https://f1metrics.wo...river-rankings/
I know there's already a driver ranking thread out there, but this one is such a detailed analytics based one that I thought it was worth its own thread.
Posted 26 November 2014 - 20:40
https://f1metrics.wo...river-rankings/
I know there's already a driver ranking thread out there, but this one is such a detailed analytics based one that I thought it was worth its own thread.
Advertisement
Posted 26 November 2014 - 21:48
Very interesting stuff. Not much I would dispute, though I would swap Rosberg with Vergne.
Posted 26 November 2014 - 22:10
Enjoyed reading that, a different perspective on things. Alonso always seems to find himself near or at the top no matter what metrics are used to evaluate a season, damn he's good.
Posted 26 November 2014 - 22:25
It must be my overly critical mind that reads stuff like this and goes "well this can't be right, and that looks questionable, etc., but in all honestly this is very impressive work.
Posted 26 November 2014 - 22:35
Posted 26 November 2014 - 23:25
Enjoyed reading that, a different perspective on things. Alonso always seems to find himself near or at the top no matter what metrics are used to evaluate a season, damn he's good.
I have a huge amount of respect for Alonso. No matter what car you give him he still finds himself up near the front somehow. As the old saying goes, he could win on a lawnmower.
I said earlier in the season that Ferrari would not win a race this year but Alonso might (if you get what I mean) and he came so damn close in Hungary. With 4 laps to go I had that prediction quoted and was going to be all boisterous about it....
...then Ricciardo got past.
Posted 26 November 2014 - 23:35
Only problem with that is Kimi is way too high on that list.
He should be below FM.
Posted 26 November 2014 - 23:38
Posted 26 November 2014 - 23:59
Only problem with that is Kimi is way too high on that list.
He should be below FM.
It's just what the model spits out based on the raw results. It doesn't take into account all the pitstop gaffes, compromised parts on the cars, and all the potential points that massa could've gained if he weren't in all those accidents.
I think the model somewhat forgives kimi for being alonso's teammate and also because the F14T really was THAT bad.
Also remember that this model is somewhat based on the comparative teammate history (it's also the reason why bianchi and chilton couldn't be ranked), so alonso is already in high standing, therefore losing to alonso wouldn't reflect as poorly as losing to bottas, who actually was comparable with maldonado in 2013.
Posted 27 November 2014 - 06:09
Brilliant. Appreciated the link.
I agree that Hamilton and particularly Alonso are the best. I feel both the Mercedes drivers were ranked slightly too low (points-wise) as they are both affected by Hamilton's previous troubles in 2011 and 2013.
Alonso is undoubtedly one of the best drivers ever, even if he retires with two titles I don't question to put him in the top 6 or 7.
Posted 27 November 2014 - 06:54
Wow
Posted 27 November 2014 - 07:13
Wow
Shouldn't happen to a top driver.
"Before 2014, Alonso’s brilliant career already put him among the all-time greats. The model rates him the 3rd greatest driver since 1950, behind only Jim Clark and Jackie Stewart, and marginally ahead of Michael Schumacher. This year further cemented his legacy......Overall, Alonso beat Raikkonen 16-1 in races and 16-3 in qualifying. In qualifying, Alonso was on average 0.53 seconds ahead. In races that both finished, Alonso was on average 33 seconds ahead."
Here we have them again - the famous 6/10s. Alonso has gotten a bit rusty, as he managed only 0,53 secs over Kimi. An average of 33 seconds ahead of your team mate Raikkonen - wow!
Posted 27 November 2014 - 08:05
I have the same top 4 for this season, it is almost impossible to separate alo, ham ,ric and bottas this year.
I think alonso and ric are probably my top 2 drivers this year
Bottas did nothing special to put him in the same category as alo, ham and ric this year.
Edited by Radion, 27 November 2014 - 08:07.
