13 replies to this topic

[bernardv]

If you look at average lap times for each of the 4 stints, Vettel was always better than Alonso. But if you look at the race as a whole it's the other way around!

How can that be?

(elementary school math)

[fdspd]

Standard deviation. In addition to the natural field spread at the end of first lap.

[Ferrari_F1_fan_2001]

Starting further back; traffic; field spread

[Zava]

Starting further back; traffic; field spread

those are all included in the lap times.

did you count a weighted average or just an average? because Vettel was on an inverse tactic, his pit stops were slipped by ~10 laps compared to Alonso, obviously 1, he was on lighter fuel for the stints, 2, had a 10 lap longer "slowest stint" and a 10 lap shorter "fastest stint" compared to Alonso.
so it is possible that the averages (just looking at the times and saying a rough average) looked like this:
Vettel: 1:44.5 - 1:43.3 - 1:41.1 - 1:38.0
Alonso: 1:45.0 - 1:43.5 - 1:41.6 - 1:40.3
but it doesn't work like averaging those, it works like this:
Vettel: 14*(1:44.5)+16*(1:43.3) etc...
and with that, I'm pretty sure you'll get a not that surprising result.

[artista]

If you look at average lap times for each of the 4 stints, Vettel was always better than Alonso. But if you look at the race as a whole it's the other way around!

How can that be?

(elementary school math)

As a good friend of mine (Uni Professor in statistics) always says, if I have eaten a whole chicken and you're starving, statistics is the science that says I have eaten 1/2 a chicken and you another half.

That doesn't mean statistics gives wrong results, it means you must know what you can/have to use in each situation and what's wrong for each situation.

What elementary school maths say, and it's absolutely right, is that Alonso's average lap time in the race (discarding lap one, which has a different length for each driver) is 1:43.2899 and Vettel average lap time in the race was 1:43.4314, or in other words: it took Alonso less time to finish the race and, therefore, he won.

What you can never do is to compare average times for stints which happened at different moments in time and with different number of laps.

[paulogman]

plus the simple fact that alonso crossed the finish line first and vettel wasn't even on the podium

[joshb]

As a good friend of mine (Uni Professor in statistics) always says, if I have eaten a whole chicken and you're starving, statistics is the science that says I have eaten 1/2 a chicken and you another half.

That doesn't mean statistics gives wrong results, it means you must know what you can/have to use in each situation and what's wrong for each situation.

What elementary school maths say, and it's absolutely right, is that Alonso's average lap time in the race (discarding lap one, which has a different length for each driver) is 1:43.2899 and Vettel average lap time in the race was 1:43.4314, or in other words: it took Alonso less time to finish the race and, therefore, he won.

What you can never do is to compare average times for stints which happened at different moments in time and with different number of laps.

To quote- "It was simple, he was (also) racing, he was faster, he passed him (twice) he won"

[bernardv]

Well, my answer would be that the stints overlap only partially (about 50%), because of the different tactics. E.g. most of Vettel's 3rd stint overlaps with Fernando's 4th (final) stint. If we compare those two, we see why Fernando came out on top. The only proper way being to compare the race as a whole of course.

[Diderlo]

those are all included in the lap times.

did you count a weighted average or just an average? because Vettel was on an inverse tactic, his pit stops were slipped by ~10 laps compared to Alonso, obviously 1, he was on lighter fuel for the stints, 2, had a 10 lap longer "slowest stint" and a 10 lap shorter "fastest stint" compared to Alonso.
so it is possible that the averages (just looking at the times and saying a rough average) looked like this:
Vettel: 1:44.5 - 1:43.3 - 1:41.1 - 1:38.0
Alonso: 1:45.0 - 1:43.5 - 1:41.6 - 1:40.3
but it doesn't work like averaging those, it works like this:
Vettel: 14*(1:44.5)+16*(1:43.3) etc...
and with that, I'm pretty sure you'll get a not that surprising result.

First thing that popped in my mind was this. Sure Vettel was faster, because he started his comparable stints later or had different tyres. So Vettel had less fuel.

[artista]

First thing that popped in my mind was this. Sure Vettel was faster, because he started his comparable stints later or had different tyres. So Vettel had less fuel.

And in the first stint, the first lap -in which Alonso is 5 seconds faster- is taken away in the calculations. That causes the average lap time to be artificially lower for Vettel than for Alonso for the only stint they started at the same time (and more or less same fuel), and you end up with all four stints in which Vettel was "faster" than the race winner, even if he finished fourth.
Ain't maths lovely?

[Seanspeed]

Well, my answer would be that the stints overlap only partially (about 50%), because of the different tactics. E.g. most of Vettel's 3rd stint overlaps with Fernando's 4th (final) stint. If we compare those two, we see why Fernando came out on top. The only proper way being to compare the race as a whole of course.

You've got it.

[Zava]

it seems Alonso made the same mistake
http://www.f1fanatic...kkonen-driving/

“In China, on race pace, Sebastian was quickest in all the stints, but had chosen a strange strategy for qualifying.”

[Kingshark]

Because Fernando is a magician.

[bl-f1]

The wheel is spinning, but the hamster is not there...