This article will look at the history of Formula One from 1950 to work out who its greatest ever drivers were, and how they compare to each other. We’ll look at several different questions, culminating in rankings of drivers:

1. Who was the best performing driver in each year of F1 history?
2. Who was the best driver in each era of F1?
3. Are F1 drivers improving over time?
4. How do the greatest drivers from each era compare?

Assisting me in this task is my mathematical model, which compares results of teammates across the history of F1 to draw conclusions on how strong each driver is (both as a whole and in individual seasons) regardless of the machinery they drove. Different drivers from the same era can therefore be ranked and directly compared. It also theoretically allows drivers from different eras to be ranked too, but as we shall it is not necessarily appropriate for them to be compared directly against each other in this way.

For new readers (welcome), here is a quick summary of what the model takes into account:

1) The percentage of the points that two team-mates score relative to each other each year. (Non-points finishes are also considered for less competitive cars.)

2) The fraction of each year they spent as team mates.

3) The points they scored compared to the maximum points available.

4) The drivers experience and age during the season.

See more information on how the model works here.

The short version is that for each driver it looks at how they have performed versus every teammate they’ve had in F1 to give them (and every other F1 driver) a relative score.

Of course such a system is never perfect. One thing to note here is that it’s only results compared to teammates that are compared. Other potentially valid methods such as outright pace or circumstances causing the results are not considered. As such, factors such as mechanical unreliability that primarily hits only one driver within a team can skew data over a short period of time (although an unreliable car does not skew the data much if both drivers are affected). The effect is generally small over a modern F1 season, and negligible over an entire career. However, the effect increases when looking at older F1 seasons. This is due to the increase in DNFs, the reduced number of races and the shorter careers of older F1 drivers.

I have previously looked at the best driver (and team) each year going back to 1985. The update below goes all the way back to 1950. It also includes additional information that has come from the 2021 and 2022 seasons. (These have mainly boosted Verstappen’s rankings as a whole).

So who is ranked as the best driver each year?

The Best Driver Each Year

Out of the 72 seasons here, the best performing driver won the world championship a total of 27 times (a little over a third of the time).

Changes from last time:

• Schumacher is now awarded the 1992 title. Schumacher, Alesi and Senna are all rated relatively closely in 1992, and this is one of the few historical years that could possibly change again in the future. The model puts the season as a weak one for Senna, a strong one for Alesi, and an typical one (given his inexperience) for Schumacher. Prost was not racing this year, but he would probably have been rated the best performer had he been present.

• Verstappen has not only gained the 2021 title, but 3 others along with it (2015, 2018 and 2020). This is due to a combination of factors, but is mainly due to Max’s ranking improving overall. However, it should be noted that these could change again over time. This is particularly true of 2015 and 2020 (which would go to Hamilton and Leclerc if Verstappen’s ratings drop). Regardless, the model considers Max a generational talent and thinks his 2021 season was one of the greatest ever.

Multiple Season Winners

The model usually believes there is one driver that stands above all others in any given time period. See below for drivers that received the title of best performer in a given year more than once:

Overall the model does a good job of spitting out the top drivers of a given era. Fangio, Clark, Lauda, Prost and Schumacher are all giants of the sport that are rightly recognised in F1 history.

The other driver that the model gives a glowing review of is Max Verstappen. Given he is a current driver that is probably not even half-way through his career, he has plenty of potential to record many other impressive seasons. However, his rating is not as secure as others on this list. This is mainly due to his teammates generally coming from the Red Bull program. This means there is a small clutch of drivers whose ratings are more dependent on each other.

Senna and Ascari are not quite considered to be the dominant driver of their era. However they could take the fight to Prost and Fangio (respectively). Both also suffered from tragic early deaths, and would likely have been more successful had their careers continued.

Stirling Moss is regularly cited as one of the greatest (of not the greatest) driver never to win the title, which is also something the model agrees with given the era he raced in. As we shall see though, his rating is not as high as other drivers such as Fangio, Ascari and Clark. There is therefore considered a strong driver, rather than a dominant one.

Similarly Fittipaldi and Angelis are also not considered to be the best of their era, but were still highly competitive. Angelis, for example, was rated the 3rd best driver of the 1980s in a previous post.

