The best Formula One driver of all time according to SCIENCE

[henry]

.....

What infuriates me and is something you touched upon, is that this article regarding the "best driver of all time" has no information as to how they got to the conclusions.
What parameters where set, how why when etc etc.

And in the end, this can be overturned to another set of parameters with a completely different set of results....aka my god is better than yours....

Here's the paper. https://www.researchgate.net/publicatio ... _1950-2014

And https://www.sheffield.ac.uk/smi is the university department from whence sprung the paper.

I think the choice of subject looks to have been made to gain publicity for an analysis methodology. The conclusions on the contribution of teams seem to be as least as significant as the driver rankings.

Hopefully this will answer some of the questions about assumptions and their ramifications.

My interpretation is that "best" means likely to gain more world championship points than all others. This would seem to favour those that have had access to the best teams but the analysis seems to downplay this and hence some who have only raced for the best are not as highly ranked as a simple measure such as count of wins would give. But then what do I know.

[FoxHound]

Much obliged Henry,

When I have a spare hour(s) and not under the influence I shall peruse the paper.

[hollus]

The description of the points system used is very strange. I can't make sense of the formula for "fractional points" for positions lower than 6th.

[Chuckjr]

A lot of folks today have a response to radical shifts in the scientific paradigm which remind me of the church's response to Copernicus and Galileo when they presented the 'radical' idea of heliocentricity. In general, I'm finding our current caretakers of science are not truly open to radical ideas -- Sheldrake's video exonerates this view.

Understandably, this is a engineering forum filled with folks that have a strong codependent relationship with Newton. Such dogmatically tight world views refuse to relinquish 18th century assumptions and this stagnates modernity's progression. Have folks forgotten science only knows what can be repeated, so it's always a secondary, reactive discipline? This fact in an of itself needs to be considered if one wishes to stand on science for any sense of understanding.

I'm simply wanting to offer a different voice here, a more artistic approach, which says we really need to question all we assume to know because understandings like quantum mechanics and electric universe theory are stating the way we have centered our scientific understanding of the world is painfully inadequate and in many cases dead wrong. The air of elitism in not allowing deep questioning of established scientific "law" only serves to render an adolescent insistence of being right despite evidence surfacing to the contrary. It's almost as if scientists are no longer open to proving themselves incorrect, but only correct. This is backwards thinking if progression is the telos.

Just a few thoughts to consider. No big deal. Sorry if this got a bit off topic. Cheers.

[Miguel]

Oh, gosh, I can't believe I'm replying to this at 1:00 a.m.. Anyway, I've only watched the first 6 minutes of Sheldrake's video, and I think it's rubbish. Some of the "dogmas" he presents aren't such. Some other facts are plain wrong. For example, the energy conservation issue is an *observed* fact. It will be thrown away if shown to be wrong. Second, he shows no evidence other than "I think it'd be nice if nature behaved this way".

I also take offence at seeing "electric universe" and "quantum mechanics" in the same sentence, in the same footing. They are not. One is a flawed model that has not a single time even mentioned why Ampere-Maxwell's law doesn't apply to it, while it does in every single experiment we can perform. The other is one of the most successful theories we have, one that has changed the way we view and understand the world, and has done so with some of the most amazing predictions ever.

All scientific ideas are wrong. What was known in the 19th century was way less wrong than before Newton. Nowadays, we are less wrong, especially with regards to very small things, or very fast and heavy objects. What makes science useful is predicting power. No matter how fancy your new ideas are, they will only be accepted if they predict new phenomena, or are able to account for previously unexplained results while still agreeing in the more normal cases.

And if you want to argue that gravity is a r^-1.98 law, rather than r^2, then the elitism you perceive actually means "please, do your homework, I'm too busy with my own stuff.". Because there are papers you can read that deal with that subject. You don't expect me to chew it all for you, do you?

[FoxHound]

I also take offence at seeing "electric universe" and "quantum mechanics" in the same sentence, in the same footing. They are not. One is a flawed model that has not a single time even mentioned why Ampere-Maxwell's law doesn't apply to it, while it does in every single experiment we can perform. The other is one of the most successful theories we have, one that has changed the way we view and understand the world, and has done so with some of the most amazing predictions ever.

Vamos Miguel,

Give us a clue?

[hollus]

A problem with this type of mathematical modeling is that it is blind to the nature of the sport. It doesn't know anything about F1, it just looks at the results. And this makes it almost impossible to distinguish if a driver is good or if a car is good. The math can do their best to separate both factors, but in the end, you can only know if a driver is better, equal or worse than its team mate. How good is the Haas this year? It it had two Grosjeans we would wondering whether it was leading the midfield, it is had to Gutierrez's we'd think it is the 8th-9th team battling with Renault.