Posted 27 November 2014 - 08:07
Based on previous seasons as well as the current season, ranking him below Kimi is quite harsh. While Vettel scores >70% of Ricciardo's points and mostly put up a fight with roughly equal reliability adjusted qualy and at least finishing near Ricciardo in races most of the time, Kimi is below 35% and was only in some cases close in races. Of course, it then depends on how you rank Ricciardo and Alonso and the latter should have a historically higher rating while Ricciardo is coming from nowhere (in fact Alonso scored 3rd best driver of all time until July this year in those ratings, Rosberg 7th, Vettel 8th, Hamilton 12th, Ricciardo only 80th), so mathematically it makes sense but looking at the season it is hard to agree. I would rank Ricciardo ahead of Alonso this season, but even if you disagree I don't think it is really such a gulf as 3rd best vs. 80th best F1 driver looking at only this season. Generally, the model does not take changes in driver skills into account (quoting the author: "The model does not explicitly account for changes in driver performance across the career arc, so some drivers may benefit by comparison to teammates who are rookies or well past their peak form"), so a rising driver like Ricciardo and his team mate is undervalued.
Nevertheless a great effort - just wanted to point out some limitations.
Posted 27 November 2014 - 08:38
Wow
Posted 27 November 2014 - 08:49
That same model ranks Frentzen and Watson higher than Ayrton Senna.
Posted 27 November 2014 - 09:16
Posted 27 November 2014 - 09:32
Here we have them again - the famous 6/10s. Alonso has gotten a bit rusty, as he managed only 0,53 secs over Kimi.
Posted 27 November 2014 - 09:42
His model sucks - all time list of his places Alonso as 3rd of all time, Nico Rosberg as 7th, Kimi 14th - while Senna is 19th and Prost 16th, Damon Hill 35th, Graham Hill 36th and Nigel Mansell 46th
Oh, Gilles Villeneuve is 51st, Nelson Piquet 55th, Hakkinen 56th, Brabham 57th well below alltime greats like Massa, Fisico, Trulli, Heidfeld, Ralf Schumacher and Frentzen
Edited by velgajski1, 27 November 2014 - 09:49.
Advertisement
Posted 27 November 2014 - 10:00
The disclaimer:
"All models should be as simple as possible and all models have their limitations, which must be kept in mind when interpreting their results. This model is about the simplest you could propose to reasonably address the question at hand. There are some important limitations.
It also fails to take into account team performance e.g poor pit-stops or mechanical issues than do not cause DNFs or DNSs.
The model attempts to simplify something that is far too complex to be simplified. It is a little bit like the Red Bull aero department, using a 50% scale 2011 Hispania to represent the RB9.
Edited by OO7, 27 November 2014 - 10:27.
Posted 27 November 2014 - 10:08
Good model. I agree Kimi still had top 10 performance this year. He had lots of bad luck, he would score podium in Monaco and in the last two races he was on Alonsos pace.
Posted 27 November 2014 - 10:12
Good model. I agree Kimi still had top 10 performance this year. He had lots of bad luck, he would score podium in Monaco and in the last two races he was on Alonsos pace.
Posted 27 November 2014 - 10:12
His model sucks - all time list of his places Alonso as 3rd of all time, Nico Rosberg as 7th, Kimi 14th - while Senna is 19th and Prost 16th, Damon Hill 35th, Graham Hill 36th and Nigel Mansell 46th
Oh, Gilles Villeneuve is 51st, Nelson Piquet 55th, Hakkinen 56th, Brabham 57th well below alltime greats like Massa, Fisico, Trulli, Heidfeld, Ralf Schumacher and Frentzen
There are some reasons for this though so everything needs to be taken into perspective.
The model obviously has its flaw but when you frame it in the right context, it sits well. Ralf Schumacher was actually a pretty special driver, he was actually a much better qualifier than Montoya in their time together.
There are obviously some odd entries but even Watson high (because he beat Lauda as teammates) and Frentzen have a lot of logic behind.
Senna and Prost seem odd exclusions from the top 10 but no model is perfect. Enjoy the rest of the analysis if you can't enjoy the ratings as its excellent!
Posted 27 November 2014 - 10:16
The model is interesting, but as has been said many times before, team mate cross comparisons don't really work.
