[stat]Team mates[/stat]

Schummy · 23 Mar 2006, 15:34 (Ref:1558636)

With just two GPs there isn't a lot that can be said about how team mates are doing in direct comparison, but some initial data can be compiled already.

To get an idea about the perfomance of team mates I have noted three "parameters": grid position, fastest lap in race and race result.

The first of them inform about one-lap raw speed and perfomance under pressure (specially this year), second one is raw pace in race conditions (something than for some reason some drivers don't know how achieve) and finally the third is global racecraft (dealing with fights, strategy and pit stops, care of the car, etc).

Anyway, this set of measures is derived from the the most known stats about F1 (poles, fastest laps, wins).

So every GP we have three comparisons, and they give us a "score" (how many of them has won one driver over the other).

I am focused in the main teams: Renault, McLaren, Ferrari, Honda and Williams. The following table shows the results. I have put first the driver usually perceived to be the "number one" in the team.

Code:

Drivers BAH / MAL   TOT Chance AlSi Sign
FA-GF   3-0 / 1-2 = 4-2  69%   ---  ---
KR-JM   1-2 / 0-3 = 1-5  22%   0-3  ---
MS-FM   3-0 / 2-1 = 5-1  22%   3-0  ---
JB-RB   3-0 / 3-0 = 6-0   3%   2-1  3-0
MW-NR   2-1 / 2-1 = 4-2  69%   ---  ---

First column is score in Bahrain, second is score in Malaysia. TOT means the aggregate of those two GPs and "Chance" is probability that total score has been got just by chance.

Renault and Williams pairs results can be explained by chance, that is, for instance, Alonso has not proved so far he is "better" than Fisi (this season), or Fisi better than Alonso. That's ok, because we have only two GPs of data!

More or less the same in McLaren and Ferrari, with some more weight towards MS and JPM, but nor significant.

In Honda, Button is outperforming Barrichello in an almost significant way. That result (6-0) is dificult to explain by fluke (but still possible).

The last two columns in the table are a bit "special". "AlSi" means the result that has to be achieved in AUS in order to beat his team mate almost significantly (5% chance). "Sign" means the result to beat his mate significantly (1%), that is, definitively (at least theoretically!).

So, if JB beats 3-0 Rubens in AUS, Jenson will have beaten RB significantly in this season (so far and probably for the rest of the season).

Menelaos · 23 Mar 2006, 15:36 (Ref:1558638)

Very nice stats, but as always they don't show the true picture, in my opinion. We eliminate too many parameters.

But still, interesting

Rennen · 23 Mar 2006, 18:23 (Ref:1558750)

Autosport.com have some interesting Anorak Stats in their 'Tales from Sepang' article...in their 'Super Stats In-depth Malaysian GP analysis from Malaysia' they have various tables that are interesting.
The one that interested me (sorry it's off thread) was that 'my driver' (Hehe!) did the least distance of all 28 drivers over the weekend, not that it held him back that much. Snigger! (Sorry)

Dutton · 23 Mar 2006, 18:42 (Ref:1558764)

How do you quantify the "Chance"? Also, you say the "Chance" is the probability the result were obtained by fluke: I was of the understanding chance and probability are different things.

Knowlesy · 23 Mar 2006, 19:02 (Ref:1558780)

Oh don't start this again Dutton!

Dani Filth · 23 Mar 2006, 20:53 (Ref:1558975)

i must say .. i just love Schummy's Stats

**Bononi** · 23 Mar 2006, 23:17 (Ref:1559093)

Quote:

Originally Posted by Dutton

How do you quantify the "Chance"? Also, you say the "Chance" is the probability the result were obtained by fluke: I was of the understanding chance and probability are different things.

very reasonable...

Dixie Flatline · 23 Mar 2006, 23:30 (Ref:1559105)

There's an interesting article on grandprix.com by "the Mole" discussing Webber's performance against Rosberg's performance in Bahrain. The article made the point that Webber outpaced Rosberg throughout the weekend, but because Rosberg sliced his way through the field after having his nose knocked off (and a fastest lap to boot), Rosberg gets the plaudits.

