Do football possession statistics indicate which team will win? Not necessarily
Jonathan Liew investigates: Perceived wisdom is that good teams keep the ball, and bad teams do not. But should we be looking at a different metric?
By Jonathan Liew
1:09PM BST 29 Apr 2014
Possession is nine-tenths of the law, so the saying goes, which meant that with 74 per cent possession against Chelsea on Sunday, Liverpool enjoyed 66.6 per cent of the law. You can prove anything with statistics.
So, possession.
You'll often see it at the bottom of the screen during a live game, and you'll think: "Oooh, that's interesting. Team A have had 61 per cent of the ball. And yet it's still goalless! What an enigmatic mistress football is sometimes."
Or you'll see it on the screen at the end of a match and think: "Wow. Team A won 1-0, with 75 per cent of the possession."
Then you'll turn to your mates with a knowing look and tell them: "The scoreline may not show it, but that's a thrashing."
This must stop.
Why do people care about possession statistics?
The main reason people care about possession statistics is that football games are, for the most part, very long and quite boring. In the age of the six-second Vine video, a sport that lasts 90 minutes and might well end with nothing happening is not the sort of sport you could ever get away with inventing today.
If you took the idea for football to a marketing expert, or perhaps one of the Dragons, they would tell you to shorten it to about 20 minutes and add a lot more goals. And get Emeli Sande involved somehow.
But we have what we have. And so, with no guarantee of a winner and long periods in which nothing very interesting will happen at all, into the vacuum step 'key performance indicators'. These help us to 'understand' how the game is going, or perhaps how it went, or perhaps how it might go.
As one of the oldest and simplest statistics, possession has taken on a sort of hallowed status within football, exalted by luminaries as diverse as Johan Cruyff (“If you play on possession, you don’t have to defend, because there’s only one ball”), Tim Sherwood ("If you don't lose the ball, you ain't got to get it back") and Mark Lawrenson in the commentary for Pro Evolution Soccer 5 ("Coaches so often say, treat the ball as a friend").
When Barcelona were winning everything, this was fine.
But how useful a stat is possession anyway?
This was all fine and dandy when Barcelona were winning everything. Possession = Goals = Wins = Trophies. A simple linear relationship.
England were condemned for having just 32 per cent possession in the goalless draw against Italy at Euro 2012, not praised for holding Italy to a goalless draw despite having just 32 per cent of possession.
Manchester United’s 2-0 win over Swansea in January was greeted with a torrent of unfavourable headlines, because their 39 per cent of possession was their worst in Premier League history.
A kind of lazy shorthand has evolved: good teams keep the ball, and by extension bad teams do not.
The trouble invariably comes when teams triumph in spite of inferior possession, not because of it. Real Madrid will take a 1-0 lead to Munich tonight, despite having just 28 per cent of possession in the Bernabéu.
When Chelsea held on for a 2-2 in the Nou Camp in 2012 despite having John Terry sent off, they had just 17.5 per cent of the ball. Xavi completed more passes (161) than the entire Chelsea team. Barcelona had 72 per cent possession over the two legs, and yet lost.
Go through the Premier League records for this season and you’ll find similar instances. Cardiff beat Manchester City on 30 per cent possession.
Aston Villa beat Southampton on less than 22. Once you’ve taken out their five shots and the kick-off, that’s a lot of running and chasing and pressing: like a training session from the 1950s, when the prevailing wisdom was that starving players of the ball during the week would make them hungrier for it on Saturday.
Lowest possession by winning teams, Premier League 2013/14
So, possession.
You'll often see it at the bottom of the screen during a live game, and you'll think: "Oooh, that's interesting. Team A have had 61 per cent of the ball. And yet it's still goalless! What an enigmatic mistress football is sometimes."
Or you'll see it on the screen at the end of a match and think: "Wow. Team A won 1-0, with 75 per cent of the possession."
Then you'll turn to your mates with a knowing look and tell them: "The scoreline may not show it, but that's a thrashing."
This must stop.
Why do people care about possession statistics?
The main reason people care about possession statistics is that football games are, for the most part, very long and quite boring. In the age of the six-second Vine video, a sport that lasts 90 minutes and might well end with nothing happening is not the sort of sport you could ever get away with inventing today.