Posted 27 November 2014 - 10:24
His model sucks - all time list of his places Alonso as 3rd of all time, Nico Rosberg as 7th, Kimi 14th - while Senna is 19th and Prost 16th, Damon Hill 35th, Graham Hill 36th and Nigel Mansell 46th
Oh, Gilles Villeneuve is 51st, Nelson Piquet 55th, Hakkinen 56th, Brabham 57th well below alltime greats like Massa, Fisico, Trulli, Heidfeld, Ralf Schumacher and Frentzen
I can't put my finger on what's exactly wrong with the model, but I suspect it has to do with the huge correction factor he uses based on car competitiveness, which he doesn't keep in check with corresponding increased uncertainty values. Alonso's 2009 season (9th) alongside a nobody being rated as the best performance of the last 20 years is a very good indicator of this problem. The bigger his correction credits are, the higher their error bars should be, but they are never even mentioned in any of his posts.
Edited by paipa, 27 November 2014 - 10:25.
Posted 27 November 2014 - 10:25
Something else that is interesting to note. Nico Rosberg was ranked 7th in the 2014 driver rankings and 7th in the all-time greatest rankings. Lewis Hamilton was ranked 2nd in 2014 standing and 12th in the all-time standing despite beating Nico two years in a row.
Posted 27 November 2014 - 10:30
Something else that is interesting to note. Nico Rosberg was ranked 7th in the 2014 driver rankings and 7th in the all-time greatest rankings. Lewis Hamilton was ranked 2nd in 2014 standing and 12th in the all-time standing despite beating Nico two years in a row.
Probably because he beat MS who's ranked all-time 4th.
Posted 27 November 2014 - 10:31
There are some reasons for this though so everything needs to be taken into perspective.
- Gilles Villenueve lost out to Scheckter and his results were never as good as they should have been (he was a rough diamond who often retired)
- Piquet lost out to Mansell, who was hammered by Elio de Angelis when they were teammates (probably due to Lotus' policy of number one drivers and Mansell's disruptive relationship with the team)
- Hakkinen lost to Herbert when they were teammates (Herbert got preferential treatment and parts. Herbert also lost against several teammates too). Hakkinen barely beat Coulthard, who was beaten by Webber and Hill comfortably when they were teammates. Admittedly at the start and end of his career but the model doesn't take age/stage of career into perspective which is a flaw
- Brabham didn't really have many competitive teammates and raced until he was very old - some of his latter races probably distort his ranking
- Rosberg got mega points by beating Schumacher who is rated 4th in the rankings
- Senna didn't have too many strong teammates (Berger isn't rated too highly in the list as he lost to Mansell and other drivers). Obviously he had Prost who I don't understand - how can Alain by that low when he beat the most champion teammates of all-time? He was the driver with the second most wins, the second most points etc.
The model obviously has its flaw but when you frame it in the right context, it sits well. Ralf Schumacher was actually a pretty special driver, he was actually a much better qualifier than Montoya in their time together.
There are obviously some odd entries but even Watson high (because he beat Lauda as teammates) and Frentzen have a lot of logic behind.
Senna and Prost seem odd exclusions from the top 10 but no model is perfect. Enjoy the rest of the analysis if you can't enjoy the ratings as its excellent!
Coulthard beat Webber in 2007, Webber beat Coulthard in 2008 because he got involved in so many accidents.
But any model that rates Frentzen as being better than Senna you have got to question
KK looks underrated, KR overrated in particular.
Posted 27 November 2014 - 10:31
Probably because he beat MS who's ranked all-time 4th.
Hence the flaw.
Posted 27 November 2014 - 10:35
Hence the flaw.
You know what to do: Get a washed up WDC, and you get promoted to the best driver ever Flawed system
Posted 27 November 2014 - 10:36
Posted 27 November 2014 - 10:42
And we are surely free to improve this model...if we can.
I don't believe such things can be accurately modeled. Also the creator has stated that it is a simplified model, which is a significant problem in itself.
Edited by OO7, 27 November 2014 - 10:49.
Posted 27 November 2014 - 10:44
Posted 27 November 2014 - 10:45
There are some reasons for this though so everything needs to be taken into perspective.
