Brought to you by TheJudge13 contributor landroni
Arguments rage over whether Sauber’s worst ever season in F1—with no points scored, which incidentally mirror Ferrrari’s worst season in over twenty years, with no wins scored—was caused by the catastrophic and overweight Swiss chassis or by the anaemic reddish Power Unit imported from Italy. Responsibility can probably be partly shared here. And in all likelihood Sauber must publicly absolve Ferrari since they’re walking a financial tightrope, and with all the bloodletting at Maranello this year, better keep happy the powers that be. This said, one element that has been forgotten in this debate is the drivers.
The kind people at F1Metrics have developed a mathematical model that, while it comes with certain limitations, attempts to disentangle car performance from driver performance, and tries to rank drivers by doing pairwise teammate comparisons so as to identify relative driver performances.
For the nitty-gritty methodological details, see the peer-reviewed paper. Many may sneer at a scientific approach to quantify driver performance, wildly preferring their own subjective bias confirmation exercises, but the model predictions do come up with several results that would seem to conform to (one would say) mainstream beliefs.
For example, given equal machinery (e.g. better cars than he has been saddled with for the past 7 years or so), Fred would have easily matched one Michael Schumacher in number of WDCs won. Or that Finger Boy would have been no more, as without the über-dominating Newey toys (e.g. in equal cars) Seb wouldn’t have won a single WDC yet, sparing onlookers from his fingery habits.
Unsurprisingly, the model lays Sauber’s predicament this year squarely at its (pay) drivers’ feet. If you think that Magnussen, Perez, Maldonado and Ericsson have been getting a lot of slack these years for their performances in an F1 car, think again. Sauber’s duo, Adrian Sutil and Esteban Gutiérrez, have been gloriously ranked dead last among the 2014 crop of drivers, 19th and 20th, respectively, behind the aforementioned drivers.
Sutil also holds the dubious record of crashing out of races more often than any other current driver, with 1 crash every 5 starts. Gutiérrez is not far behind on that record, being 3rd last, only just in front of Grosjean. (Still remember Grosjean’s crashy-crashy form from early in his career?)
And if you thought that was bad, the duo are still behind our beloved Crashtor, aka Pastor Maldonado, who comes in 5th last. If you put Sutil’s crashing performance in a historical perspective, he comes in neck and neck with the crash-prone Andrea de Cesaris, affectionately nicknamed “de Crasheris”.
Damningly, though, if you assume identically skilled drivers (e.g. equal driver performances), then Sauber’s paltry 2014 performance with 0 points and behind Lotus and Marussia suddenly transforms, according to the model, into a 34 points tally and in front of the Lotus. Put differently, stick Grosjean and Maldonado into a 2014 Sauber, and all of a sudden the Swiss heavyweight chassis would out-perform the Enstone prodigy with fridge-like aerodynamic characteristics.
In this light, it makes you wonder what Sergio Marchionne and Maurizio Arrivabene were smoking… ahem… thinking when they decided to enrol Esteban into Ferrari’s car development programme…
That was just fascinating! I’m not sure if Ericsson will manage to do any better though…I was astonished Sauber had hired him…until I remembered that he had sponsorship:)
I love how high they rated JEV…I think he’s capable of doing what Daniel has done but not sure if he’ll ever get the opportunity…
Lastly, was watching Super Speedway last night and Mario Andretti made a really interesting comment that sometimes you can have a brilliant car and a brilliant driver but the two don’t mesh…there’s more to racing than mathematical models and statistics…
Thanks for such an interesting article!
great insite, Jennie. IMHO, one has to look at Hill, Clark, Surtees, Stewart, Sheckter, AJ, Parnelli, Gurney, Mansell, Andretti and a bunch more to find the true heros of racing. guys who could and DID go to many multiple venues and seriously kicked ASS over and over! I know I am gonna feel the wrath of many TJ13 commentors, but I seriously look at Senna, Shummi, Seb and Hami as mere 1 string dueling banjo fiddle players along a Southern US Cracker Hillbilly Kuntry river… oops. my bad, but that is where I stand!!!
get er done, or pretend to…
It would probably serve you well to read a Senna biography before you place him amongst the drivers you mentioned. He began car racing in 1981, won the titles, in 982, won the titles. In 1983 he took the British F3 title and won in Macau. Even Hamilton who won on his way up to F1 needed a second season to win.
