Football Statistics: Their uses in the modern game...

[AyeDubbleYoo]

Stats are for people with no knowledge, it's raw information without insight.

 

But that's what analysis is for.

[La Parka]

Brentford are/were? The biggest exponents of the moneyball method.

[gjohnson]

http://www.computerweekly.com/news/450296302/Foxy-Leicester-City-FC-won-Premiership-with-data-analytics

 

 

How do they pass off the other 18 teams that didn't win the league and athe least two that were relegated with Prozone?

[thomas]

http://www.computerweekly.com/news/450296302/Foxy-Leicester-City-FC-won-Premiership-with-data-analytics

 

 

How do they pass off the other 18 teams that didn't win the league and athe least two that were relegated with Prozone?

The article is a bit stupid and reads like a barely concealed advertisement - there's no silver bullet in football; not stats, not a manager, not money. There are unpredictable things outside of your club that no technology or methodology can control for.

 

However on the extreme other end, glibly dismissing the use of statistics altogether with the "they don't tell the full story" or "you can twist stats to make them say whatever you want" chestnuts is equally stupid.

 

It's only as useful as capabilities of the people who are making decisions based on it.

[LucaAltieri]

http://www.computerweekly.com/news/450296302/Foxy-Leicester-City-FC-won-Premiership-with-data-analytics

 

 

How do they pass off the other 18 teams that didn't win the league and athe least two that were relegated with Prozone?

 

Who knows? They do say that Leicester's work with data is "one of the most advanced in the Premier League." But who knows how they justify that statement.

 

I'm pretty skeptical of data for things like passing and shooting statistics, overall game play, etc. Because there are so many other factors that you can't quantify that have massive effects on the game. Strict adherence to the numbers in that context leads to things like signing Downing and Carroll for massive sums of money. But for things like athletic performance there does seem to be something there.

[LucaAltieri]

I listened to an interesting podcast yesterday that had an interview with a stats guy from Opta. It sounds like Opta is way behind what the individual clubs/coaches are doing with stats, but we can't be sure because it's all behind closed doors.

 

As a coach (or as any educated observer of the the game could see) it's obvious they're missing some massive data points by only tracking on-the-ball actions.

 

Anyway, I thought it was interesting. I'd really like to see an open data movement in football. Here's a link for anyone wanting to listen:

 

http://fivethirtyeight.com/features/in-the-era-of-advanced-stats-soccer-still-lags-behind/

 

[Dinho lad]

Stats are for people with no knowledge, it's raw information without insight.

 

Do you still stand by this, out of interest?

[Guest firetotheworks]

Stats are for people with no knowledge, it's raw information without insight.

 

Do you still stand by this, out of interest?

 

It was a bit of an inflammatory remark, but I stand by the crux of what I meant. Stats are obviously very useful when they're analysed correctly and used in conjunction with a knowledge of the game, but raw stats on their own are generally worthless without any knowledge for the non-quantifiable circumstances (e.g. your own opinion of a player/partnership/team regardless of the stats, how a player will lift the team/crowd, and then there's form/confidence and who the opposition is etc)

 

Beyond that I think stats tend be used very heavily by managers who are overly defensive and use stats on their own to play percentage football. It's how you end up with situations where flair players who can win matches single-handedly are dropped in favour of inferior players because their stats aren't up to scratch.

 

Good managers with knowledge beyond stats use them well, Rafa, Mourinho for example. Others such as Pardew, Allardyce, and Pulis, use them as gospel, because that's the majority of what they have to offer - percentages.

[TRon]

Stats are for people with no knowledge, it's raw information without insight.

 

Do you still stand by this, out of interest?

 

It was a bit of an inflammatory remark, but I stand by the crux of what I meant. Stats are obviously very useful when they're analysed correctly and used in conjunction with a knowledge of the game, but raw stats on their own are generally worthless without any knowledge for the non-quantifiable circumstances (e.g. your own opinion of a player/partnership/team regardless of the stats, how a player will lift the team/crowd, and then there's form/confidence and who the opposition is etc)

 

Beyond that I think stats tend be used very heavily by managers who are overly defensive and use stats on their own to play percentage football. It's how you end up with situations where flair players who can win matches single-handedly are dropped in favour of inferior players because their stats aren't up to scratch.

 

Good managers with knowledge beyond stats use them well, Rafa, Mourinho for example. Others such as Pardew, Allardyce, and Pulis, use them as gospel, because that's the majority of what they have to offer - percentages.

 

 

It would be daft to dismiss stats, they should be part of a coaching analysis set up. But unless stats can measure the drop of a shoulder, or a feint which opens up a defence, then they are only going to give you a rudimentary picture. If stats were the answer, then a team like Stoke under Pulis would have been wiping the floor with Barcelona every time. But of course if you put that to someone like Allardyce he'd claim that Barcelona play long ball football if you watch them closely.

