Probability of Win 2013

[Schummy]

It is the probability of win in a next generic GP, i.e. not having in account the possible special characteristics of a particular GP (for example, Monaco).

Currently I calculate it using the 1st and 2nd places achieved so far. A discount factor is used to weight progressively less the older GPs, and another factor considers the possible wet weather event.

     A M C B S    PoW
ALO  1 0 2 0 2 => 35%
RAI  2 0 1 1 1 => 33%
VET  0 2 0 2 0 => 27% 
WEB  0 1 0 0 0 =>  6%

Webber's 6% is surely a bit less actually because in the event of he being 1st and Vettel being 2nd, we know they are going to swap places. Conversely, Vettel real probability should be more than 27%. Anyway this effect is small, because that scenario is of low probability.

Hamilton has not yet any chances, apparently. His car is too bad fitted to races. But the potential is there, Lewis has had 2 third positions already.

Now, let's see how those PoWs translate in future wins:

    PoW  EFW  STD  IFW
ALO 35% 4.90 1.78 [2..8]
RAI 33% 4.62 1.76 [2..8]
VET 27% 3.78 1.66 [1..7]
WEB  6% 0.84 0.89 [0..2]

EFW = Expected future wins (in the season)
STD = Standard deviation
IFW = Interval of future wins (about 95% probability, Tchebycheff 2 sigma)

Alonso, Kimi and Vettel are very near in terms of prognosis. Poor Mark has an expectation of barely one win in the season (if Marko allows it), but his interval reaches 2 wins.

All these data is considering just the real (sample) data in the season. It doesn't take in account theoretical a priori reasons.

[Bononi]

What about Gutierrezz ???

[Schummy]

ummmm...

Well, the reason Guty doesn't appear here is that there is not negative probabilities.

[Aysedasi]

Some (nasty unkind) people might suggest that (notwithstanding the stats) that applies to Webber as well.....

[Adam43]

Originally Posted by Schummy

Webber's 6% is surely a bit less actually because in the event of he being 1st and Vettel being 2nd, we know they are going to swap places. Conversely, Vettel real probability should be more than 27%. Anyway this effect is small, because that scenario is of low probability.

You mean it is unlikely Webber will be ahead of Vettel and in a 1-2?!

Your result feels right. It is about three drivers, Alonso, Vettel and Raikkonen.

[Schummy]

There is just 6% of MW being 1st, so it is even less probable having MW 1st and SV 2nd. I suppose if MW is 1st then probably RBR is a very good car in that race :-) , so it would not surprising that SV was 2nd. But anyway it is a low probability combination.

[steve_r]

Do you really believe that Webber will move over for Vettel after what happened a couple of races ago.

[Schummy]

I am pretty sure he will do it. Otherwise maybe the team will "adjust" his car or pitstop :-) . Anyway the chances of him disobeying orders are small, because it must happen in the last laps to be successful and, of course, he has to want it.

He is an employed and he gets a lot of money, I am not sure he will have a rapture of romanticism. Maybe I am wrong.

[Razzzor]

there will be no team orders next time

[Schummy]

Right, they directly will use the remote control.

[Bononi]

Originally Posted by Razzzor

there will be no team orders next time

Originally Posted by Schummy

Right, they directly will use the remote control.

Brilliant !!!

[Schummy]

Markimedes's Principle:

"Any second driver is hindered by a force proportional to his displacement in the team."

[JeremySmith]

Originally Posted by Schummy

Right, they directly will use the remote control.

You are kick arse..

[daallo]

Originally Posted by Schummy

ummmm...

Well, the reason Guty doesn't appear here is that there is not negative probabilities.

He has had a better start than Grosjean, Maldonado, Sutil or Ricciardo so there can be a chance of him fighting for a win in a couple of years and for grabbing a podium next year.

[Schummy]

I am sure he will improve, if they give him the future opportunity; how much improvement I don't know. This year he has not been doing a good job against his team mate (1-3 in races and 0-5 in qualifies, with an average of three rows behind); he is very young and still inexperienced. And the car is not very capable of big things, neither.