Whatever he raced in, he won. In his FF2000 season in the UK, he took part in a Talbot Sunbeam race and beat the opposition. In 1984, he took part in the inaugural opening of the Nurburgring with a number of ex World Champions all driving Mercedes 190-E and won.
There is an actual article of Senna trying rally cars in Wales in his Lotus days, and he tried RWD, FWD and 4WD cars. His learning was so rapid that professional rally drivers declared he would be World Championship level with very little practice.
I think people forget that when Senna was the dominant driver in F1, the sport was changing and drivers didn’t have the opportunity to compete in other series.
As an aside, he tested an Indycar in 1993, Fittipaldi was the teams driver and had won four races that season. On fresh tyres he set a time of 49.7 seconds. Senna went out with same car, took a few laps to get used to sequential gear-changes etc and then on same tyres, he lowered lap record to 49.12. In 28 laps..
Carlo.. it was always interesting how Williams didn’t sign him up straight after that first F1 test.. instead persevering with their favourite Laffite.. imagine Senna in the 86/87 Williams! Two titles racked up already before meeting Prost…
The comparison with Hamilton would be – how impressive would Lewis’ career be if he started in 2010? Or perhaps 2009 was that year with Toleman. 2007 and 2008 would be lost to karting/climbing the ladder.
Senna could probably have done an Alonso/Kimi/Button/Massa and climbed up to F1 faster, although he still did only 3 years total after karts.. something which probably let those 4 do what they did in following in his footsteps to F1 in 2/3.
And of course getting into the 2014 Merc is about the same time when Senna finally got into the 1988 McLaren, after coming close to the title on merit, but being let down by the car…..
of course, your examples of Senna’s speed are dead spot on! embedded in my point of view is that they were one-off forays once he reached F1. indeed, the time had passed when our heros could jump around to win the WDC, Indy 500, LeMans, F5000, CanAm, TransAm, etc.. potential to have been one of MY all time greats? you bettcha. proof at being an all time great? not so much
Thanks Jennie!
I was also very intrigued by JEV’s ranking… Vergne mostly matched Ricciardo race-wise, even if he was dominated in quali, despite Ricciardo having a head-start. A pity his career has stalled, then, and should be fun to see a matured JEV in an F1 car again. Remember that originally Red Bull evaluated JEV’s long term potential higher than that of Ricciardo!
I have always been puzzled by Gutierrez’s lack of speed in Formula 1. He was fast in junior series and I rated him always higher that Perez, mainly because he seems to be a lot more intelligent than Perez. But in Formula 1 he just looks slow.
I have a feeling that for Gutierrez it really was a case of “way too early”. It seems to me that he never had the psychological maturity to fully accommodate to the rigors of F1. As the Sky crew pointed out once, it always seemed like the car was one step ahead of Gutierrez, instead of it being the other way round.
And in this business slight lapses in confidence can be deadly career-wise.
Same here.. I always wondered if he was a momentum based driver, hence excelling in small formula cars, but not as much in the high end equipment. You could argue it’s the opposite for Perez, who spent much of his youth in shifter karts.
Once upon a time Sutil was fast. He never seemed to recover from his enforced vacation. But it was Perez who really did them in as he took Adrian out at COTA where they might have actually had a chance to score a point at least.
Gutierrez seems to have just gone backwards.
The crashing thing is very interesting, I would be curious to see how this stacked up in Sutil’s career, crashwise and DNF wise in a year over year comparison. Would probably be an entertaining graph.
Great article landroni!
Thanks, Matt.
There is also Monaco. Both Sutil and Gutierrez found fit to crash themselves on a day when Bianchi managed to eke it out in the points. Bad idea!
http://thejudge13.com/2014/06/16/f1-victims-of-circumstance-monte-carlo-2014-monacogp/
I always wondered if that arose from his late entry to racing – 14 when he first sat in a kart. To then make F1 in ten years is no mean achievement – it takes some drivers 10 years in single seaters after 10 in karts to make that level. Max V would have already been in F1 had he taken the same approach 😛
It’s notable in some top performers who lose that ultimate control under pressure – Lee Westwood for example also started Golf around the same age. When the pressure is on.. they need that extra ‘done it all your life’ experience to come out of the unconscious and rescue them. Max V could be a good marker for that when he kicks arse on his debut.
