GP AUS: Statistics about lap times

----------------------------
WARNING! MANY NUMBERS AHEAD!
----------------------------

I have done some calculations about lap times in Melbourne.

Firstly, I've calculated median lap times per driver. Median is convenient because it is barely impacted by in and out laps or other anomalous laps (for example, first lap). So median gives a good aprox of the typical lap time for a driver in the race. In the following table, I have pointed out with "*" the fastest driver of each team. "DIF" means lap time gap to the fastest driver.

DRV MEDLAP CAR DIF
----------------------
RAI 87.338 *FER 0.0 *
HAM 87.790 *MCL 0.5 ·····*
ALO 87.791 MCL 0.5 ·····*
HEI 88.048 *BMW 0.7 ·······*
KUB 88.133 BMW 0.8 ········*
MAS 88.442 FER 1.1 ···········*
FIS 88.524 *REN 1.2 ············*
ROS 88.935 *WIL 1.6 ················*
KOV 89.245 REN 1.9 ···················*
BAR 89.275 *HON 1.9 ···················*
RSC 89.366 *TOY 2.0 ····················*
TRU 89.439 TOY 2.1 ·····················*
SAT 89.477 *SAG 2.1 ·····················*
COU 89.728 *RDB 2.4 ························*
WEB 89.868 RDB 2.5 ·························*
BUT 89.871 HON 2.5 ·························*
WUR 89.892 WIL 2.6 ··························*
LIU 90.034 *TOR 2.7 ···························*
DAV 90.430 SAG 3.1 ·······························*
SPE 90.511 TOR 3.2 ································*
SUT 90.653 *SPY 3.3 ·································*
ALB 92.316 SPY 5.0 ··················································*

Only Albers's result is unreliable as he did just 10 laps. Speed a bit shortchanged too because he did a bit less than 50% of laps, but his number of laps is relatively reliable.

Kovalainen 0.7 secs slower than Fisi
Rosberg 1.0 secs fastest than Wurz
Barry Cello ;) 0.6 secs faster than "Disappointing" Button.
Great weekend for Sato. In the race he was 1.0 sec faster than Davidson (but AD was possibly handicapped by the accident with Sutil).
Speed 0.5 secs behind Liuzzi.
Sutil was unstable but did very reasonable laptimes.
Extremely close lap times between team mates in McLaren, BMW, Toyota y Red Bull. The most surprising case is Alonso-Hamilton, but when FA was clear of LH, he seemed to run somewhat faster. We will see how it develops...

In the next table, "LAP" is the average between the two drivers in each team, "DIF2" is the gap to the fastest team. "DIF1" is the gap to the fastest team using only the fastest driver in each team (possibly a better alternative than DIF1). "RANK" is the ordering according to DIF1, that's basically the current order of teams in terms of peak perfomance.

CAR LAP DIF2 DIF1 RANK
--------------------------
MCL 87.8 0.0 0.5 2 ·····*
FER 87.9 0.1 0.0 1 *
BMW 88.1 0.3 0.7 3 ·······*
REN 88.9 1.1 1.2 4 ············*
WIL 89.4 1.6 1.6 5 ················*
TOY 89.4 1.6 2.0 7 ····················*
HON 89.6 1.8 1.9 6 ···················*
RDB 89.8 2.0 2.4 9 ························*
SAG 90.0 2.2 2.1 8 ······················*
TOR 90.3 2.5 2.7 10 ···························*
SPY 91.0 3.2 3.3 11 ·································*


Top group: Ferrari, McLaren and BMW, with a sizeable edge for Ferrari of half a second (about half a minute in a whole race).
Renault is hanging between the top and the middle zone. My subjective personal opinion is Renault maybe will be in trouble if Kova doesn't turn to be another Hamilton or Kubica, because this team lacks currently of a world class top F1 driver who can lead and motivate the team, from the mechanic to the engineers at the factory. F1 history shows many cases of top teams decaying when top drivers left the (mother)ship.
A long string of teams in the middle zone in this aprox order: Williams, Honda, Toyota, Super Aguri (this name...), Red Bull, Toro Rosso. There is not any big gap between them but the absolute difference between Williams and Toro is big (> 1 sec).
Last position for Spyker, but they are not totally detached from the former group

Finally, and seeing that I had yet hours ahead to burn in this masochistic quest, I have calculated the number of "Anomalous Laps" per driver. ALs are laps much slower than rest of laps for that particular driver. In the following table "AL2" and "AL3" are the important columns.

Usually ALs by default are: first lap, in laps and out laps. "AL1" is the raw number of ALs, "PITS" are the actual number of pitstops during the race (if there is any error, please say me). "AL2" is the number of ALs except the default ones (1+2*n, where n is the number of pitstops). And "AL3" is AL2 corrected taking in account the number of laps done by the driver (basically a little correction on AL3).