It is apparent that some drivers receive the best performance title almost every year that they competed, with 6 drivers hitting a rate of >50%. Unsurprisingly, 3 drivers with the highest percentage of top performances are relatively inexperienced.

We can also see that comebacks after at least 2 seasons out of the sport do not generally lead to further top performances. This is true to a range of factors, including age related decline and the grid as a whole improving. Even Lauda’s miraculous 1984 championship victory against teammate Prost was not enough to give him a top performing score that year.

It is very rare for the model to think that there are 2 great drivers competing at any given time. Senna/Prost (late 80s – early 90s) and Schumacher/Alonso (mid 00s) are 2 exceptions. Although the results above do not clearly show this, the model sees potential for Verstappen and Leclerc to be the defining drivers of the next generation.

In case it was not already clear, the probability of the best driver of the era being ranked as the top performer is significantly higher than their chance of securing the championship.

Single Season Winners

There are 2 reasons why the model would give a single championship to a driver. Firstly it could be because the model thinks there was no dominant driver at that time. This mostly explains the single championship successes up until 1979.

Other drivers here (e.g. Frentzen and Ricciardo) can be considered fairly lucky to have a best performing season. The model sees them as top 5 drivers for several years. They just happened to peak when the absolute best driver(s) of the era were having an off season.

That just leaves Lewis Hamilton. His 2007 season is considered to be phenomenal considering his rookie status. Surprisingly though, the model doesn’t even rank it in the top 5 years of his career. This is because of the way a rookie season is assessed. Given that Hamilton was one of the best prepared rookies in history and hit the ground running, it’s likely that his 2007 season is underrated. Overall Hamilton is considered to the strongest driver on the grid across his career bar Alonso (and more recently Verstappen).

Another surprising result is that Jackie Stewart is considered a driver that is very good, rather than an obvious all-time great. His top performance in 1969 is joined by 2 close runs in 1970 and 1971. Both expert rankings and other models rank him among the best of all time. I have tried to explore reasons why this does not translate to my model, but have not come up with any obvious answers.

What is true is that Stewart’s ranking has a large amount of uncertainty to it. Partially this is due to the era as a whole (he retired from 9 out of 11 races in 1967, for example). Stewart also spent his last 4 seasons, including 2 championship wins, partnering François Cevert, a man who had very few external drivers connections. This, along with a middling comparison with Graham Hill (another driver who is probably rated lower than general perceptions) is perhaps a cause of his lack of top performances. Despite this, Stewart is still ranted as one of the best of a competitive era.

Meanwhile, incredibly successful drivers such as Piquet, Mansell, Häkkinen and Brabham are thought to be good drivers that were flattered by their machinery. It’s always difficult to judge people’s perceptions, but I would bet that most would concede that none of those drivers were the absolute greatest of their era. (How many would pick the above names over Prost, Senna, Schumacher and Clark?)

Who is the GOAT: Comparing Different Eras of F1

Power Creep

I have previously looked at the ‘power creep’ that the model seems to think exists within Formula One. (That is to say, the idea that F1 drivers are generally getting better over time). Although that analysis was for the 21st century, the process appears to have existed for most of the history of F1.

The general trend of the grid improving over time is present across the history of F1, but particularly prevalent since the 1970s. As a rough guide, a performance good enough for 5th best of a season would be only around the rated as the 10th best performance 20 years later. As looked at previously, I cannot find anything within the model that explains this. (Drivers do not, for example, get worse over time which makes new drivers artificially look better). The most reasonable conclusion is that the effect is real. I should also add that the best fit lines here are mainly just for illustrative purposes, and it’s not clear from this data that the improvement is naturally linear.

It is notable that in sports where individual performances are more easily measured the competitors are consistently improving too. A classic example is the men’s 100m race. The slowest time in the 2020 Olympic final (9.98s) would have been a world record time during the 1960s.

In the previous analysis I noted that the rule of improvement did not obviously apply to the absolute best driver, due to it being more volatile. Whilst the volatility is still there, there is also an overall trend of improvement among the best driver of the era. This means that the best rated young drivers in F1, currently Verstappen and Leclerc, are currently rated as two of the strongest drivers of all time. (Higher than Fangio, Prost, Senna, Schumacher and Hamilton, for example). It’s of course unknown whether they will stay so highly rated over their entire careers, but it’s a good example of the overall trend.