Inigo Montoya · 23 Mar 2006, 23:56 (Ref:1559117)

These stats are wicked Schummy. Keep them coming

Schummy · 24 Mar 2006, 01:15 (Ref:1559155)

Quote:

Originally Posted by Dutton

How do you quantify the "Chance"? Also, you say the "Chance" is the probability the result were obtained by fluke: I was of the understanding chance and probability are different things.

Well "chance" is a multi-purpose word but I'm using it here in the meaning of likeliness of appearing an event (i.e. probability).

Anyway, I'm not going to analyze the subtleness of terms as risk, randomness, probability, etc

It is too removed for the main purpose of these little stats

Let's roll even more "senseless" F1 stats!

Nicholosophy · 24 Mar 2006, 08:11 (Ref:1559406)

Could I suggest something in regards to qualifying?

Should we look at grid position or fastest lap achieved in all sessions to indicate the fastest of the two?

Schummy · 24 Mar 2006, 08:48 (Ref:1559427)

I think I prefer grid position. The other sessions are run in possibly different fuel/tyres combos and drivers don't necessarily make flat out laps.

More or less the same argument can be made for segments within qualifying session. Only the last and definitive qualifying segments have (most of) drivers really fighting for those precious grid positions.

Everything has a bad side, fuel strategies make less reliable grid positions as well. Strategies and traffic can affect fastest race laps or race results. Kimi Raikkonen was victim in MAL of a 1st lap unprovoked accident that damaged his race FL and result. Anyway, that's one reason to take in count the existence "chance" elements in order to make reasonable conclusions.

Schummy · 2 Apr 2006, 14:56 (Ref:1567961)

Comparison after Australia

Code:

Drivers BAH / MAL / AUS   TOT Chance AlSi Sign
FA-GF   3-0 / 1-2 / 2-1 = 6-3  51%   ---  ---
KR-JM   1-2 / 0-3 / 3-0 = 4-5 100%   ---  ---
MS-FM   3-0 / 2-1 / 3-0 = 8-1   4%   2-1  3-0
JB-RB   3-0 / 3-0 / 1-2 = 7-2  18%   3-0  ---
MW-NR   2-1 / 2-1 / 3-0 = 7-2  18%   3-0  ---

Alonso is doing better than Fisi, but just now it can be due to chance.
Kimi has equalled Montoya, after a bad start of the season.
Massa is already almost significantly doing worse that MS.
Button and Webber are forcing the hands of Barrichello and Rosberg, but not significantly yet. Anyway Jenson has lost some ground in Melbourne.
A strong result for MS against FM in the next GP, would produce the first "victim": Massa would be significantly beaten by Schumi.

Schummy · 26 Apr 2006, 18:47 (Ref:1595622)

An update of scores between team mates (one point for (each) better grid position, better fastest lap and better final classification) after Imola. We have now:

Code:


Teams  Score Chance A.S. Sig.
FA-GF   9-3  15%    3-0
MS-FM  11-1   0.6%  1-2  2-1   
KR-JM   5-7  77%    
JB-RB  10-2   3.8%  2-1  3-0
MW-NR  10-2   3.8%  2-1  3-0
RS-JT  10-2   3.8%  2-1  3-0

In Imola, Alonso, MS, Button, Webber and Ralf have beaten their mates by 3-0, however Kimi has lost again to Montoya by 1-2 (I put KR before JPM because he is perceived loosely as sort of num. 1 driver).

With 4 GPs we already have a "victim": Felipe Massa is (statistically) significantly beaten by Schuey. So I will drop this pair in further calculations, as it has been proved than MS has done better than FM.

Three pair are in the edge to reach a significant conclusion as well: Button, Webber and Ralf are "almost significantly" better than their team mates. If they get 3-0 in Nurburgring they will become "proclaimed" better than their mates.

Alonso will need something as two 3-0 GPs against Fisi to "prove" he is better statistically speaking. Somehow it is expected Fernando will achieve it sooner or later.

The only really close fight is in McLaren team. Just now Juan Pablo leads by 7-5, but this is far for being significant. Anyway, he is nearer to achieve it than KR. If they go on this way it is possible to go along the season without any significant decision between those two.