- Gilles Villenueve lost out to Scheckter and his results were never as good as they should have been (he was a rough diamond who often retired)
- Piquet lost out to Mansell, who was hammered by Elio de Angelis when they were teammates (probably due to Lotus' policy of number one drivers and Mansell's disruptive relationship with the team)
- Hakkinen lost to Herbert when they were teammates (Herbert got preferential treatment and parts. Herbert also lost against several teammates too). Hakkinen barely beat Coulthard, who was beaten by Webber and Hill comfortably when they were teammates. Admittedly at the start and end of his career but the model doesn't take age/stage of career into perspective which is a flaw
- Brabham didn't really have many competitive teammates and raced until he was very old - some of his latter races probably distort his ranking
- Rosberg got mega points by beating Schumacher who is rated 4th in the rankings
- Senna didn't have too many strong teammates (Berger isn't rated too highly in the list as he lost to Mansell and other drivers). Obviously he had Prost who I don't understand - how can Alain by that low when he beat the most champion teammates of all-time? He was the driver with the second most wins, the second most points etc.
The model obviously has its flaw but when you frame it in the right context, it sits well. Ralf Schumacher was actually a pretty special driver, he was actually a much better qualifier than Montoya in their time together.
There are obviously some odd entries but even Watson high (because he beat Lauda as teammates) and Frentzen have a lot of logic behind.
Senna and Prost seem odd exclusions from the top 10 but no model is perfect. Enjoy the rest of the analysis if you can't enjoy the ratings as its excellent!
His analysis is pretty good, but the model is really bad / pointless. Simply ranking first by % of titles won followed by % of GP's won followed by % of points won would give much more 'accurate' (more consistent with what people think in general) results over driver careers in my opinion. His model is something I'd use as corrective for drivers that had exceptional teammate results but didn't have that impressive number of titles - prime example for that would be Alonso. So something that would enter into overall model with 20-30% of weight with 70-80% being based on real results over career - favoring titles, then wins, then points.
Edited by velgajski1, 27 November 2014 - 10:52.
Posted 27 November 2014 - 10:47
Something else that is interesting to note. Nico Rosberg was ranked 7th in the 2014 driver rankings and 7th in the all-time greatest rankings. Lewis Hamilton was ranked 2nd in 2014 standing and 12th in the all-time standing despite beating Nico two years in a row.
Posted 27 November 2014 - 10:51
Yeah and this is where the "all time" version of the model falls down. Rosberg gets major uplift for beating Schumacher, because Schu was a god. But in reality, he beat a faded Schu, not the godlike version.
Hamilton gets marked down for being beaten by Button in 2011 etc.
Despite beating Alonso (yes I understand they were tied in points) who is rated 3rd all-time.
Posted 27 November 2014 - 10:55
Yeah and this is where the "all time" version of the model falls down. Rosberg gets major uplift for beating Schumacher, because Schu was a god. But in reality, he beat a faded Schu, not the godlike version.
Hamilton gets marked down for being beaten by Button in 2011 etc.
I agree. The model makes a very flawed assumption that driver form is constant across the entire span of a career, which I don't think is true even for the most consistent drivers.
Posted 27 November 2014 - 10:56
Well yes. Alonso was not beaten by Hamilton massively, but only very closely.Despite beating Alonso (yes I understand they were tied in points) who is rated 3rd all-time.
Posted 27 November 2014 - 10:57
I agree. The model makes a very flawed assumption that driver form is constant across the entire span of a career, which I don't think is true even for the most consistent drivers.
Especially in cases like Schumacher who probably lost huge amount of points for his Mercedes years - while Rosberg gained those.
Advertisement
Posted 27 November 2014 - 11:25
Ridiculous method to assess drivers' performance,just like any other primarily based on results/points.
Posted 27 November 2014 - 11:48
Especially in cases like Schumacher who probably lost huge amount of points for his Mercedes years - while Rosberg gained those.
I'm working on an all time list myself and worked out that if Schumacher hadn't made his comeback, he would have been 3 places higher within the top 10 group. It's a dilemma because if we only look at a drivers peak seasons, that can introduce bias of ignoring when they weren't at their peak. We don't know how much Schumacher was off his peak. I think Rosberg vs Hamilton has shown, perhaps not as much below as we thought at the end of 2012.
Only focusing on peaks can also be unfair to those drivers whose peaks may have been a little below a rival, but they were more consistent. For example, Gilles Villenueve could put in some amazing performances, but in the 37 races where he did not experience car trouble or a disqualification or similar, he crashed out of no less than 10. And it wasn't as if most of them came early in his career, either.
