Overtaking in 2012: statistics and analysis

[bhall]

This is clearly not true. Statistics have validity, they do not 'mean whatever anyone wants them to mean'.
Just because they are not perfectly objective in many situations does NOT mean they can be 100% dismissed in all situations.

Statistics used in a meaningful way often lead to useful conclusions and insights. You just have to be careful how you use them.
Like the scientists at CERN with the Higgs Boson particle etc - statistical probability is very meaningful for them.

Correlation does not equal causation. The scientists at CERN would unequivocally agree with me on this. That's why they run so many tests and conduct so many experiments. Otherwise, everything after any first results would be a waste of time and energy.

I think the statistics submitted here can be dismissed, because they are woefully inadequate for anything other than extremely broad generalizations. And that's to say nothing of the fact that the raw data from which this "analysis" is based is admittedly incomplete and subjectively collated.

The final criteria involves subjective judgements and consequently figures can never be regarded as ‘definitive’. Gaps in the available data, such as moves missed by TV cameras or obscured on lap charts by pit stops or retirements, mean that the data do not lend themselves to detailed analysis at the micro level, but are indicative of general trends.

The title of this thread should be "Selected Overtaking in 2012: statistics and analysis." Instead, it pisses on me and tells me it's raining.

[mnmracer]

@mnmracer i´m sure you´ll find what you need at your site.

Please do not try to deceit people into thinking that statement is true without providing evidence.
Seriously dude, resorting to lying?

WTF are you talking about?
You have a site containing all the information you need, so i said "I´m sure you will find what you need at your site"
Your site = the site you took statistics from!

As I said, that statistic is not available on there. Do not make it look like I have the information but am not sharing it.
Those who people who care enough to invest time and research in it -be it CTA, F1F, Pirelli- do not consider it relevant information.

[Nando]

Not much of a statistics site if you can´t find out what happens on lap 1 of every single race on the calendar.
I´m positive you can find the information there.

But by all means, please continue with your flawed statistics.

Also, the numbers i gave you works just as fine as Bhall has so nicely proven. It´s cherry-picked just like your stats.

[Cylinder]

Stats can say anything, according to stats vettel overtook about 21 cars at Abu Dhabi, his fanboys tried to say this was proof he was a great racer, until you realise he overtook a bunch of caterhams and HRT and Marussias twice over, then the Torrso rossos just jumped out the way for him, then drivers infront like webber and massa crashed out infront of him or were pitted so he couldnt race them.

Stats can be manipulated to show all sorts, not really worth a big argument.

Hamilton is the best overtaker in the sport, followed by Alonso, that's my stat, the rest i dont care about.

Or you could just try and read...

Overtakes on the 5 top teams:
1. Sebastian Vettel - 17 on-track overtakes
2. Kimi Räikkönen - 16
3. Jenson Button - 14
4. Lewis Hamilton - 13
5. Felipe Massa - 12

You are clearly mixing up stat with personal opinion.

So let's get this straight, you are gonna sit here and tell me Jenson button is a better overtaker than Lewis Hamilton and that Massa is a better overtaker than Alonso.



Do me a favour, log off and go back to autosport forums and tell Bourbon, Sakae, Alacron and the rest of the crazy gang i said hello.

What you posted above is more proof if ever needed that numbers can be made and interpreted to prove lots of points, they are interpretations, they are not undeniable proof.

It's like saying Senna and Vettel are equal cos they both have 3 world titles, when in reality Senna would destroy webber week in week out and he wouldnt lose 9 qualis in a red bull to him either.

[mnmracer]

Not much of a statistics site if you can´t find out what happens on lap 1 of every single race on the calendar.
I´m positive you can find the information there.

But by all means, please continue with your flawed statistics.

Also, the numbers i gave you works just as fine as Bhall has so nicely proven. It´s cherry-picked just like your stats.

Let's start by explaining how 'your numbers' say who overtook who.
And how a negative number indicates the amount of overtakes.

Humor me.

[mbvinnie]

You are clearly biased against VET then - better? Or against mnmracer? You are arguing with emotion.
(I was going off your name, not your avatar for being an 'Nando' fan)

You are free to put me in whatever folder you like, seeing as you fire off without thinking it doesn´t really matter what i say.

I get it, starts or the first complete lap is not something to count but DRS overtakes is.
It´s flawed statistics that was only posted because of Vettel´s nr1 fan.

I don´t care who comes out on top to be honest, i just think you can´t pick and choose what to put in the statistics.

No, I don't think DRS overtakes are as valid as non-DRS overtakes. I agree with you. Let's look further then and count a DRS pass as a 1/2 weight.
I am all about refining the dataset and looking at it further, which is why I think the inclusion of launch passes would not help. I would want Lap 1 passes included if we could remove the Launch passes - I think that would help.

