A comparison between 3 mathematical models

[cider_and_toast]

You say Lauda withdraws, I give you Hunt deliberately breaking his drive shaft by rubbing his wheel against the armco and squirting the throttle.

 

[Spinodontosaurus]

Mr Phillips sure has an interesting model

6. Hunt
17. Frentzen
25. Heidfeld

I think these are largely cases of drivers being perpetually underrated by fans relative to how well they actually performed, although Frentzen's placement has other factors affecting it which I'll get into. I'm not familiar enough with Hunt's career to talk about him, so I'll simply link you to his entry in Phillips' original article.

I feel more comfortable talking about Heidfeld and Frentzen, so I will.
The other two models in the article actually both rate Heidfeld higher than Phillips' model does, and if you examine his career it isn't hard to see why he is rated so highly. Heidfeld had 12 teammates over the course of his career, but 3 of them were only for a very limited time so we will ignore them (Kobayashi, Glock and Vettel combined had just 10 races as Heidfeld's teammate). The remaining 9 teammates were as follows, with the points tallies in brackets (all have been converted into the modern system):

• Alesi (8-3)
• Raikkonen (77-66)
• Massa (65-46)
• Frentzen (29-37)
• Pantano (7-0)
• Webber (68-63)
• Villeneueve (39-23)
• Kubica (395-352)
• Petrov (34-32)

Heidfeld outscored all of them bar Frentzen.
Imagine if Button never ended up at Brawn GP, and then McLaren afterwards - he wouldn't have been any worse a driver than he was in reality but he wouldn't have achieved anywhere near the level of success, and the general perception of his abilities would be much lower. That was Heidfeld throughout his entire career - a very strong driver, but never had the right car to show it, and his conservative driving style failed to win over many fans (unlike say Montoya).


Frentzen, rather like Heidfeld, was also only outperformed by one teammate (in this case Villeneueve). However, a big reason for Frentzen's high placement in Phillips' model is the 1999 season, where he outscored an old and demotivated Damon Hill 54-7. A quote from Phillips' 2014 article:

Hill was at this point just shy of 40 years old, so Frentzen’s ranking likely benefits from comparison to Hill near the end of his career arc. This can be explored by completely removing Hill’s 1999 results from the analysis, which drops Frentzen’s ranking to 57th and raises Hill’s ranking to 26th. This treatment is obviously unfair on Frentzen, given it ignores probably his strongest ever year, but it also suggests he may be overrated by the model based on this particular teammate comparison.

If we look at Frentzen on a season-by-season basis (based on Phillips' model), I think he deserves to be ranked ahead of Hakkinen, but behind Hill and perhaps comparable to Villeneueve. The consensus list of the 3 models actually supports this placement; Hill is 24th, Frentzen is 36th, Villeneueve is 42nd while Hakkinen is outside the Top 50.
We also see that 1999 is not too far out there compared to a couple of other strong seasons, including 2000 (beat Trulli 11-6) and 2003 (beat Heidfeld 13-6)

 

[teabagyokel]

Imagine if Button never ended up at Brawn GP, and then McLaren afterwards - he wouldn't have been any worse a driver than he was in reality but he wouldn't have achieved anywhere near the level of success, and the general perception of his abilities would be much lower.

Even more appropriate considering what might have happened if Alonso hadn't walled it in Budapest in 2006, added to a more permanent Honda collapse in 2008. Button becomes Heidfeld...

I hope he gets a win in Formula E soon.

A huge difficulty in these analyses is drivers' changing fortunes. Before 2013, Nico Rosberg hadn't had a comparible team-mate since his debut season in 2006. (Nakajima plus Wurz and Schumacher after long breaks.) There was a tranche of Red Bull drivers who had also only races each other. Effectively, these judgements all hinged up to 2013 solely on Mark Webber!

Hakkinen is also an interesting case. He spent more than half of his career as a team-mate of David Coulthard only. Difficult to judge him in the peak years of both drivers, is it not?

Basically, it can't be done.