DRV AL1 PITS AL2 AL3
---------------------
HAM 4 2 -1 -1
COU 4 2 -1 -1.2
ALO 5 2 0 0
TRU 5 2 0 0
SAT 5 2 0 0
KUB 3 1 0 0
ALB() 1 0 0 0
RAI 6 2 1 1
HEI 6 2 1 1
FIS 6 2 1 1
MAS 4 1 1 1
ROS 6 2 1 1
RSC 6 2 1 1
WEB 6 2 1 1
LIU 6 2 1 1
DAV 6 2 1 1
SPE() 4 1 1 2.1
BAR 7 2 2 2
WUR 5 1 2 2.4
KOV 8 2 3 3.1
BUT 8 2 3 3.1
SUT 8 2 3 3.1


Hamilton and Coulthard were extremely consistent. Brilliant work by Lewis in his first GP in an unkown track.
Alonso, Trulli(!), Sato(!!) and Kubica were very consistent. Alonso's case is for granted, but Kubica's is impressive. Very noteworthy are the respective progress in race terms for Trulli and Sato.
Barry and Wurtz were a bit distracted, with two unmotivated (AFAIK) anomalous laps each.
Kovalainen, Button and Sutil were disastrous in terms of reliable "lapping". If Earth is this unstable we are in big trouble :p . Sutil, as formerly mentioned, however did good lap times.

And... that's all, folks!

Makes Hamilton's debut look that little bit more impressive IMO.

I think this is enough reason to warrent a "Button: OVER THE HILL??" Thread or two, or maybe a "What now for Webber?" Thread.

Good stuff.

I'd say we should wait for a clean weekend from the two, but that'd be misleading as to what went wrong.

Great stuff as usual Schummy. :)

Err,yes,great stuff Schummy. *now my head hurts*

The one thing that we can deduce from all that,is that some drivers need to pull their weight a bit more,whereas others are doing a grand job.:)

And don't we already have "Button" and "Webber" threads on a daily routine basis.

The median laps have a lot that can influence them. A big one is getting stuck behind someone who is slow. For instance Rubins was stuck behind Button for quiet some time. His laps behind Button are going to be adversely effected and his median lap will likely be worse. Overall I'd say that perhaps the Honda in RB's hands was an okay car. Massa had the fastest car (same as Kimis), but he was so effected by traffic that his fastest lap isn't that good. Rosberg was held up at times whereas Wurz held others up so that teammate comparison gap was likely larger in reality (Wurz's car had issues though).

Alonso was "managing the gap" behind Hamilton for most of the race, therefore his laps are mostly dictated by Hamilton's pace. Kimi, Heidfeld and Kubica and Fisi basically ran their own races, not particularly impacted by others.

A couple of interesting aspects I take from the numbers that I didn't realize. Webber was slow. Sato was on fire and that Aguri was pretty close to the Toyotas, but the new Honda is a bit faster than both.

Median and "averages" in general can be afected also by strategy. FM was using a combination of fuel/tyres strategy that (supoosedly) helped him through the traffic, but didn't help his lap times. As you say, the jury is out about the ultimate pace of FA vs LH (probably 0.1 or 0.2 secs according to qualifying and last laps).

But fastest laps have problems too. Peak perfomance is different from consistent pace. Alonso and Schumacher (Michael, that is) were able to put many consistently fast laps through a race, meanwhile other fast drivers were not able to do. Fastest laps are also influence by each particular car behavior with empty tanks or new tyres. So we have two interesting statistically here: peak lap times and consistent lap time; but they point at slightly different things.

An interesting possibility is using a "biased" average, as for example the 25% quantile (quartile 1), that extract the "consistent fast lap" (to say that). Or comparing cummulative frequencies (very informative). But, frankly, I didn't want to complicate things further in those tables :) There will be others threads to explore more issues.

Top stuff Schummy. :beer:

Some interesting stats there.

:wine:

(Why knowlesy and I are always drinking together, no matter what thread it is? :D )

I'l just echo Knowlesy and say "top stuff." Awesome effort Schummy.

Have to point out though that most of Webber's lap times were set with an open fuel flap and supposedly dodgy electronics, so it's anyone's guess as to how much that affected the times.

And you're right about Massa, it may have been a brilliant performance to go half the race on softs like he (apparently) did... then again, there weren't too many others doing the same thing in the same car so we don't know exactly how well he did, or how poorly.

That was truely awesome. Great set of stats there :)

One thing I wonder though; where did you get all the information for all of the lap times?

Hamilton and Coulthard were extremely consistent. Brilliant work by Lewis in his first GP in an unkown track.

Hamilton very consistent despite moments like this (http://www.autosport.com/gallery/photo.php/id/68138) during the race!