Catania - Inter (3 Mar 13)

victor_inter · Mar 4, 2013

i cant wait to see those stats. I am quite interested to know lol if there is actually a correlation between winning games and possession. Based on the league standings in serie a specifically i would have to say yes. But it is totally different in UCL imo. Motivation factors, determination, financial incentives, increased wages all play a part in UCL because this is a competition where only elite teams can participate, besides it is probably the most prestigious competition in Europe.

.h. · Mar 4, 2013

I4E said:
Liverpool have won ~45% of their games. Man Utd have won ~90%.

They have the same possession percentage and pass success.

How can their be a correlation for one but not the other ???

You're misunderstanding correlation. Its not done on that locally a basis - its global. For any one team, its not a predicator, but it can imply correlation - e.g. a team with more possession is more likely to do better, but its no guarantee (unless you have 1:1 correlation of course)

Bzzlightyear · Mar 4, 2013

browha said:
You're misunderstanding correlation. Its not done on that locally a basis - its global. For any one team, its not a predicator, but it can imply correlation - e.g. a team with more possession is more likely to do better, but its no guarantee (unless you have 1:1 correlation of course)

Browha 1:1 correlation is a bit hard to get specialy to analyse a data with " winning a game " by position. I will say as well if we take one game only one game data were one team have most position and wins the game, then yes possible to get 1:1 correlation, but when you take multiple matches " data " then it becomes more and more complicated

victor_inter · Mar 4, 2013

oh damn it browha, i guess u r not posting those stats after all lol. Anyways looking forward to reading your next article under writer's section.

.h. · Mar 4, 2013

bzzdk said:
Browha 1:1 correlation is a bit hard to get specialy to analyse a data with " winning a game " by position. I will say as well if we take one game only one game data were one team have most position and wins the game, then yes possible to get 1:1 correlation, but when you take multiple matches " data " then it becomes more and more complicated

I'm certainly not expecting 1:1 correlation out of this. Not at all. I'm just stating that IF we did have 1:1 correlation, it means for every 1% more possession you have, you have some very linear 1 more point (or something)

I'm also going to plot it as a function of points in the league. That seems the most reasonable discriminator.

- - - Updated - - -

victor_inter said:
oh damn it browha, i guess u r not posting those stats after all lol. Anyways looking forward to reading your next article under writer's section.

I will post the stats - I'm just busy with work. Give me a few hours after I get home today, we'll see how it goes

Bzzlightyear · Mar 4, 2013

yes you get linear funktion f(x)= ax+b

victor_inter · Mar 4, 2013

nha its all good, take your time. MAybe just post it as a new article ? Could be interesting to know what people think of those stats.

Michael · Mar 4, 2013

dynasty27 said:
Michael, do you also translate tuttosport?

Not really for two reasons: one I do not have a subscription for it, two I hate the paper. But when there's a big story like the interview with MM few days back, I'll find a way to do it.

Pharaoh · Mar 4, 2013

Michael said:
Not really for two reasons: one I do not have a substitution for it, two I hate the paper. But when there's a big story like the interview with MM few days back, I'll find a way to do it.

Do you mean subscription? Sorry that confused the fuck outta me.

.h. · Mar 4, 2013

I have all the data inputted for the correlation matricies... just got to automate the actual calculation so it creates the correlation coefficients for each league and combined

Bzzlightyear · Mar 4, 2013

browha said:
I have all the data inputted for the correlation matricies... just got to automate the actual calculation so it creates the correlation coefficients for each league and combined

Could you send the data, and as well did you use plot as "x= position and y = winning" the game ?

.h. · Mar 4, 2013

I'll make it all publicly available once I'm finished.

The preliminary result is that the weakest correlation is ~0.58 (La Liga), 0.82 (Bundesliga), 0.72 (Serie A), and 0.67 (Premiership). This means ALL the leagues are strongly correlated, possession with points gained. It is pretty damn good correlation in the Bundesliga, and weakest in La Liga (but 0.58 correlation is still bloody strong)

Michael · Mar 4, 2013

Pharaoh said:
Do you mean subscription? Sorry that confused the fuck outta me.

LOL substitution... :chan: Yes subscription.

.h. · Mar 4, 2013

Posted. See my sig for the article.

Wallace · Mar 4, 2013

Pajo said:
Lol, some people just don't know to give credits where's due. Bipolar fucks.

I don't get it, I think you're misusing the word bipolar.

In fact, I think rockball is very stubborn in repressing the number of credits he gives out, just like he does usually.

Therefore, I think he is a polar fuck, not a bipolar fuck, if that makes any sense.

shingayi · Mar 4, 2013

Shaun · Mar 4, 2013

I'm not sure why this is even an argument. Having more possession is generally a good thing because the team with the more possession more often than not wins. Of course there are exceptions, but Browha never claimed there wasn't.

Seriously, why the fuck is this an argument? It's like arguing whether good looking people get laid more often than ugly ones. Of course, usually, but there are exceptions

Bzzlightyear · Mar 4, 2013

Well but there is always a solution but calculating the outcome, and facts does approve. After all we are statistics, and statistics can be calculated, if the data is False or truth.

Pimpin · Mar 4, 2013

Shaun said:
I'm not sure why this is even an argument. Having more possession is generally a good thing because the team with the more possession more often than not wins. Of course there are exceptions, but Browha never claimed there wasn't.

Seriously, why the fuck is this an argument? It's like arguing whether good looking people get laid more often than ugly ones. Of course, usually, but there are exceptions

yeah and han, if a dude has a bigger penis and cant get erected there is a pill for that.. just sayin

.h. · Mar 4, 2013

I stop arguing when Han gets involved, tbh. As a rule of survival

Catania - Inter (3 Mar 13)

Capitano

Part time Lazarus

Capitano

Capitano

Part time Lazarus

Capitano

Capitano

La Grande Inter

Allenatore

Part time Lazarus

Capitano

Part time Lazarus

La Grande Inter

Part time Lazarus

Marotta FC

Capitano

La Grande Inter

Capitano

I'm better than Icardi

Part time Lazarus