Throwing it away

[Éperons]

Ep, please update this thread with more jiggery pokery. Since I finished reading Soccernomics you're becoming my new stat dealer.

I wonder what the correlation is between league performance and clean sheets, but I'd need the last five seasons of PL clean sheet data. It's not on MyFootbalFacts , afaict, and I've not found it elsewhere. I suppose I could programmatically do it with the match matrices, but fuck me that sounds like work.

 

[YidoBuckler]

Regardless we should be keeping more clean sheets than we have.

That I agree with. Settled back five next season and I think we will improve in this area. Still finished with a GD of 20, bet we haven't beaten that many times in Premier League.

 

[Pairsy1882]

That I agree with. Settled back five next season and I think we will improve in this area. Still finished with a GD of 20, bet we haven't beaten that many times in Premier League.

Totally agree, looking forward to seeing us at full strength next season. How much of a difference can Kaboul make? I still want a new left back but the choice of centre backs next season is going to be the envy of most Premier League side; Kaboul, Vertonghen, Caulker and Dawson, take your pick

I agree that we may not have beaten a GD of +20 too many times before but I would argue that you only have to look at our GD difference compared to the teams above us to see that keeping a few more clean sheets could make a difference to our final standing.

 

[Éperons]

I wonder what the correlation is between league performance and clean sheets, but I'd need the last five seasons of PL clean sheet data. It's not on MyFootbalFacts , afaict, and I've not found it elsewhere. I suppose I could programmatically do it with the match matrices, but fuck me that sounds like work.

Guess it wasn't that hard...

So this is the regression, over the past five PL seasons, of number of clean sheets (y-axis) against place in the PL (x-axis):

The correlation coefficient is -.8

Stone cold correlation. But who knows about causation :balelol:

 

[Pairsy1882]

Guess it wasn't that hard...

So this is the regression, over the past five PL seasons, of number of clean sheets (y-axis) against place in the PL (x-axis):

The correlation coefficient is -.8

Stone cold correlation. But who knows about causation :balelol:

Is there an Eperons to English translator on Google?

 

[Éperons]

Is there an Eperons to English translator on Google?

Over the past five seasons, it's been the case that the higher you finished up the table, the more clean sheets you had. Sounds obvious, but I wanted to test it.

 

[Pairsy1882]

Over the past five seasons, it's been the case that the higher you finished up the table, the more clean sheets you had. Sounds obvious, but I wanted to test it.

Thank you very much for the translation. I thought that was probably the case as well. It would be interesting to research the causes behind the anomalies. e.g how we managed to finish 6 points off of 2nd place keeping only nine clean sheets.

 

[Pairsy1882]

Éperons please can you perform some of you ANALysis magic on this...

 

[Éperons]

Thank you very much for the translation. I thought that was probably the case as well. It would be interesting to research the causes behind the anomalies. e.g how we managed to finish 6 points off of 2nd place keeping only nine clean sheets.

Because correlation isn't causation. It's the same thing with having a huge wagebill. There's a definite connection between leading the league in wages and winning the championship, but that doesn't mean that a) a huge wage bill = trophy or b) poor teams can't win.

It just means that if you have big wages, it seems to be the case that you also have a bigger chance of winning the trophy. Similarly, if you have more clean sheets, it seems to be the case that you finish higher in the table.

So since Spurs are a motherfucking emo special snowflake, they've figured out a way to break the trend.

How special are we? This season we finished fifth overall and 12th in clean sheets. Last season, 4th/4th. Season before, 5th/12th. Season before, 4th/7th. Season before, 8th/7th.

Woo-hoo.

 

[YidoBuckler]

Pairsy1882 I can do some. If you don't concede, you don't lose.

 

[Pairsy1882]

Because correlation isn't causation. It's the same thing with having a huge wagebill. There's a definite connection between leading the league in wages and winning the championship, but that doesn't mean that a) a huge wage bill = trophy or b) poor teams can't win.

It just means that if you have big wages, it seems to be the case that you also have a bigger chance of winning the trophy. Similarly, if you have more clean sheets, it seems to be the case that you finish higher in the table.

So since Spurs are a motherfucking emo special snowflake, they've figured out a way to break the trend.

How special are we? This season we finished fifth overall and 12th in clean sheets. Last season, 4th/4th. Season before, 5th/12th. Season before, 4th/7th. Season before, 8th/7th.

Woo-hoo.

I am interested as to what it is that allows us to keep so few clean sheets and still achieve the points tally we have this season. Could it be something to do with the fact that we only failed to score in 4 games this season, only United failed to score in fewer games.

Interesting that some people have suggested that we are a more solid unit this season than we were under Redknapp, yet we finished higher up the clean sheet table under HR.

 

[Pairsy1882]

Pairsy1882 I can do some. If you don't concede, you don't lose.

Your powers of observation baffle me. Please enlighten me.

 

[Éperons]

I am interested as to what it is that allows us to keep so few clean sheets and still achieve the points tally we have this season. Could it be something to do with the fact that we only failed to score in 4 games this season, only United failed to score in fewer games.

I think the fact that we've been able to score is precisely what lessened the importance of the clean sheet. How many matches did we lose 1–0? Fulham and Wigan home… Furthermore, we had four 1–1 draws. So our failure to keep a clean sheet cost us a total of—under ideal circumstances—ten points. I don't know if that's a lot or not.

The Wanderers, by comparison, had four 1–1s like we did. But they only lost one match 1–0. So they dropped nine points this season by not keeping a clean sheet.

Chelsea had one 1–1 and no 1–0 defeats… pretty impressive. City had three 1–1 and lost once 1–0 (seven dropped points). United, on the other hand, had three 1–1s and three 1–0 defeats… so that's nine points dropped because they couldn't keep a clean sheet.

So though that measly one point keeps us out of CL (well, not really… we needed to finish with two more points than we got), we can see that dropping nine or ten points because of a lack of clean sheets isn't crazy. Woolwich had the third most clean sheets and still lost nine points!

I guess I'm not terribly worried, then. But it is something we should try to do better.

 

[tricky]

Is there an Eperons to English translator on Google?

whilst we either draw low scoring games or lose by small margins it is impossible to tell if it was keeping the clean sheet or our failure to score which has dictated the outcome. Hence you can prove correlation (league position to clean sheet ratio) without causation (understanding why it happens).

As an aside, given the nature of the game, there is the cliche (in part truth) of the opposition can always 'nick a goal' in low scoring games even if they are not dominating the game, the season is full of the for all teams. But as the converse is also true 'good teams not dominating and playing badly and still winning by small margins, it would be difficult to separate out the two, statistically speaking.