So, with that in mind, who were Christian Fittipaldi's team mates? In 1993, Felipe Barbazza and Pierluigi Martini. He dragged his Minardi to 5 points (scores only to 6th) and a lot of top ten finishes. More than double the points and almost double the top 10s than his team mates.
In 1994, in Footwork, he had Gianni Morbidelli as team mate. Christian managed two 4th positions, tripled his teammate's points and more than doubled his top 10 finishes.
These are Alonso vs Raikonnen in Ferrari numbers. So for all the algorithm can know, he is a super-genius that destroyed his opposition 100% of the seasons. And honestly, we haven't any info to say otherwise. The only good comparison teammate there is Pierluigi Martini, a guy which destroyed his teammates most years!
What I find not so sophisticated of this analysis is not to count that fact that n=2 (in seasons) is not exactly the most significant statistical sample.

On the other side, who were Niki Lauda's teammates? Regazzoni and Reutermann in Ferrari, he outpaced them rather well overall but still lost to Regazzoni in his rookie season. Watson and Piquet in Brabham, he outpaced Watson but not Piquet, as least not according to results where DNF dominate the whole season with two finishes in 14 races (Piquet stayed on track longer and thus "won" by a lot of positions according to the analysis' points system. Then in McLaren Watson again, with Watson dominating him in 82 and 83; and then Prost, it was a virtual tie in 84 and complete, thorough, massive domination by Prost in 85.
He was unfortunate to be dominated in his first and last seasons (happens to many) while finding himself in a very good or dominant car in both seasons (doesn't happen to most).
So if you were an algorithm with a very cold mind you'd look at this driver X (called Lauda) with teammates Y and Z in teams A, B and C and conclude: X is a bit better than Regazzoni but didn't do better than the car's average, outpaced Reutermann, lost overall to Watson, both hardly stars, lost by a fair margin to Piquet (unfair assessment driven by DNFs) and then was a lot worse than Prost, winning only when in a dominant car and even losing badly while in a dominant car. Assessment: meh, fastish but inconsistent. Plus, if you "just" win when in a dominant car and "totally" suck when in a crappy car... I can see why just looking at finishing order makes him look not that good. Still, dropping to about 100...

I think the other problem with this algorithm is that it likely rewards gaining many positions (compared to the teammate or the car's potential). The problem is that that rewards midfield battles, where skill is the difference between 12th and 6th. If your car sucks too bad and even DNFs more often than not, skills take you from 20th to 18th, while if your car rocks, skills take you from 3rd to 1st. And Lauda was only ever in those two positions.
The algorithm also likely treats drivers and teams as independent entities, assuming that drivers drive for random teams, and fails to factor in that good teams hire good drivers and hence Cristian Fittipaldi outclassing Pierluigi Martini is not quite the same as Lauda outclassing Reutermann. No, it wasn't only that the Ferrari was that much better than the Minardi, part of that difference stemmed from it having to top drivers driving it and only part from the car. And this is a classic problem in math. You have to (want to, tend to) assume that somethings are independent. But in sports, they are not, and thus Lauda can end up... there.

[henry]

@hollus

Your historical knowledge is impressive.

The model seems to provide a framework for discussing the relative merits of drivers and teams in the way you do. It would be interesting to be able to tweak it

Given the quantified importance of teams the selection of drivers by the best teams looks increasingly like taking out insurance. If, as the author suggests, teams use statistical methods to assess drivers I wonder what criteria they include? I wonder what the probability of a 1m80 driver being WDC is?

[hollus]

No knowledge. Wiki. I only remember Martini because Minardi has had so any Spanish drivers.

[FoxHound]

https://www.researchgate.net/publicatio ... _1950-2014

Right, I've had a bit of time to go over the paper. The math may well be sound, but the torturous method used to get to the numbers are flawed.
I'll pick a case in point.

According to the paper, Schumacher was rated highly, until his Mercedes days when Rosberg beat him.

Michael Schumacher, who holds the record for the most championships and race victories of any driver in Formula 1, comes in a relatively modest eighth place. This is in part because those victories were won in an excellent car, but also because his ranking is dragged down by his more recent post-retirement performances (2010-2012) when he performed less well than in the main part of his career and crucially was generally outperformed by his Mercedes teammate Nico Rosberg. Thus, we re-ran the analysis with the latter section of Schumacher’s career treated as a separate driver. In this formulation, pre-2006 Schumacher’s ranking rises to 3rd and Nico Rosberg’s ranking falls from 13th to 49th. This is because Schumacher’s high standing as a driver in the model effectively deflated 2010-2012 Mercedes’ team ranking in the first model, meaning Rosberg’s performances appeared more impressive. When treated as separate drivers, post-retirement Schumacher performed less well, the Mercedes team effect appears greater, and so Rosberg’s performances no longer stand out compared to his team

There was no mention of 3 critical factors.

Age, and the fact Schumacher was 41 years old on his return.
Retirement in 2006, leaving him out of the game for 3 years.
Rule changes, which left the Formula very different from the last time he'd raced.