[Mole_Toonfan]

KI

[Dinho lad]

Stats are for people with no knowledge, it's raw information without insight.

 

Do you still stand by this, out of interest?

 

It was a bit of an inflammatory remark, but I stand by the crux of what I meant. Stats are obviously very useful when they're analysed correctly and used in conjunction with a knowledge of the game, but raw stats on their own are generally worthless without any knowledge for the non-quantifiable circumstances (e.g. your own opinion of a player/partnership/team regardless of the stats, how a player will lift the team/crowd, and then there's form/confidence and who the opposition is etc)

 

Beyond that I think stats tend be used very heavily by managers who are overly defensive and use stats on their own to play percentage football. It's how you end up with situations where flair players who can win matches single-handedly are dropped in favour of inferior players because their stats aren't up to scratch.

 

Good managers with knowledge beyond stats use them well, Rafa, Mourinho for example. Others such as Pardew, Allardyce, and Pulis, use them as gospel, because that's the majority of what they have to offer - percentages.

 

Agree.

[Guest Howaythetoon]

The thing with stats, its how they get interpreted. If used correctly they can only add to things. You cannot ignore them in the modern game as they are pretty much essential these days.

 

That said, the best side I have ever seen and the best season we have had since we last won something were not told about their opponents, shown stats or anything. They were just sent out to score as many goals as they could, to entertain the fans and to express themselves as individuals and as a team.

 

I'd love to see such an approach today, I reckon such a team if it had enough good players, would do very well indeed.

[LucaAltieri]

Stats are for people with no knowledge, it's raw information without insight.

 

Do you still stand by this, out of interest?

 

It was a bit of an inflammatory remark, but I stand by the crux of what I meant. Stats are obviously very useful when they're analysed correctly and used in conjunction with a knowledge of the game, but raw stats on their own are generally worthless without any knowledge for the non-quantifiable circumstances (e.g. your own opinion of a player/partnership/team regardless of the stats, how a player will lift the team/crowd, and then there's form/confidence and who the opposition is etc)

 

Beyond that I think stats tend be used very heavily by managers who are overly defensive and use stats on their own to play percentage football. It's how you end up with situations where flair players who can win matches single-handedly are dropped in favour of inferior players because their stats aren't up to scratch.

 

Good managers with knowledge beyond stats use them well, Rafa, Mourinho for example. Others such as Pardew, Allardyce, and Pulis, use them as gospel, because that's the majority of what they have to offer - percentages.

 

 

It would be daft to dismiss stats, they should be part of a coaching analysis set up. But unless stats can measure the drop of a shoulder, or a feint which opens up a defence, then they are only going to give you a rudimentary picture. If stats were the answer, then a team like Stoke under Pulis would have been wiping the floor with Barcelona every time. But of course if you put that to someone like Allardyce he'd claim that Barcelona play long ball football if you watch them closely.

 

It depends what you're looking for. If you're looking at individual players, then sure. If you're looking at it from a team game plan or system of play point of view then you don't really care how a player beats an opponent. The specific move doesn't matter.

 

From a coaching point of view, you sort of have to look at it as a series of battles in different parts of the pitch. The are lots of variables to consider. Most of them measurable. But the specific parts that are focused on varies slightly from manager to manager.

 

Overall though, managers in the Pep, van Gaal, and Bielsa style are using different metrics to calculate penetration into the "right" parts of the pitch. Passing, moving, dribbling into key zones for their style of build-up play. They put their own spin on it... Bielsa with his direct forward passing, a lot of player rotation, frequent use of the up-back-and-through pattern. Pep focuses more on trying to drag defenders out of position through ball rotation then penetrate quickly. There are lots of useful stats but they're all particular to the style of play deployed. Roger Schmidt is less about the position on the field and more about the shape of his team in transition (ie. changing from defense to attack and vice versa) and trying to control the game just in the transitions. But even then you're talking about relative position on the pitch, position of teammates, position of opponents as a potential trigger for pressing.

 

Anyway, point being, if a manager has a game plan of wanting to penetrate behind the right full-back and get the ball in towards the penalty spot, then he doesn't really care about body feints. It's enough to just measure how many passes the winger got on the end of, how much time and space they could create for themselves, how many times did they dribble past the full-back, how many times did they get the ball into the right area, that sort of thing.

 

I'm probably using a poor example there with an attacking goal. The consensus seems to be that defences need to be organized and drilled to manage things predictably. Whereas the final third of the field should be open to creative endeavors with no hard and fast targets. It needs to be fluid and unpredictable or you're easy to defend against. Famously Pep didn't have a game plan for the final third other than get the ball to the forwards in particular areas then left them go with it.

 

Still, I'm sure you get the idea.

[LucaAltieri]

Pretty interesting.

 

Also uninteresting. Didn't particularly need stats to come to the conclusion Bravo is shit.

 

http://statsbomb.com/2017/05/the-worst-transfers-of-the-2016-17-premier-league-season/