

The result of the standard deviation is inverted -dividing 1 by the standard deviation. The higher the standard deviation, the lower the level of competition.

Below the C-levels of all seasons 1950-2024:

The Formula

In the final step of the methodology, the outcome of a driver's percentage of wins, poles and podiums is multiplied by his personal C-factor. 

Below the final formula with two examples:

where

• ds = driver score

• wi = wins/net races

• pp = poles/net Q's

• pd = podiums/net races

• cf = c-factor

The final score is multiplied by 100 and converted to absolute number.

Examples

References

• Stats F1 is used as the preferred database for the OneFormula model.

• Peeters, R., Wesselbaum,D,.(2023)  Competitiveness in Formula One, Sports Economic review

• Bell, A., Smith, J., Sabel, C. E., and Jones, K. (2016). Formula for success: multilevel modeling of formula one driver and constructor performance, 1950–2014. Journal ofQuantitative Analysis in Sports, 12(2):99–112.

• Bol, R. (2020). How to win in formula one: is it the driver or the car? The Correspondent.

• Budzinski, Oliver and Feddersen, Arne, Measuring Competitive Balance in Formula One Racing (March 16, 2019).

• Burkner, P.-C. (2017). brms: An R package for bayesian multilevel models using Stan. Journal of statistical software, 80(1):1–28.

• Eichenberger, R. and Stadelmann, D. (2009). Who is the best formula 1 driver? An economic approach to evaluating talent. Economic Analysis & Policy, 39(3).

• Elo, A. (1978). The rating of chess players, past and present. Arco, New York.

• Henderson, D. A., Kirrane, L. J., et al. (2018). A comparison of truncated and time-weighted Plackett–Luce models for probabilistic forecasting of formula one results. Bayesian Analysis, 13(2):335–358.

• Ingram, M. (2021). A first model to rate formula 1 drivers. accessed March 2022).

• Phillips, A. J. (2014). Uncovering formula one driver performances from 1950 to 2013 by adjusting for team and competition effects. Journal of Quantitative Analysis in Sports,10(2):261–278.

• Van Kesteren, E.-J. and Bergkamp, T. L. G. (2022). Code Repository: Bayesian Analysis ofFormula One Race Results.. Quant. Anal. Sports 2023; 19(4): 273–293

• formula1points.com visitors can select from a number of criteria and their weighting factors. Based on the selection, the site produces a ranking. It uses criteria similar to the OneFormula Top 25.