Eras of F1

To try to prevent the issue of power creep distributing the results too much, I have presented ratings for top drivers of each era in graphs below. Each graph is recalibrated so that a value of 1 is equal to a median performance of the top driver in the era. (In the case where there’s no identified top driver it’s the median of the best performances over the era.)

This means that that graphs from different eras shouldn’t be directly compared to work out whether Fangio could beat Schumacher, for example. Instead, they are designed to show how good each driver was given the quality of other strong drivers in that era. However, each graph has a small crossover with the next, so you can see what recalibration is taking place each time. In general, the quality of drivers improves over time.

One last thing to keep in mind is that the title of “Best Other Driver” is somewhat fluid. It may mean the 1st, 2nd or 3rd best driver that year depending on how many other drivers are considered. It is there to give more context to the drivers that are being focused on.

F1 GOAT: Juan Manuel Fangio and Alberto Ascari

Fangio is ranked as the number 1 driver of the decade, with Ascari a close second. The pair only spent 3 proper seasons together on the grid (1950, 1951 and 1953), and generally seem closely matched during that period.

Fangio’s much lower 1950 ranking suffers from poor reliability. An even more extreme example of this is Ascari’s 1954 season. Ascari DNF’d in every race he entered! He actually still scored in half as these races due to the fastest lap rules at the time. The results of this season are not included here as he raced in fewer than half the GPs

José Froilán González is largely forgotten from F1 history, but is also quite highly ranked and held up relatively well as both Ascari’s and Fangio’s teammate. His contributions are the higher than average scores for Next Best Driver in 1951, 1953 and 1954. In 19 race finishes he recorded an impressive 15 podiums, although only 2 wins.

Behind him is a competitive small group of competent drivers, which 1950 champion Farina is a part of. The quality of drivers on the grid during this period varied significantly. Many drivers also only raced only in a few Grand Prix and teammates changed regularly within a season.

As noted above, Fangio has a remarkable conversion rate of both championships and top performances. His win rate was also incredibly impressive, especially when DNFs are discounted. Fangio is ranked as the strongest driver in the first 30 years of F1 history. He’s also one of only two non 21st century drivers to feature in the top 10 list of best F1 drivers. This is all the more remarkable given that he was in his 40s for almost all of his F1 career.

Despite how close Ascari’s scores are to Fangio’s, he is somewhat lucky to have 2 best performing years. In one Fangio had reliability issues (1950) and in the other he didn’t even compete (1952). However, had Ascari survived his fatal crash and continued racing he would certainly have more years at the top. He’d also probably be better remembered by fans of F1 stats too.

F1 GOAT: Stirling Moss

Once Fangio retired in 1958, Moss is considered to be the best of the rest going into the 1960s. However, not only is he rated someway short of the top drivers before and after him (Fangio/Ascari and Clark respectively), but he is also much closer to the rest of the grid. His inclusion as an era defining driver is therefore much more debatable than others.

His reputation as one of the greatest (if not the greatest) driver to never win the world championship is a deserved one though, with Elio de Angelis being the only other non-world champion to be ranked as the best performer twice.

1958 world champion Mike Hawthorn is rated at a similar level to Moss. Meanwhile 1959 champion Brabham is seen as strong at his peak, but quite erratic across his career.

F1 GOAT: Jim Clark

Jim Clark is easily considered the best driver of his era. After a couple of years finding his feet (which is common for new drivers throughout F1 history), he is regarded as the best performing driver for 6 years in a row until his untimely death. Like Ascari before him, he surely would have won more and be better remembered by the F1 community at large had he been able to continue racing.

His 1963 seasons is considered to be one of the greatest in F1 history. Clark won 7 races in a season where only the top 6 results counted towards the championship. Without reliability troubles it’s quite possible he would have won every race. Whilst it’s tempting to dismiss this dominance as due to the car, the fact that his teammates scored a combined total of 1 point for the entire season puts things into perspective.

