18 April 2011

Possession isn't nine-tenths of the law anymore

Yesterday, Liverpool managed to earn a draw despite dramatically losing the possession battle. Arsenal had 62% of the ball, the most any opposition have had since Dalglish took the reins. Liverpool were under pressure and pegged back for long stretches of the match, but only conceded thanks to a sloppy penalty eight minutes into injury time.

Normally, you'd think that a bad thing. Even after the season we've seen, most of us are still used to the Rafa Benitez method deployed over six seasons: keep possession first and foremost. Blunt the opposition, set the tempo, play keep-away. Spain and Barcelona have won everything under the sun in recent years with similar (if far more effective) tactics, so it must be the "right method." But that hasn't been the case for Liverpool this season. And more often than not, losing the possession battle has actually been a good thing.

Simply put, whether under Hodgson or Dalglish, Liverpool have been better in matches where the opposition has more of the ball.

In the league, Liverpool have eight wins, four draws, and nine losses – an average of 1.33 points per game – when they out-possess their opponents. When the opposition has more possession, Liverpool have won six, drawn three, and lost three – an average of 1.75 points per game, and a difference of almost a half-point per game.

Liverpool have been held under 45% possession in six matches this season: v Arsenal, v Chelsea, at Wolves, at Chelsea, v United, and at Arsenal. Twice under Hodgson, four times under Dalglish. And Liverpool are unbeaten in all six, winning four while drawing twice against the Gunners.

The disparity grows when just considering Dalglish's 13 games: 3W-2D-2L with more possession than the opposition, 4W-1D-1L when the opposition out-possesses Liverpool. An average of 1.57 points per game versus 2.17 points per game. The games won with more possession have been against City, Stoke, and Fulham (all at Anfield); the games won with less possession were against Sunderland, United, Chelsea, and Wolves (all but United away from home). Getting results on the road has happened with defense and counter-attacking football. Finally.

It's been a trend all season, but the January changes were the turning point. Dalglish and Clarke have done a far better job organizing the team, especially the defense. Exchanging Torres for Suarez and Carroll improved Liverpool's ability to counter-attack, with both better at bringing in midfield runners such as Kuyt and Meireles. Liverpool are scoring more goals – having notched in every one of Dalglish's league matches – and conceding less.

But they're doing it with some similarities to Hodgson's team. As said above, players have changed, as have certain tactics (pressing higher up the pitch being the most noticeable). But Liverpool's still reliant on the counter-attack, back in a 4-4-2/4-2-2-2 formation since injuries to Gerrard and multiple defenders, and still getting better results when conceding more possession to the opposition.

Clearly, it's not so simple as just having less possession or the likes of Stoke, Bolton, and Blackburn (or whoever Hodgson's managing) would win the league every season. But it's strange to see such a variation in results contingent upon how much of the ball Liverpool has.

At the end of August, I lamented Liverpool's lack of possession, unhappy with early performances and results under the new manager and still expecting Benitez's metronomic tactics.

Sometimes, a tiger can change its stripes. But it needs the right manager to do so.


nate said...

Full list of possession statistics for league matches. Games where Liverpool had less possession are in italics.

LFC 38-62% Arsenal (a)
LFC 53-47% MCFC (h)
LFC 48-51% WBA (a)
LFC 48-52% Sunderland (a)
LFC 42-58% United (h)
LFC 58-42% West Ham (a)
LFC 53-47% Wigan (h)
LFC 43-57% Chelsea (a)
LFC 67-33% Stoke (h)
LFC 54-46% Fulham (h)
LFC 41-59% Wolves (a)
LFC 56-44% Everton (h)
LFC 53-47% Blackpool (a)
LFC 67-33% Blackburn (a)
LFC 63-37% Bolton (h)
LFC 56-44% Wolves (h)
LFC 55-45% Newcastle (a)
LFC 49-51% Villa (h)
LFC 49.5-50.5% Spurs (a)
LFC 54-46% West Ham (h)
LFC 60-40% Stoke (a)
LFC 46-54% Wigan (a)
LFC 35-65% Chelsea (h)
LFC 59-41% Bolton (a)
LFC 58-42% Blackburn (h)
LFC 67-33% Everton (a)
LFC 57-42% Blackpool (h)
LFC 51-49% WBA (h)
LFC 45-55% MCFC (a)
LFC 36-64% Arsenal (h)

Seth said...

I haven't looked into this, but I'd be interested to know how much of our opponents' possession is due to us trying to hold onto a lead, and therefore packing it in later in games. Could be wrong, but I think perhaps a better indicator of whether possession (or lack there of) = results would be to analyze possession stats during minutes where the score is tied.

nate said...

Yeah, I don't have access to that kind of data.

There are some obvious games where that happened – conceding possession to hold a lead – under both managers. Villa's possession advantage in the 3-0 home win, Wolves' advantage in the 3-0 away win, Chelsea's advantage in the 2-0 home win, United's advantage in the 3-1 home win. But in matches like 1-0 Chelsea (a) and 1-1 Arsenal (a), the tactic was to concede possession and attack on the counter, which has worked surprisingly well this season, especially under Dalglish.

Really, it's just a quirky statistic which caught my eye, like many I post about. Each match has to be assessed on its own. But there are some interesting conclusions to be drawn.

Anonymous said...

Very interesting post, Nate. I do find it odd that Dalglish's side is giving up similar amounts of possession to the opponent. Something I don't think that can be emphasized enough is the change in confidence. Under Roy, it seemed as the the life was being sucked from Liverpool. Things are starting to look up know, despite the many injuries. I imagine that the players are more confident to play under Dalglish, and I think we're likely to retain the talent we have (I'm thinking Pepe) but also attract more talent. Suarez's addition was phenomenal for the side, and Carroll is looking good too.

red2death said...

Seems like the age-old problem of us not being able to break teams down. Nevermind Dalglish and Hodgson, that's been a problem all through Rafa's time and even for part of Houllier's. When the opposition puts 9 men behind the ball, we just can't find a way through, and often end up dropping points.

And naturally we have more possession in those games where opponents choose to go defensive because they give it to us. So, I'd say possession stats are just the symptom here, not the cause.

Seth said...

I do wish we had more data. As an American soccer fan, I feel compelled to offer some more money-ball esque contributions to analyzing the game (and I very much appreciate what you do with your blog here on that front).

I do think someone should start tracking what I'd like to coin now as "neutral possession" - a team's possession % during the game when the score is tied. This rests upon the foundation that not all possession is equal. No one likes for a team who gives up a goal 15 minutes in and loses 1-0 to complain after the game that they "had all of the ball and were unlucky to lose" when in fact the other team ceded possession for the remaining 75 minutes in order to cement their well earned goal.