If you took the idea for football to a marketing expert, or perhaps one of the Dragons, they would tell you to shorten it to about 20 minutes and add a lot more goals. And get Emeli Sande involved somehow.
But we have what we have. And so, with no guarantee of a winner and long periods in which nothing very interesting will happen at all, into the vacuum step 'key performance indicators'. These help us to 'understand' how the game is going, or perhaps how it went, or perhaps how it might go.
As one of the oldest and simplest statistics, possession has taken on a sort of hallowed status within football, exalted by luminaries as diverse as Johan Cruyff (“If you play on possession, you don’t have to defend, because there’s only one ball”), Tim Sherwood ("If you don't lose the ball, you ain't got to get it back") and Mark Lawrenson in the commentary for Pro Evolution Soccer 5 ("Coaches so often say, treat the ball as a friend").
When Barcelona were winning everything, this was fine.
But how useful a stat is possession anyway?
This was all fine and dandy when Barcelona were winning everything. Possession = Goals = Wins = Trophies. A simple linear relationship.
England were condemned for having just 32 per cent possession in the goalless draw against Italy at Euro 2012, not praised for holding Italy to a goalless draw despite having just 32 per cent of possession.
Manchester United’s 2-0 win over Swansea in January was greeted with a torrent of unfavourable headlines, because their 39 per cent of possession was their worst in Premier League history.
A kind of lazy shorthand has evolved: good teams keep the ball, and by extension bad teams do not.
The trouble invariably comes when teams triumph in spite of inferior possession, not because of it. Real Madrid will take a 1-0 lead to Munich tonight, despite having just 28 per cent of possession in the Bernabéu.
When Chelsea held on for a 2-2 in the Nou Camp in 2012 despite having John Terry sent off, they had just 17.5 per cent of the ball. Xavi completed more passes (161) than the entire Chelsea team. Barcelona had 72 per cent possession over the two legs, and yet lost.
Go through the Premier League records for this season and you’ll find similar instances. Cardiff beat Manchester City on 30 per cent possession.
Aston Villa beat Southampton on less than 22. Once you’ve taken out their five shots and the kick-off, that’s a lot of running and chasing and pressing: like a training session from the 1950s, when the prevailing wisdom was that starving players of the ball during the week would make them hungrier for it on Saturday.
Lowest possession by winning teams, Premier League 2013/14
So, possession.
You'll often see it at the bottom of the screen during a live game, and you'll think: "Oooh, that's interesting. Team A have had 61 per cent of the ball. And yet it's still goalless! What an enigmatic mistress football is sometimes."
Or you'll see it on the screen at the end of a match and think: "Wow. Team A won 1-0, with 75 per cent of the possession."
Then you'll turn to your mates with a knowing look and tell them: "The scoreline may not show it, but that's a thrashing."
This must stop.
Why do people care about possession statistics?
The main reason people care about possession statistics is that football games are, for the most part, very long and quite boring. In the age of the six-second Vine video, a sport that lasts 90 minutes and might well end with nothing happening is not the sort of sport you could ever get away with inventing today.
If you took the idea for football to a marketing expert, or perhaps one of the Dragons, they would tell you to shorten it to about 20 minutes and add a lot more goals. And get Emeli Sande involved somehow.
But we have what we have. And so, with no guarantee of a winner and long periods in which nothing very interesting will happen at all, into the vacuum step 'key performance indicators'. These help us to 'understand' how the game is going, or perhaps how it went, or perhaps how it might go.
As one of the oldest and simplest statistics, possession has taken on a sort of hallowed status within football, exalted by luminaries as diverse as Johan Cruyff (“If you play on possession, you don’t have to defend, because there’s only one ball”), Tim Sherwood ("If you don't lose the ball, you ain't got to get it back") and Mark Lawrenson in the commentary for Pro Evolution Soccer 5 ("Coaches so often say, treat the ball as a friend").
When Barcelona were winning everything, this was fine.
But how useful a stat is possession anyway?
This was all fine and dandy when Barcelona were winning everything. Possession = Goals = Wins = Trophies. A simple linear relationship.