MSc, Kimi, Sutil.. these guys totally rely on their feel and proprioception.. simulators are useless for them.
But yes, 2012 out really took that edge off his pace… being beaten by Di Resta in 2013 would have seen Paul continue in F1 in most circumstances.
The F1 Metrics model is crap. HHF ahead of Senna? Rosberg (Nico) ahead of Senna?
Models are just that – models. Being peer reviewed by publishing it doesn’t make it a good model.
“Essentially, all models are wrong, but some are useful.”
All statistical models are necessarily simplifications of real life conditions and interactions, simply because reality is so bewilderingly complex. But they can still be useful to describe reality. Just ask any serious F1 team! For instance, there are 1001 things that could go wrong with an average, another statistical tool, but we’re still using that in daily real life, don’t we?
This said, don’t get me wrong: I also tend to use “peer-reviewed” as a slur… But this model does have some nice features. It explains 82% of the variance in the data. It arrives to a rough, reasonable ranking of drivers that can be corroborated using a qualitative analysis. And it allows drawing some general conclusions which fit in nicely with mainstream beliefs. Some of its features will necessarily be rough around the edges, as are the points that you raise.
As mentioned in the beginning, this model comes with several limitations. One of its limitations is that it compares teammates at different stages of their careers, more or less irrespective of their current performance (vs career performance). So, as explained in the F1Metrics article, when Rosberg dominates Schumacher, a 7x WDC, for 3 consecutive years, then he is bound to be ranked highly by the model. Removing Schumacher’s results post 2010 from the model made Rosberg’s ranking fall to 15th, mostly as a result of his closely matching Hamilton at Merc. Since this would bump Hamilton to 11th, this would mirror what many think of Rosberg: very close to Hamilton’s performance, but only that very, very little bit slower overall.
As for HHF, the article addresses this head-on:
“Frentzen’s appearance at 17th is surprising, but arises from the fact that Frentzen was outperformed only once. Unfortunately for Frentzen, this occurred in what should have been his breakthrough year. Failing to capitalize on a dominant car scuppered his chances of racing for a front-running team again, and he subsequently spent his career in midfield.”
“Comparing Frentzen with Mika Hakkinen gives a lesson on the importance of timing for success in Formula 1. Most fans and pundits rate Hakkinen well above Frentzen, whereas my model does the opposite. It is Frentzen’s failure to perform in his one year in a top car that captures attention. As a hypothetical, imagine that Hakkinen had performed to the exact same level in each of his years of racing, but also imagine that he spent only 1997 in a top car, like Frentzen — a year in which he was outperformed by Coulthard.”
And before we start waxing lyrical about Senna (wink wink Jackal), remember that Senna operated in an era where competition was relatively weak, a fact controlled for in the model. Also Senna may have been the best driver over one lap ever, but overall his race performances were slightly less impressive. For instance, “in 1988 Senna outscored Alain Prost in the best 11 finishes that counted towards the championship, whereas Prost outscored Senna over all races”, meaning that taking results over the entire season season Prost would have been WDC, leaving Senna to a Alonsoesque 2x WDC. Over the course of their careers, Prost proved more efficient at translating performance into WDCs.
Just like Hamilton vs Rosberg today, Senna and Prost were ultimately very closely matched in their performance. Even if Senna obliterated Prost in qualifying 28-4, Prost proved more consistent in eking out results and consequently they were much closer 14-9 in races and 154-186 in points (150-163 in points that counted towards the championship). Another modern-day example example would be Button vs. Hamilton at McLaren from 2010-2012. Even if Hamilton dominated qualifying 44-14, he was ultimately outscored 657-672 by Button, through a combination of bad luck and inconsistent form.
And you know as they say: The more risk you take, the higher the potential peak rewards. But, you know, the more risk you take of falling flat on your face. Not an outlandish characterization of both Senna and Hamilton…..