Clark’s 2nd championship in 1965 season was equally dominant, scoring 6 wins when again only the top 6 results counted. His lower score is due to the improved fortunes of his teammate. Spence scored 10 points across the season including a podium at the final race.

Other strong drivers of the era include Gurney, Surtees and McLaren. Two time champion Graham Hill is not as highly rated. This is at least in part due to a long career with varying degrees of success.

F1 GOAT: No Dominant Driver

After Clark’s death in 1968 there is a period of several drivers vying for the spot of best performer. The spikes in best performer prior to this period (Clark) and afterwards (Lauda) also indicate that the model is not convinced anyone in this era was truly exceptional. This does not necessarily mean that the overall quality of driving was poor though. In fact the model has concluded that there was a large number of closely ranked drivers delivering similar performances.

The model records 7 drivers (Surtees, Stewart, Ickx, Fittipaldi, Peterson, Rodriguez, Rindt) who were capable of delivering top performances during this era.

F1 GOAT: Niki Lauda

In a trajectory similar to Clark, Lauda took a few years to find his groove. After this he’s clearly considered to be the top driver of the 1970s. Other top drivers of the era such as Hunt and Fittipaldi are ranked highly too. This makes it an era where the model follows drivers actual successes (and probably perceptions as well) quite closely.

This graph above also shows the inching up of the Next Best Driver as the decade progresses, a trend that is not obvious in the previous era graphs. By the end of his first retirement, he was overtaken by other drivers. This trend is regularly maintained by the model, with Prost, Schumacher and Alonso suffering similar fates.

F1 GOAT: Alain Prost

Prost established himself as one of the top drivers on the grid almost immediately after debuting in 1980. Within a few years he is considered well clear of everyone, a position he does not relinquish until the 1990s. Prost’s long period of success (both by the model’s reckoning and his actual results) is much more similar to modern F1, where a career spanning over a decade is no longer considered unusual.

Elio de Angelis is not considered to be on Prost’s level, but is good enough to take a couple of top performances during Prost’s early years when his results are less consistent. Senna’s performances never reach Prost’s level at his absolute peak, but their averages are pretty close. I have compared Prost and Senna in more detail here.

One final detail in this graph is the apparent drop of form of the Next Best Driver as the 1980s progresses. This is mostly caused by a trick: If you consider the Next Best Driver to be Senna then you can see there’s no significant drop. Another point is that even with this “drop” in the 1980s, the Next Best Driver score is still higher than a decade ago.

F1 GOAT: Michael Schumacher

Senna and Prost continued their dominance into the early 90s. Following Prost’s retirement and Senna’s death, Schumacher is easily ranked as the best driver for almost an entire decade. 1994 is considered his peak, a year in which he won 8 of the 10 races he finished. The season was however full of controversy, with Schumacher having 2 disqualifications, a 2 race ban and a championship deciding collision with rival Hill. His team were also accused of cheating on more than one occasion.

The short answer for the massive dip in 1999 is that Schumacher broke his legs mid-season. However, this by itself would not effect his score. The result is mainly caused by an inconsistent start of the year. Whilst his support for Irvine’s title bid upon returning had an impact too, at the point of his injury he was just 6 points clear of his teammate, far below where he was in every other year they competed against each other.

Below Schumacher is a wealth of good drivers, with Frentzen, Villeneuve and Hill being the highest ranked. Frentzen excelled in midfield cars but couldn’t deliver when he got his major chance (Williams, 1997). Hill and Villeneuve both took championships for Williams with a significant car advantage, but delivered some decent performances for less competitive teams too.

Likewise the model concludes that double world champion Häkkinen was aided significantly by his car. His peak season was his first championship win in 1998, although even then there are several other drivers that are rated higher.

Lastly, we can see that Schumacher’s last 4 years at Ferrari were seen as slightly below average. This includes his dominant 2004 season, which many found questionable when I previously discussed this. As a simple explanation, Schumacher’s points total was almost identical for 2004 and 2002 (his other dominant year), despite having an extra grand prix and a more generous points system. Meanwhile, teammate Barrichello’s points total increased by almost 50%. Whilst there are many reasons for this, the fact is that Schumacher wasn’t able to show the same consistency in 2004 despite having a more dominant car underneath him. (For those wondering,the points system change adjusts the numbers a bit but doesn’t undermine the fundamental conclusion.)

