This is an article by Jonathon Wilson for the Guardian, normally I'm a really big fan of his work, anyone with an interest in football tactics should read his book inverting the pyramid. Not his greatest work this article but some relatively interesting facts I suppose.
Not since 1903 have Liverpool had a worse start to the season but for the most part their supporters seem relatively sanguine. The fixture list has not been kind, offering up home games against three of the sides likely to fill the top four positions in the Premier League at the end of the season and testing away games at West Bromwich Albion and Sunderland. And Liverpool have played pretty well – even in the home defeat by Arsenal, when they ended up comfortably beaten 2-0, there was an hour or so in which they controlled the ball.
The perception is of individual errors having undermined decent team performances and the assumption is that once those are ironed out, particularly as the fixtures become a little easier, there will be significant improvement.
But then Liverpool fans last season got used to results not matching up to performances. The statistics are well known and hard to comprehend; Liverpool finished eighth, their worst finish in over half a century, and yet in the top five leagues in Europe only Barcelona had a greater proportion of their play in the attacking third of the pitch.
In 33 of 38 league games Liverpool had more shots than their opponents. In a table of shots on goal in league games, Liverpool were fourth. They missed five of their six penalties and hit the woodwork a total of 33 times; had even half of those attempts gone in Liverpool would have finished the season with the third best goal difference.
Egil Olsen, the Norway coach who has invested vast amounts of time studying chance creation, insists that in three out of four games the team who creates more chances wins. Liverpool last season were in the other quarter an awful lot.The theory I put forward then – and it's not one I have great confidence in – was that they might be creating the wrong sort of chance. Could it be that a game based on crossing creates more chances that are harder to take than the fewer chances created by a more cautious passing game? And was that exacerbated with a forward of Luis Suárez's type?
In an interview with Lars Sivertsen in Issue Three of The Blizzard, Olsen was adamant that the quality of the chance doesn't matter too much. "It would be possible to try to create chance one, two and three, according to sizes, big, medium and small," he said. "But it turns out that with those of us who work with this, when we're counting chances we usually end up on the same number; there could be some variations because it's a subjective matter, but I think it's the best indicator of the nature of a football game, after the result." That doesn't take into account the type of chance, admittedly, but Olsen insists most teams take one chance in three or four. This actually sounds remarkably high – Charles Reep maintained it was around one in nine – which suggests Olsen has a strict interpretation of what constitutes a chance.
So how have things changed since Brendan Rodgers took charge of Liverpool and began to institute his more possession-based approach?
Average possession hasn't actually gone up quite as much as you might expect, with Opta figures showing an increase from 55.2% to 55.7%. Nor has the number of passes attempted gone up that much: Liverpool have attempted 2,599, which extrapolates to 19,752 over the whole season, as opposed to 18,794 last season.
As Liverpool face weaker opposition that figure is likely to rise further. Already there has been a marked improvement in pass completion, up from 80.9% to 86%, and from 73.24% to 78.9% in the opposition half. To put that into context, last year's figures were the best for Liverpool over a full season since Opta began collating data in 2006-07.
The problem is all too familiar. Last season Liverpool converted only 9.13% of their chances, the lowest figure since Opta began producing its reports. It had been 13.59% the previous season but then Liverpool managed just 11.4 shots per game. When Liverpool finished second in 2008-09, they had 15.1 shots per game and converted 13.46% of their chances. But if 9.13% looked bad, this season's figure is far worse: a meagre 5.97%.
That's not the only issue. A downturn in cross completion is only to be expected and a slight fall-off in tackles won is too small to be significant. But Liverpool have won only 48.11% of duels this season as opposed to 53.17% last. It's not a vast difference and it may improve, particularly as Liverpool face weaker teams, but that's a big enough fall off to raise the odd question.
Even more striking are José Reina's stats. Goalkeepers are hard to analyse – Peter Shilton once claimed his best games often came when he hardly touched the ball but had organised his defence so well he didn't have to – and, perhaps, saves are overrated (apologies for another plug but for more on this, see my book on goalkeepers, The Outsider, out in December), but two things stand out about the modern Reina compared to the player who joined the club in 2005. For seven seasons his save percentage hovered between 69% (last season) and 75% (2008-09); this season it stands at 44%. Again, it's early, but when you consider that of the six mistakes that Opta categorise as "leading to a goal" he has committed at Liverpool, five have come since the start of last season, the impression is confirmed of a keeper suffering a sustained spell of indifferent form.