Exactly, it was really only 2012 where Schumi was fully back to form and beating Rosberg in race pace (not to mention to Monaco pole position), just he got very unlucky with reliability.
Rosberg 15th, Hamilton 11th would fit much better with what most people feel, along with Ricciardo now smashing his way into the top 60. Frentzen was also at his best in the 1999 Jordan, and just didn’t gel with Williams’ autocratic nature.
Hamilton vs. Button is a good analogy for how they’d have fared now.. I think Alonso/Button will be another! Just like Alonso/Hamilton was too. Japan 2014 also backs up Button’s rating as the best wet weather driver.
One of its limitations is that it compares teammates at different stages of their careers, more or less irrespective of their current performance (vs career performance)
It’s a very interesting model, but that is undoubtedly its biggest weakness. Trouble is, correcting for that would be very complicated indeed, particularly for current drivers.
A separate score for the first couple of years of drivers careers would make sense, as it’s quite uncommon for them to reach peak performance before that. How you might treat the career twilight is quite a puzzle. Given the limited data on any single driver’s career, mucking around with it too much doesn’t leave you with a lot.
One striking analysis was the consistency of performance – on which Alonso scores every highly indeed.
In the end, the model does provide a very interesting perspective – another useful corrective to one’s own subjective judgments.
Btw, look at the calculations for performance in the wet – even more interesting, IMO.
MONEY! Lots of Mexican money. With a new sticker to discreetly place on the car.
Plus he has experienced the 2014 Lego motor that Ferrari produced and will be able to confirm improvements to drivability and power as they are dialed into the SIM and I would imagine he will get an on track test day at some point to evaluate how effectively the SIM recreates the real driving experience. Not to mention they also have JEV, who has done many many miles in the RedBull SIM, so can give feedback on how good Ferrari’s simulator is against what I would imagine to be one of, if not the best simulator on the grid. But Gutierrez is strictly business, cold hard cash.
Sure, they were definitely smoking Mexican pesos…
But I’m more worried by the signal that Ferrari is sending out: “We’re now ready to take on-board any and all unproven youngsters with truckloads of cash”. This is what the Caterham’s, Marussia’s and Sauber’s do, NOT Ferrari! Seriously. Ferrari has literally recruited someone who got dumped by Sauber, after several years of consistently unfulfilled potential!
Before Ferrari was always skittish about taking on-board unproven fellas. They even sent Bianchi to an extended internship at Marussia, for all the good that it did to him, without bothering to take him on as 3rd driver or whatever within Ferrari. And now look: Arrivabene arrives well, and suddenly Guttierez is recruited. It shall be fun to see Guttierez racing in a Ferrari should either Seb break a finger or Kimi have an extended AAA meeting; should bring back fond memories of Luca Badoer or Giancarlo Fisichella in a Ferrari…
And what next? Chilton, Sutil and Maldonado? I’m sure the smoke of British pounds and Bolívar fuertes can be intoxicating, but it may also start an uncontrolled fire…..
The irony is that if Ferrari had replaced Massa with Bianchi for 2012, they wouldn’t have needed to sign Raikkonen, would have saved a truckload of cash, and had a faster driver to boot, ready to now partner (and maybe even surpass) Vettel for the future.
PS. Why is the young stroll on the Ferrari junior line up? Nothing to do with his dad owning Mont Tremblant? 😛
True, however the only problem with that is that Sauber don’t have a simulator, which is probably what caught out Gutierrez in his attempt to break into F1 with them. I imagine Vergne will do sim work, while Gutierrez will do any track testing time.
Not to mention the strong possibility of a Vergne-Gutierrez, Sutil-Gutierrez or Gutierrez-Marciello line up at Haas in 2016. Vergne-Marciello would be a great line up to have.
For The First Time Ever In 22 Consecutive Seasons, SAUBER Ended 2014 With Zero Points.
They Should Have Kept Some Of Their Best Drivers From Recent Years.
Even MARUSSIA Got Some Points With The Same Engine – Live And, Hopefully, Learn.
GO, 44 !
The articles in that site are great, by the way.
I dunno… it seems to me that both of these guys, but Gutierrez in particular, have too small a sample, leading to the equivalent of very high variance. Had he scored, say 20 points, he might be ranked top 10.