If more pairs are significantly "resolved", I can include Villeneuve-Heidfeld in the table. I am afraid that even deciding who I put first or second is going to irritate someone!

shiny side up! · 26 Apr 2006, 22:56 (Ref:1595807)

Do it Schummy, do it! Who has the lower car number between JV and NH? That would be a fair way to determine the order...

rocketracer · 27 Apr 2006, 00:40 (Ref:1595848)

yeah go on schummy, do it! we would all love to see it - bmw is panning out to be the hotly contested battle of 06 (as dixie noted in another thread - nothing like fighting for your career!), but would it be too much to ask to do DC vs klien as well? theres certainly plenty of contentious talk over their battle...

oh, and i've got 3 words for you - alph-a-betically!

PS loving your work...

Nicholosophy · 27 Apr 2006, 01:59 (Ref:1595874)

Good job on the stats. It's interesting to see how drivers are going on a statistical basis and not just a personal opinion basis.

Schummy · 27 Apr 2006, 15:43 (Ref:1596322)

OK, here goes the scores for BMW and Red Bull:

I have put JV first because of his WDC title

Code:


Team  BAH  MAL  AUS  SMA  Total
JV-NH 010  101  000  111 = 6-6
DC-CK 000  000  111  101 = 5-7

As you can see both teams have a great equality between team mates. But in both cases the more experienced driver started rather low (3-6 Jacques and 0-6 Coulthard).

So it will be interesting to look at how it goes on in the future.

**Bononi** · 29 Apr 2006, 12:45 (Ref:1597494)

Very interesting. Good work Schummy

Schummy · 7 May 2006, 23:22 (Ref:1602885)

After GP Europe in Nurburgring:

Code:


Drivers BAH / MAL / AUS / SMA / EUR    TOT Prob  AlSi Sign
FA-GF   3-0 / 1-2 / 2-1 / 3-0 / 3-0 = 12-3* 96.5% 2-1  3-0
KR-JM   1-2 / 0-3 / 3-0 / 1-2 / 3-0 =  8-7   0%   ---  ---
JB-RB   3-0 / 3-0 / 1-2 / 3-0 / 0-3 = 10-5  70%   ---  ---
MW-NR   2-1 / 2-1 / 3-0 / 3-0 / 1-2 = 11-4  88%   3-0  ---
RS-JT   2-1 / 2-1 / 3-0 / 3-0 / 1-2 = 11-4  88%   3-0  ---
JV-NH   1-2 / 2-1 / 0-3 / 3-0 / 2-1 =  8-7   0%   ---  ---
DC-CK   0-3 / 0-3 / 3-0 / 2-1 / 1-2 =  6-9  39%   ---  ---

MS has already "beaten" Massa, so that pair will not appear again.

Alonso has almost beaten Fisi (99% confidence is what we call here "beaten", 95% is "almost beaten"). If he beats GF 3-0 in GP Spain, he will have "finished" Fisi.

Rubens, Jarno and Nico have saved their "compromised" situation winning against respective team mates in Nurburg. But Webber and Ralf can "almost beat" them if they achieve a 3-0 in Montmelo.

McLaren, BMW and RBR pairs are equilibrated, although Klein has some advantage over David "this-is-my-year" Coulthard. The way they are going is difficult one of those three teams can beat significantly his t-mate.

I have changed the "Prob" column in the table. Now shows something we can call "probabilist level of confidence" of one being better than the other driver. When 99% is reached we will pronounce the victim "significantly beaten". If 95% is achieved we will label it as "almost beaten".

Schummy · 8 Jul 2006, 22:17 (Ref:1651046)

Updating after GP USA.