F1 GOAT: Fernando Alonso

Schumacher continued to dominate at the start of the decade, but the model considers Alonso to be the dominant force from 2003 until around 2015. At points he is considered well clear of the rest. His 2012 season is regularly mentioned as an all time great performance. However, the model considers his championship Renault years to be even better.

I’ve commented on Schumacher’s 2004 season, but Alonso’s deserves a quick mention too. How has he ended up as the top performer is he was outscored by teammate Jarno Trulli? Firstly, Alonso scored more points per race at Renault than Trulli during the season despite having more retirements. This is because Trulli was fired before the end of the season and Alonso enjoyed a strong end to the year. This, coupled with an impressive comparison vs new teammate Jacques Villeneuve, even accounting for Villeneuve joining towards the end of the year, means that his the model thinks it was indeed a slightly below average year for Alonso, but still an impressive year by most driver’s standards.

The period of Alonso’s dominance in the model’s eyes includes many other world champions: Räikkönen, Hamilton, Button, Vettel and Rosberg. Of those, Hamilton is considered to be the strongest, with the model ranking him the best driver other than Alonso every year between 2007 and 2010.

Vettel, Rosberg and Button meanwhile are thought of as being near identical overall. All 3 considered to be Hamilton’s equal at their peak, but unable to maintain such form across their entire career. This demonstrates the depth of skill within the more recent F1 grids, without even mentioning the number of quality non-champions such as Ricciardo and Kubica.

That leaves Kimi Räikkönen, who is still rated as an above average driver, despite being considered the weakest world champion of the 21st century.

Of all the driver’s mentioned so far, Alonso has perhaps the biggest disparity between the model’s rating of him and his success in Formula One. This is not to undersell his success. He is still a double world champion with the 6th highest number of wins in the sport’s history after all. But even that feels disappointing given the model’s result of 12X best performing driver.

The reasons for this have been discussed many times elsewhere, but it is at least partially a consequence of Formula One becoming a more car dominated formula over time. Whether this is a good or a bad thing is up for debate though (on reason it is the case is that there are more good drivers than ever, which naturally makes the difference between drivers smaller).

F1 GOAT: Max Verstappen

Max Verstappen burst onto the F1 scene, winning his first race in just his second season and quickly establishing himself as a top driver. However, the model thinks he did not reach his peak level until 2018. Whilst his 2021 performance was lauded in my year review, the previous few years are of similar ratings. There is one major other consideration though: Verstappen’s teammates during this those years were all inexperienced drivers in the Red Bull program. The years 2019 and 2020 and therefore more vulnerable to change with additional data.

It is unlikely that Alonso, Hamilton or Vettel will be able secure a top performing ranking this decade. (Hamilton is by far the most likely of the 3 to achieve though). This means that any challenge to a Verstappen dominated decade will be from other relatively young drivers, with Leclerc being the prime candidate.

From the graph you can see that Leclerc’s level is not as high as Verstappen’s over his time in F1. However, when comparing their first 4 years in the sport they are roughly equal. This suggests that he definitely has the potential to take the fight to Verstappen long term. As of the 2022 Monaco GP Leclerc is also the highest ranked driver of 2022. In short, both are ranked very highly, and their long-term paths are still uncertain.

There are 2 other drivers currently on the grid that the model thinks has the potential to deliver great things . Norris’ rating is currently just a touch below Leclerc’s. It remains to be seen whether he is currently overrated due to Ricciardo’s McLaren struggles. He may even be underrated if he took time to reach his current form. Finally, Russell’s ranking is gradually increasing over time due to strong showings versus Hamilton. His early years suffered from an inability to get points on the board despite showing obvious speed, but that does not appear to be an issue at Mercedes.

Summary of Eras

Rankings of GOATs

Finally, I want to give some rankings of drivers based on the model. Rather than rank every F1 driver in history, I’ve decided to make 2 criteria for a driver’s inclusion:

1. They have been discussed above as defining an era based on the model’s results. This includes: Fangio, Ascari, Moss, Clark, Lauda, Prost, Senna, Schumacher, Alonso, Verstappen.
2. Alternatively, they have at least 3 world championships. This also includes: Brabham, Stewart, Piquet, Vettel, Hamilton.