The whole thing about statistics is there should be some picking or choosing. That's the skill in analyzing. Just including everything often obscures the very trends and patterns you are looking for. You have to break it down.
And you can't have a pre-conceived outcome that you want to arrive at (that's like religion vs science).

Statistics done badly are awful and meaningless, statistics done well are very very powerful.
It wasn't the best starting point in this thread, but it was a start and had some potential to show something. Even if the answers are not definitive. It is just one strand of analysis.

[mnmracer]

So let's get this straight, you are gonna sit here and tell me Jenson button is a better overtaker than Lewis Hamilton and that Massa is a better overtaker than Alonso.

Do me a favor and quote me where I have uttered the words "best overtaker".

[Nando]

Let's start by explaining how 'your numbers' say who overtook who.
And how a negative number indicates the amount of overtakes.

Humor me.

Rather simple really.

Positions gained/lost on lap 1.
Negative number is average over a whole season on lap 1. the other Negative number is all overtakes and being overtaken on lap one compiled by plus or minus until you are finished counting.

If it´s incomplete or not satisfactory to you then go and watch every race and make a proper overtake-statistic where overtakes are not cherry picked.

[mnmracer]

The whole thing about statistics is there should be some picking or choosing. That's the skill in analyzing. Just including everything often obscures the very trends and patterns you are looking for. You have to break it down.
And you can't have a pre-conceived outcome that you want to arrive at (that's like religion vs science).

At least someone understands.
This is the data set that is available, and this is an analysis of that data.

[Nando]

"The whole thing about statistics is there should be some picking or choosing."

No that´s called tweaking statistics to get a different result then what actually was there had it been counted properly.

If a person makes an overtake before start/finish on lap 1 it won´t count, if he does the same pass just after start/finish it all of a sudden count.

Some nice stats there.

[mbvinnie]

I think the statistics submitted here can be dismissed, because they are woefully inadequate for anything other than extremely broad generalizations. And that's to say nothing of the fact that the raw data from which this "analysis" is based is admittedly incomplete and subjectively collated.

I agree that they are not definitive, and I have said several times that we can really only take some broad generalizations, but the analysis is kinda fun and interesting, and there is some limited usefulness. I agree that there is some bias, so let's look at how we can improve it rather than dismissing it.

I couldn't agree more about the 'correlation not equaling causation' thing. That's essentially my job is.
But that doesn't mean statistics can always be dismissed, or

[Cylinder]

So let's get this straight, you are gonna sit here and tell me Jenson button is a better overtaker than Lewis Hamilton and that Massa is a better overtaker than Alonso.

Do me a favor and quote me where I have uttered the words "best overtaker".

I said "Better overtaker". So you failed right out of the box yet again.

As i said before, stats can be pulled out of box to show all sorts, enjoy looking at the numbers, but dont try and delude yourself into thinking they make your man somekind of senna like overtaker.

Here are my stats. Vettel got out qualified 9 times by Mark Webber.

Alonso and Hamilton only got qualified 3 times.

Therefore Hamilton and Alonso are better qualifiers than Vettel.

How you like them stats?

[mnmracer]

Let's start by explaining how 'your numbers' say who overtook who.
And how a negative number indicates the amount of overtakes.

Humor me.

Rather simple really.

Positions gained/lost on lap 1.
Negative number is average over a whole season on lap 1. the other Negative number is all overtakes and being overtaken on lap one compiled by plus or minus until you are finished counting.

If it´s incomplete or not satisfactory to you then go and watch every race and make a proper overtake-statistic where overtakes are not cherry picked.

Now that's rich.
You -somebody with no interests in the subject whatsoever- are the one who is not satisfied with the available data.
Everyone relevant in this field considers this the only data relevant.

You make a claim, you back it up.

[mnmracer]

So let's get this straight, you are gonna sit here and tell me Jenson button is a better overtaker than Lewis Hamilton and that Massa is a better overtaker than Alonso.

Do me a favor and quote me where I have uttered the words "best overtaker".

I said "Better overtaker"

"you are gonna sit here and tell me Jenson button is a better overtaker"
"I said "Better overtaker"

So it's very simple, did I tell you Jenson Button was a better overtaker or not?

[mbvinnie]

"The whole thing about statistics is there should be some picking or choosing."

No that´s called tweaking statistics to get a different result then what actually was there had it been counted properly.

If a person makes an overtake before start/finish on lap 1 it won´t count, if he does the same pass just after start/finish it all of a sudden count.

Some nice stats there.

You're not listening. I WANT the Lap 1 overtakes (excluding Launch) included. I just don't think they are available.

You do have to select data carefully. Very rarely is an analysis run on 'everything' without looking into the data further. The key is to be careful, to be thorough and to not have a pre-determined outcome.