Code:

Drivers BAH / MAL / AUS / SMA / EUR / SPA / MON / GBR / CAN / USA TOT Prob FA-GF 3-0 / 1-2 / 2-1 / 3-0 / 3-0 / 3-0 / 3-0 / 3-0 / 3-0 / 0-3 = 24- 6 0.1% beaten! KR-JM 1-2 / 0-3 / 3-0 / 1-2 / 3-0 / 3-0 / 2-1 / 3-0 / 3-0 / 2-1 = 21- 9 4.3% almost beaten! MS-FM 3-0 / 2-1 / 3-0 / 3-0 / 3-0 / 2-1 / 3-0 / 3-0 / 3-0 / 3-0 = 28- 2 0.0% beaten! JB-RB 3-0 / 3-0 / 1-2 / 3-0 / 0-3 / 2-1 / 1-2 / 0-3 / 3-0 / 0-3 = 16-14 86% equality MW-NR 2-1 / 2-1 / 3-0 / 3-0 / 1-2 / 2-1 / 2-1 / 0-3 / 2-1 / 1-2 = 18-12 36% ? RS-JT 2-1 / 2-1 / 3-0 / 3-0 / 1-2 / 1-2 / 1-2 / 1-2 / 0-3 / 2-1 = 16-14 86% equality JV-NH 1-2 / 2-1 / 0-3 / 3-0 / 2-1 / 0-3 / 1-2 / 0-3 / 2-1 / 3-0 = 14-16 86% equality DC-CK 0-3 / 0-3 / 3-0 / 2-1 / 1-2 / 1-2 / 3-0 / 2-1 / 1-2 / 2-1 = 15-15 100% equality

Some circustances needed consideration: in Monaco MS is given a victory over FM in qualif despite his disqualif of times, because here we want measure perfomances, not behaviour. In USA, KR and JPM are considered tied in fastest lap and race results.

As can be seen, many pairs are very close. Probably some cases will be a surprise for some people. Only FA and MS (and perhaps KR) have definitively beaten their mates. Of the rest the only one who can achieve this in a short term is MW against NR. He will need two 3-0 in the next two races.

Apparently the table is too big for a comfortable view. I'm sorry but I'm not responsible for that

Schummy · 4 Aug 2006, 03:55 (Ref:1672759)

Update after GP Germany.

Code:

Drivers BAH / MAL / AUS / SMA / EUR / SPA / MON / GBR / CAN / USA / FRA / GER    TOT   Likelihood  
FA-GF   3-0 / 1-2 / 2-1 / 3-0 / 3-0 / 3-0 / 3-0 / 3-0 / 3-0 / 0-3 / 3-0 / 1-2 = 28- 8   0.1% beaten!
KR-PR                                                             / 2-1 / 3-0 =  5- 1  22%   ?
MS-FM   3-0 / 2-1 / 3-0 / 3-0 / 3-0 / 2-1 / 3-0 / 3-0 / 3-0 / 3-0 / 3-0 / 3-0 = 34- 2   0.0% beaten!
JB-RB   3-0 / 3-0 / 1-2 / 3-0 / 0-3 / 2-1 / 1-2 / 0-3 / 3-0 / 0-3 / 2-1 / 3-0 = 21-15  41%   ?
MW-NR   2-1 / 2-1 / 3-0 / 3-0 / 1-2 / 2-1 / 2-1 / 0-3 / 2-1 / 1-2 / 1-2 / 3-0 = 22-14  24%   ?
RS-JT   2-1 / 2-1 / 3-0 / 3-0 / 1-2 / 1-2 / 1-2 / 1-2 / 0-3 / 2-1 / 2-1 / 2-1 = 20-16  62%   equality
JV-NH   1-2 / 2-1 / 0-3 / 3-0 / 2-1 / 0-3 / 1-2 / 0-3 / 2-1 / 3-0 / 1-2 / 3-0 = 18-18 100%   equality
DC-CK   0-3 / 0-3 / 3-0 / 2-1 / 1-2 / 1-2 / 3-0 / 2-1 / 1-2 / 2-1 / 2-1 / 1-2 = 18-18 100%   equality

Equality in Toyota, BMV and Red Bull.
Overwhelming advantage of MS, big, decisive advantage for FA.
Button has some advantage over Barrichello, but not much.
Webber can beat significantly Rosberg but he needs more than this.
Kimi can beat the likeable Pedro in short time, I think.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Bahrain: some fastest laps little stat	Schummy	Formula One	28	18 Mar 2006 17:11
James Kaye - Stat required	Andy H	Touring Car Racing	6	7 Mar 2003 18:12
Paging Dr. Austin, Paging Dr. Austin....enemas needed STAT	stretch51	IRL Indycar Series	1	5 Apr 2002 00:41