Obviously this is a non-exhaustive list that misses out some fairly big names (e.g. Mansell, Häkkinen and Räikkönen). It is also a somewhat arbitrary distinction based on the model’s results. Fittipaldi and Angelis could be included, or Moss could be left out. However, without ranking every F1 driver in history there are always going to be some cuts.

For this ranking I decided to use the driver’s 3rd best year as rated by the model. The idea is to strike a balance between consistency and peak performance.

Absolute Ranking of F1 Drivers

So, Verstappen is currently considered to be the greatest F1 driver of all time. However, due to the power creep factor these results will naturally tend to favour modern drivers. How prevalent is that effect? It may not be immediately obvious, but it’s actually incredibly strong. For example, running the results again with every F1 driver in history we get a pretty shocking outcome: 7 of the top 10 F1 drivers ever by this metric are on the current grid!

It’s an open question whether this is a fair way to compare drivers from different eras. After all, they can only beat the competition in front of them. The likes of Fangio, Lauda and Prost would also undoubtedly have been better drivers if they had the ability to use the training, professionalism and technology involved in modern F1.

Another way to approach the problem is to adjust driver scores based on the era they raced. This requires a fitted function to be able to adjust drivers ratings based on their era. I ended up using a basically flat profile for the first 20 years of F1, and a linear one afterwards. (In contrast to the line of best fit presented for the previous power creep graphs). This provides a good fit to the overall trend of the data, although of course there is some noise.

The idea here is to rank how good a driver was given the era they raced in. There are pros and cons to this approach as opposed to the results above, but I think on balance it is a fairer way of assessing which driver(s) can be considered the greatest ever. It gives the following rankings:

Ranking of F1 Drivers Adjusting for their Era.

Conclusions from this result:

• The 15 drivers as a whole close up due to the adjustment for their era.
• Fangio & Prost actually have near identical rankings, with the model considering them the 2 strongest drivers ever.
• There is an obvious leaning towards drivers that are considered dominant in their era. The top 4 places are occupied by the 4 drivers that have spent the most seasons as the top performer.
• Meanwhile, of the bottom 6, none truly dominated their era in the model’s eyes.

Finally, another metric that appealed to me but I ended up not perusing was consistency across a driver’s career. I found that it was almost identical for most of the top drivers I looked at. However, surprisingly, there was little correlation between consistency across a driver’s entire career and across their peak period. For example, Lauda was considered a very consistent driver at his peak, but his form varied a lot over his career. This is mostly attributed to a slow start in the sport. By contrast Vettel has a poor consistency from one year to another, but his overall level of variance was not considered to be poor (i.e. other drivers would be more likely to have periods of high/low performance lasting several years, whilst Vettel’s would be more likely to last a single year).

Conclusions

-The model generally thinks there is 1 dominant driver at any given time in F1 history. Occasionally there are crossover points or times where there isn’t such a driver.

-Drivers considered the best of their era are: Fangio, (Moss,) Clark, Lauda, Prost, (Senna), Schumacher, Alonso and Verstappen.

-It’s Alonso, rather than Hamilton or Vettel that is considered the best driver for the majority of the last 20 years. However, both #44 and #5 are thought of as well above average drivers. This is true both in their era and across F1 history.

-Verstappen is currently considered the best driver in F1. However, his ranking within F1’s hierarchy of all-time greats is still fluctuating. It is also possible that another driver could provide a consistent challenge in the coming years.

-The model considers the F1 grid to be getting stronger over time. This is true for the best drivers of the era and the top half of the grid as a whole.

-Not adjusting for this fact leaves Max Verstappen as the greatest F1 driver of all time.

-Adjusting for the grid increasing in strength leaves Fangio and Prost as the 2 best drivers given the era that they drove in.

Thanks for Reading!

• WordPress
• Twitter
• Facebook
• Reddit

• Why Is F1 Suddenly Popular in China?
• F1 To Race On The Moon In 2025
• 2010 F1 Driver Rankings
• 2024 F1 Driver Predictions
• How Well Does the First Race Predict the Championship Winner?