Pass completion statistics are fast becoming a popular measure of player evaluation, no doubt aided by the fact that Barcelona have set a benchmark at the top of club football.
Like all statistics though, and perhaps even more so for passing data, more context is required, particularly when you note thatLeon Britton has trumped Xavi in the pass completion stakes so far this season. However, looking at the other numbers it’s clear to see why Xavi is a Champions League and World Cup winner, and why Leon Britton is not.
I’m particularly interested in passing efficiency in the final third; an area of the pitch that you’d think requires tremendous ability to be successful. But first, I’ve attempted to find some correlations in Premier League passing data this season.
It’s important to bear in mind none of these can be attributed as causal effects, but nevertheless:
- Both higher pass completion and a greater number of successful passes are significantly correlated with winning.
- Both higher pass completion in the final third and a greater number of successful passes in this area are also correlated with winning.
- A greater number of final third entries is uncorrelated with winning.
- Making a greater proportion of your passes in the final third is uncorrelated with winning.
The problem with the significant correlations is that passing success and winning can be a self-perpetuating cycle; teams that pass better may be more inclined to score, but scoring can also give teams confidence to pass it better. Thus we cannot definitively say that better pass completion increases your chances of winning.
To identify whether passing success really leads to success in results, we’d need to break down the data to see if successful passing is correlated with scoring the next (and particularly first) goal in a match.
Without this separation, I’ve just looked at general trends in passing so far this season. First, as you’d expect, teams with better pass completions across the entire pitch tend to have better pass completions in the final third too.
Teams below the line of best fit pass worse in the final third than you’d expect, and teams above the line pass better in the final third. Two notable teams, therefore, are Blackburn and Swansea, who somewhat underachieve when trying to pass the ball in the final third. The chart suggests that much of Swansea’s passing success is biased by passes completed in easier areas of the pitch. Blackburn, meanwhile, are only better than Stoke when it comes to completing passes in the final third.
Swansea’s style of play is a recurring theme in passing analysis this season, the Welsh team have grown into the division after a cautious passing display on the opening day at Man City. That match was typified by an inability to get the ball into dangerous areas, and whilst they have improved, the vast majority of their time on the ball is spent in parts of the pitch that will not trouble the opposition.
That’s not to say it’s a bad thing, clearly good pass completion in the defensive two-thirds of the pitch will tire opposition players, opening up opportunities later in the game. It’s interesting that two sides that have had frustrating seasons – Merseyside clubs Liverpool and Everton – have completed a number of their passes in attacking areas. Clearly there’s been plenty of probing football, and opposition teams have been happy to accommodate this.
This is supported by the number of successful final third passes it takes the two clubs to score a goal; more than twice as many as the most efficient team, Blackburn.
Blackburn complete so few passes in the final third, that when they can put a few together they have proved effective, perhaps indicative of teams pushing forward against Steve Kean’s side.
The league’s top two teams have been equally efficient. This is impressive; we’d expect teams would make it hard for the Manchester clubs to score, and even if they have, City and United been ruthless in the final third, needing fewer than 50 successful passes on average before scoring.
They have also been two of the most efficient teams defensively, and as ever, defensive success is better correlated with good results than attacking ability. Arsenal’s problems off the ball do not appear to have improved on last season, whilst Swansea can be pleased that even though they often out-pass the opposition, they aren’t vulnerable without the ball in dangerous areas. Other than Sunderland, the top six teams on the chart above haven’t had to ‘absorb’ pressure as much during games, which may help them appear more efficient defensively.
There are a long of strong relationships between passing across the entire pitch and passing in the final third, but focusing on the attacking and defensive thirds of the field help gives context and perspective to some teams’ numbers.
This post is also available on 5 Added Minutes.
All of the Opta stats used in this article are from our EPLIndex Stats Centre. Subscribe now for Access and write your own article on EPLIndex.com!
- Excellent
- Informative
- Awesome
- Good Read
- ok
Categories: Arsenal (NN), Aston Villa, Blackburn Rovers, Bolton Wanderers, Chelsea, EPL, EPL Index Featured Article, EPL Index Statistical Comparisons, EPL Opta Stats, Everton, Fulham, Liverpool, Manchester City, Manchester Utd, Newcastle Utd, Norwich City, Stoke City, Sunderland, Swansea City, Tottenham Hotspur, West Bromwich Albion, Wigan, Wolves
Tags: english Premier league, EPL, epl opta stats, EPL Stats, Final Third Passing, Opta Stats, premier league, Premier League Stats
This article has had 20,658 Views








Superb analysis. Thanks for posting this Omar. Appreciate it.
Not at all, the numbers hopefully speak for themselves!
What might be interesting is comparing the “per goal scored” chart to “successful passes in final third per shot made” and “per shot on goal made.” It might point out where Liverpool’s problems are, which could be any of: 1. opportunity creation once passes are made; 2. shooting accuracy; or 3. ball striking ability.
Excellent post, Omar. Would like stats that could be sorted by minute and situation. Example – MCFC have frequently had a lead late in matches. Are they making more ‘defensive’ passes in the attacking third to run down the clock? That would raise their passing numbers but only be incidental to the scoreline.
Absolutely – scoreline is a huge amount of context that is missing from these numbers, which is why I wouldn’t draw any causal conclusions from the correlations yet. Disaggregated data is the next step, hopefully!
This is a top read.
Statistics are good for those who swear by them, but you can make them say what you want them to say, it’s points in May that win trophies, not pass completion rate
jack is admin @ http://www.manutdreds.net
Well that’s precisely what I’ve tried to avoid here; I’ve found a correlation between winning and pass completion, then sought to find trends. Of course points win trophies – but how do you collect points? Clearly good pass completion helps, and is a factor in winning games.
I love manchester united its an awesome team
Top notch piece Omar.Fair play!Excellent stuff
Firstly you need to define
What is a successful pass.
You need to define … Was it a final third entry or a final third possession entry.
There certainly is a correlation between winning and final third possesion entries.
The key thing in terms of passing is the teams ability to pass the ball into the final third with possesion when you win it in the defending and middle thirds. If you can’t do this then pass completion counts for nothing.
So I am sorry but much of this data is wide off the mark..
1) a successful pass: “An intentional played ball from one player to another.”
2) Final third entry – a pass into the final third
3) I found no correlation, but that may be due to teams who are winning backing off and not playing more final third entries – a lack of causation which I acknowledge.
4) Got any numbers to back that up? Genuinely interested
5) How is the data wide of the mark? The data is reliably sourced from Opta…
Omer
I am sorry to say please stick to economics since many of your conclusions are without substance.
I’m all ears to your points, but you’ve not really said where I’ve gone wrong. I’m genuinely interested where you get your conclusions from.
Omar this is good stuff, keep up the good work.
John, stop trolling and don’t read it if you don’t like it.
Omar
Thanks this very interesting.
Do you have stats for number of passes in the moves leading to shots and those leading to goals?
How many crosses lead to goals?
Pass length in the moves leading to shots and to goals?