The good news for Rodgers is that passing, the absolute core of his philosophy, has improved, even given the tough start. The bad news is that individual errors keep undermining the effectiveness of the overall structure.
Not since 1903 have Liverpool had a worse start to the season but for the most part their supporters seem relatively sanguine. The fixture list has not been kind, offering up home games against three of the sides likely to fill the top four positions in the Premier League at the end of the season and testing away games at West Bromwich Albion and Sunderland. And Liverpool have played pretty well – even in the home defeat by Arsenal, when they ended up comfortably beaten 2-0, there was an hour or so in which they controlled the ball.
The perception is of individual errors having undermined decent team performances and the assumption is that once those are ironed out, particularly as the fixtures become a little easier, there will be significant improvement.
But then Liverpool fans last season got used to results not matching up to performances. The statistics are well known and hard to comprehend; Liverpool finished eighth, their worst finish in over half a century, and yet in the top five leagues in Europe only Barcelona had a greater proportion of their play in the attacking third of the pitch.
In 33 of 38 league games Liverpool had more shots than their opponents. In a table of shots on goal in league games, Liverpool were fourth. They missed five of their six penalties and hit the woodwork a total of 33 times; had even half of those attempts gone in Liverpool would have finished the season with the third best goal difference.
Egil Olsen, the Norway coach who has invested vast amounts of time studying chance creation, insists that in three out of four games the team who creates more chances wins. Liverpool last season were in the other quarter an awful lot.The theory I put forward then – and it's not one I have great confidence in – was that they might be creating the wrong sort of chance. Could it be that a game based on crossing creates more chances that are harder to take than the fewer chances created by a more cautious passing game? And was that exacerbated with a forward of Luis Suárez's type?
In an interview with Lars Sivertsen in Issue Three of The Blizzard, Olsen was adamant that the quality of the chance doesn't matter too much. "It would be possible to try to create chance one, two and three, according to sizes, big, medium and small," he said. "But it turns out that with those of us who work with this, when we're counting chances we usually end up on the same number; there could be some variations because it's a subjective matter, but I think it's the best indicator of the nature of a football game, after the result." That doesn't take into account the type of chance, admittedly, but Olsen insists most teams take one chance in three or four. This actually sounds remarkably high – Charles Reep maintained it was around one in nine – which suggests Olsen has a strict interpretation of what constitutes a chance.
So how have things changed since Brendan Rodgers took charge of Liverpool and began to institute his more possession-based approach?
Average possession hasn't actually gone up quite as much as you might expect, with Opta figures showing an increase from 55.2% to 55.7%. Nor has the number of passes attempted gone up that much: Liverpool have attempted 2,599, which extrapolates to 19,752 over the whole season, as opposed to 18,794 last season.
As Liverpool face weaker opposition that figure is likely to rise further. Already there has been a marked improvement in pass completion, up from 80.9% to 86%, and from 73.24% to 78.9% in the opposition half. To put that into context, last year's figures were the best for Liverpool over a full season since Opta began collating data in 2006-07.
The problem is all too familiar. Last season Liverpool converted only 9.13% of their chances, the lowest figure since Opta began producing its reports. It had been 13.59% the previous season but then Liverpool managed just 11.4 shots per game. When Liverpool finished second in 2008-09, they had 15.1 shots per game and converted 13.46% of their chances. But if 9.13% looked bad, this season's figure is far worse: a meagre 5.97%.
That's not the only issue. A downturn in cross completion is only to be expected and a slight fall-off in tackles won is too small to be significant. But Liverpool have won only 48.11% of duels this season as opposed to 53.17% last. It's not a vast difference and it may improve, particularly as Liverpool face weaker teams, but that's a big enough fall off to raise the odd question.
Even more striking are José Reina's stats. Goalkeepers are hard to analyse – Peter Shilton once claimed his best games often came when he hardly touched the ball but had organised his defence so well he didn't have to – and, perhaps, saves are overrated (apologies for another plug but for more on this, see my book on goalkeepers, The Outsider, out in December), but two things stand out about the modern Reina compared to the player who joined the club in 2005. For seven seasons his save percentage hovered between 69% (last season) and 75% (2008-09); this season it stands at 44%. Again, it's early, but when you consider that of the six mistakes that Opta categorise as "leading to a goal" he has committed at Liverpool, five have come since the start of last season, the impression is confirmed of a keeper suffering a sustained spell of indifferent form.
The good news for Rodgers is that passing, the absolute core of his philosophy, has improved, even given the tough start. The bad news is that individual errors keep undermining the effectiveness of the overall structure.