Posted 27 November 2014 - 12:44
This guy has been promoting his flawed model for the last year or so.
Irrespective of the statistical methods he's used to produce his lists the fact is if you asked those in the paddock who was a better racer Senna or Heinz Harald Frentzen or Nico Rosberg very few people would say Senna was a poorer driver than either.
Likewise Raikkonen, Vettel and Hamilton were all rated below Rosberg in his initial list.
If it tastes wrong it probably is wrong.
Adding verbiage to try and explain away flaws in the model isn't a convincing scientific method when so many other factors are ignored in the analysis.
Edited by BCM, 27 November 2014 - 12:44.
Posted 27 November 2014 - 13:10
I don't understand how critical people are of these things, be it rankings about drivers or any other type of sportsmen/teams (for example the widely mocked FIFA Rankings), it's well clear this is an attempt to estimate performance through a mathematical model, he himself admits its flaws, it's not like it's a madman shouting "WE HAVE SCIENTIFICAL PROOF FRENTZEN WAS BETTER THAN SENNA". No method will ever be 100% reliable, but it can help trim out cognitive bias.
I think it's highly interesting, very well thought of and very well explained, my only criticism is that indeed age/experience is a missing factor in his equations. I think it would be very possible to introduce this factor and it would give out more accurate overall results, although it might introduce some other biases (for example overestimating highly prepared rookies like Hamilton, who didn't improve much on their rookie seasons), but smaller than the bias we have at the moment due to the lack of this factor. And not only would it help the results, a by-product of introducing this factor is that maybe we could also analyse how important is the level of experience, what is the "ideal" age for performing in F1, and how sharp is the decline at a veteran age and/or the handicap of going in at an early age.
Posted 27 November 2014 - 13:45
There are quite a few cases where 1 drivers rating is boosted significantly due to 1 year or 1 team mate. I guess it would be possible to somehow make more recent years of a drivers career be more important when comparing 2 team mates, that way making it less likely that 1 driver makes a significant gain from being a driver past their best etc. But to be honest considering it's just a model I think it does quite a good job
Posted 27 November 2014 - 18:34
Despite beating Alonso (yes I understand they were tied in points) who is rated 3rd all-time.
Straight question - should 2007 mean that Hamilton is always considered the equal or better of Alonso?
Posted 27 November 2014 - 18:49
Straight question - should 2007 mean that Hamilton is always considered the equal or better of Alonso?
Not solely on the basis of 2007 which, all the dark and complex politics of that season aside, we don't know if Alonso was at his peak. Drivers form goes up and down. This is countered by it being Hamilton's first season in F1. Magnussen and co might say "not as difficult to be on it in your first season when you have all that testing." In my opinion, individual seasons such as 2007 can form the basis for distinctions between very closely matched and similar careers, but can't be the sole basis for answering your question If so, we could end up with teammate comparison chains like this based on individual seasons: Schumacher<Rosberg<Hamilton<Button
Edited by hittheapex, 27 November 2014 - 18:50.
Posted 27 November 2014 - 18:51
Posted 27 November 2014 - 19:00
Not solely on the basis of 2007 which, all the dark and complex politics of that season aside, we don't know if Alonso was at his peak. Drivers form goes up and down. This is countered by it being Hamilton's first season in F1. Magnussen and co might say "not as difficult to be on it in your first season when you have all that testing." In my opinion, individual seasons such as 2007 can form the basis for distinctions between very closely matched and similar careers, but can't be the sole basis for answering your question If so, we could end up with teammate comparison chains like this based on individual seasons: Schumacher<Rosberg<Hamilton<Button
Yeah, I don't think it is definitive. This model (the thread subject) is in good accord with my rating of the drivers. It is not just that Alonso is the best, but that there is big gap to Hamilton with smaller gaps then between successive drivers, who are listed in the right order for 2014 IMO.
Obviously the career length version has anomalies (how can Prost be so low?) but maybe these don't factor into the one year version.
Posted 27 November 2014 - 19:01
in 2007 we had Bridgestone tyres and refueling. People need to stop with that theory : X driver beats Y driver ; Z driver beats Y driver = Z driver would match or beat X driver.
Posted 27 November 2014 - 19:14
Very impressive!