England were condemned for having just 32 per cent possession in the goalless draw against Italy at Euro 2012, not praised for holding Italy to a goalless draw despite having just 32 per cent of possession.
Manchester United’s 2-0 win over Swansea in January was greeted with a torrent of unfavourable headlines, because their 39 per cent of possession was their worst in Premier League history.
A kind of lazy shorthand has evolved: good teams keep the ball, and by extension bad teams do not.
The trouble invariably comes when teams triumph in spite of inferior possession, not because of it. Real Madrid will take a 1-0 lead to Munich tonight, despite having just 28 per cent of possession in the Bernabéu.
When Chelsea held on for a 2-2 in the Nou Camp in 2012 despite having John Terry sent off, they had just 17.5 per cent of the ball. Xavi completed more passes (161) than the entire Chelsea team. Barcelona had 72 per cent possession over the two legs, and yet lost.
Go through the Premier League records for this season and you’ll find similar instances. Cardiff beat Manchester City on 30 per cent possession.
Aston Villa beat Southampton on less than 22. Once you’ve taken out their five shots and the kick-off, that’s a lot of running and chasing and pressing: like a training session from the 1950s, when the prevailing wisdom was that starving players of the ball during the week would make them hungrier for it on Saturday.
Lowest possession by winning teams, Premier League 2013/14
The question is whether possession is actually any use at telling us who’s going to win a football game. And the answer is: it really depends where you look.
In this season’s Premier League, the team with more possession has won 55.3 per cent of games. In the Champions League, the figure is 66.7 per cent. That would seem to indicate a strong correlation between keeping the ball and winning the game.
But as a means for predicting the winner of a football game, possession is deeply unreliable. A far better metric in this regard is shots on goal, or “shot supremacy”: the ratio of shots on goal to shots conceded, which has been proven to have a strong correlation with points.
More often, possession is the by-product of a good team, rather than the other way round. The higher the standard of the competition, the more likely you are to find players with the skill levels required to play successful possession football.
So what happens when the standard is lower? To study this, we looked at a competition some distance removed from the dizzy heights and fancy pirouettes of the Champions League.
Completely at random, and with apologies, we picked the Australian A-League. The standard may be improving, but only a complete dingo dog would claim it was a serious rival to the big European leagues. So, with help from the excellent FourFourTwo Stats Zone app, we looked at all 105 matches played so far this season. And the results were mildly startling.
In the Champions League, remember, the team with more possession wins 67 per cent of the time, if you exclude draws. In the A-League, the team with more possession loses 57 per cent of the time. More remarkably still, the team that completes more passes in the opposition’s final third – proper passing, not just knocking it around the defence – loses 59 per cent of the time.
If you are setting up a team to play in the Australian A-League, it seems that the best thing to do is to kick the ball away as fast as you can. Conclusions
Why should this be? Perhaps it’s because players of a lower standard are more likely to make mistakes on the ball, more likely to send passes astray, more likely to make a fatal error in a dangerous position. Possession football requires a certain skill level to carry off well, and yet even then it is not foolproof, as Barcelona’s occasional slip-ups have proved. When inferior players try to play keep-ball, it often proves counterproductive.
According to the Spanish journalist Diego Torres in a recent book, Jose Mourinho’s attitude to possession at Real Madrid was summed up thus.
1. The game is won by the team who commits fewer errors.
2. Football favours whoever provokes more errors in the opposition.
3. Away from home, instead of trying to be superior to the opposition, it's better to encourage their mistakes.
4. Whoever has the ball is more likely to make a mistake.
5. Whoever renounces possession reduces the possibility of making a mistake.
6. Whoever has the ball has fear.
7. Whoever does not have it is therefore stronger.
This is tiki-taka’s direct nemesis: not necessarily anti-football, but definitely anti-something. The data from Australia proves that below a certain skill threshold, football is primarily a game of mistakes.
Mourinho’s genius is in enacting his theory at the very highest level of football.
http://www.telegraph.co.uk/sport/football/competitions/champions-league/10793482/Do-football-possession-statistics-indicate-which-team-will-win-Not-necessarily.html