Not being Schumacher's biggest fan, I'm all for displacing his status in the all timers list.
But I cannot do so without being dishonest.
All 3 have not been accommodated in the comparisons, leaving Schumacher in a lowly 8th if the math is to be believed.
My opinion, based on what I've seen and the numbers MS generated, is that he has got to be in the top 3/4 drivers if not maybe higher.
It is not anomalous that he is absent from the top 5, it is because the wrong parameters were set in achieving this lists outcome.
Yes he benefited from some wondrous machinery, but then so has every world champion.
So why is it the math is skewed to accommodate this fact for MS but not others?

Additionally, it adds that drivers who perform better at "street tracks" and in "certain weather conditions" bore out stronger points due to less dependency on the car.
Again it falls flat....

Schumacher won at Monaco 5 times, and was renowned as the "regenmeister" in "certain weather conditions", some of these victories labelled amongst the greatest of all time.

So lets look at who they have labelled the greatest. Christian Fittipaldi. I'll focus on points finishes and "street tracks".
Note, I do not intend to denigrate Fittipaldi here...just adding some perspective.

His first points came in Japan in 1992. He finished 6th, and just for perspective the legendary Aguri Suzuki finishing in 8th and Ukyo Katayama 11th on the same lap.
https://en.wikipedia.org/wiki/1992_Japanese_Grand_Prix

Christian Fittipaldi finished 8th in Monaco in 1992, behind a Larousse Lamborghini of a fairly average Bertrand Gachot.
https://en.wikipedia.org/wiki/1992_Monaco_Grand_Prix

And then creditably, finished 5th in 1993, albeit being an attritional race with 5 of the 7 of the drivers finishing behind him having either qualified behind him or involved in race incidents.
https://en.wikipedia.org/wiki/1993_Monaco_Grand_Prix

In 1994 he was running in 4th before his pitstop, and retired due to a gearbox issue some laps later while running in 5th.
Michele Alboreto running in a Minardi Ford finished in 6th at this race....
https://en.wikipedia.org/wiki/1994_Monaco_Grand_Prix

At Kyalami in 1993, Fittipaldi finished 4th out of a total 5 classified finishers, with the 5th being JJ Lehto in a damaged Sauber. It was a wet race.
https://en.wikipedia.org/wiki/1993_Sout ... Grand_Prix

At the Pacific GP in 1994 he finished a creditable 4th, Erik Comas finished in 6th having qualified 6 places lower.
It should be noted that higher qualified cars suffered mechanical failures at a very high rate at this race.
https://en.wikipedia.org/wiki/1994_Pacific_Grand_Prix

In the German GP of 1994, he finished in 4th again with his teammate(Morbidelli) in 5th 8 seconds down the road, and that man Erik Comas again in a Larousse in 6th.
It did help that 10 cars were wiped of the grid in the first lap due to an almighty collision, and a further 8 cars suffered mechanical failures.
https://en.wikipedia.org/wiki/1994_German_Grand_Prix


So by my reckoning, what is borne out of this is that Fittipaldi was a bit better than Morbidelli, who himself scored an unlikely podium in Aus in 1995(In a Footwork!) at a street circuit in pretty much identical circumstances to Fittipladi's points results.

Further, what is apparent in both cases is the results are formed on the basis of attritional races whereby finishing was the most rewarded aspect. In most these races the car also qualified higher(not universally) than the cars it beat in the race(that finished or suffered an issue and finished). Luck opportunism etc are factors also unaccounted for.

So it's with the respect to Fittipaldi, that I say this list is innacurate, and to the creators of the list, that I say this needs recalibration, re-jigging, reworking, reformatting and rethinking to be taken seriously.
And if Fittipaldi comes out on top again, you know, you just know....you have been asking the maths to calculate the wrong questions.

[lebesset]

I may have followed F1 continuously since 1950 , but I don't believe that this qualifies me to judge who was the best ever F1 driver
the only people who can , in my view , are the drivers who take part , and you can only be the best of your generation
but during the last 65 years there have only been 2 drivers who were universally acknowledged by their contemporaries as being in a class of their own
Fangio , who was never contracted to any team ....any team would always provide him with a car if he would condescend to drive for them
jim clark , who famously asked a friend why other drivers in the same car went so slowly , only to be told that they didn't , he could just drive so fast

[OneAlex]

I was dubious when I saw the words "according to science", but then I saw it was the Daily Mail so it seems legit.

It raises a fun idea though. Prost vs Alonso. How close are the two in terms of talent?

[strad]

You can argue it with science or whatever you want, the fact is...It's Fangio.

[Just_a_fan]

I was dubious when I saw the words "according to science", but then I saw it was the Daily Mail so it seems legit.

That's a joke, yes? The Daily Mail is one place one should never look for facts. It's a rag.

[OneAlex]

I was dubious when I saw the words "according to science", but then I saw it was the Daily Mail so it seems legit.

That's a joke, yes? The Daily Mail is one place one should never look for facts. It's a rag.

Sorry, to be clear, yes, it is a joke; the Daily Mail is angry tosh.*

*No offence to people who read the Daily Mail, of course.