Next years, who knows? For example, Grosjean has become a pretty fast driver and now he is even consistent. But, by now, even Roman has not a realistic chance of wins neither!

[Bononi]

Who knows, who knows ?? That's the motto of F1 !

[daallo]

Yep that's true, or some drivers may have a great start and then just fade. I actually thought he did realy bad the two races just before Spain and was one of the worst, but it was in Spain where I think he was one of the best of the race and he's talented and since his first days he has shown he knows how to learn an improve so let's wait and see.

[Schummy]

Evolution of Probability of Win across the season (so far).



Main points:
- Raikkonen is losing momentum as probable race winner. Still he continues in a high position.
- Alonso and Vettel evolve complementarily, in the long term they maintain positions, not increasing or decreasing them.
- Rosberg's sudden surge puts him in the top position as likely race winner, but just by a minimal margin against FA, KR and SV.
- Massa's chances as a winner looked relatively good at the start of the season, with a fast car and a good driving form. As the races have past by, he is losing likelihood as a race winner. Currently is a very remote possibility.
- Hamilton has not never given the sensation of being a race winner, so far. He can in podiums, but, for one reason of another his races never had all the pieces together. But it is difficult to imagine that Nico is a top candidate for wins and Lewis is almost out of that picture. Surely something have to change sooner than later in LH's races.

Rosberg's position is a bit misleading. He has had two convincing poles and one very decent win in Monaco. But Monaco is not really a car race, is not it? Therefore, it's difficult to know the real reach of current Mercedes's pace in races.

The "structure" of the season is clear (... until Monaco): Three drivers are entangled between them and consistently above the rest. The novelty is Rosberg, who, apparently, is now between the best positioned for future wins.

[Schummy]

Actually the former graph was a kind of mess because two race got glued in one and it played havoc with data.

This a better version, with that problem fixed (hopefully).



Most of remarks are valid, but now some of them are not. These are "new" added remarks:

- Raikkonen is keeping fairly consistent, even if he didn't reach the high expectations (for wins) in the beginning of the season.

- Alonso is not doing badly in terms of win chances, but he is not so good in the title points race.

Basically, expected winners are Vettel, Rosberg, Raikkonen and Alonso (at around 20%-25% each one), with a side shoot by Hamilton (8%) and a rare opportunity for Webber (3%) and frankly unlikely for Massa (1%).

Mercedes, surprisingly, leads the team ranking as probable race winners, but actually is a tight tie with Red Bull:

MER 31%
RBR 29%
FER 20%
LOT 20%

[Schummy]

The entropy of the season is increasing:

1st race: 3.56
2nd race: 4.02
3rd race: 4.97
4th race: 5.06
5th race: 5.36
6th race: 5.36

With entropy I mean the uncertainty on determining the next race winner. It is a function on probabilities of race win, given in the former post. The increment in entropy is good news, as it means races are more unpredictable (at least in terms of race winner).

As long of different race winners appear (and other good classifications for more people) the entropy will keep increasing. I suppose there is not much more entropy remaining, because there is not many more potential race winners (Hamilton, mainly).

The numbers in the table are actually "2^entropy" (2 raised to entropy) and it intuitively means approximately the number of potential winners in the next race. We started the season with 3 or 4 possible race winners and now we have 5 or 6 (according to the entropy).

[stripedcat]

Other thoughts:-

I think that McLaren will probably come on strong when F1 gets to Hungary. Maybe a win there?

Peréz might get at least 2 wins.

Hamilton - I get the feeling he'll get podiums but no wins this season. Only, an unexpected set of circumstances might net him one.

Likewise, with Rosberg scoring another win. The Mercedes is strong in qualifying, but not that good in the race. He'll get podiums, but no more wins.

Webber - Yes, I think he'll pick up at least 2. Vettel allowing!

Kimi - I think he might pick up a few more, but I do think the general consensus is that the Lotus has lost a lot of its speed. I don't think he has that strong a car now.

I do think it is going to come down to another Alonso and Vettel battle.