A question for the stats guys, which are the best stats to measure passing ability? Everyone knows about pass completion but I am aware that there are many more interesting ones out there but don't really know them properly. I thought it could be helpful if posters were aware of the broader measures so as to better inform discussions - rather than X has pass completion under 90% he must be awful or Y has a pass completion of 95%, he must be amazing. If we had like five stats (for example) that everyone knew, it would improve the discussion.
What are the best stats to measure passing ability?
- Thread starter Physiocrat
- Start date
Best measure for me is stat in comparison, e.g. percentile compared to players in similar positions in the following categories:
1. Forward passes (ball traveling more than 10 m forward)
2. Passes in final third
For defensive players general stat like passing accuracy is a pretty good because it reflects their reliability in terms of keeping procession. For attacking players it’s useless, IMO. For example, Bruno.
1. Forward passes (ball traveling more than 10 m forward)
2. Passes in final third
For defensive players general stat like passing accuracy is a pretty good because it reflects their reliability in terms of keeping procession. For attacking players it’s useless, IMO. For example, Bruno.
First and foremost, I would use "per 90 minutes" to normalize the data and allow for better comparisons among players. A simple approach from there, would be to segment by the pass type (e.g. short, medium, long, through balls, crosses) and analyze attempts, completions, and completion percentage.
As for the more advanced stats, key passes and xA (expected assists) are useful for players in attacking roles and will give you a general idea of a player's ability to create goals for their team.
Another interesting one, Progressive Distance (per 90), would tell you how many yards a player's successful passes traveled toward the opponent's goal (backwards passes count as 0). This is useful to analyze how much a player's passes push their team up the pitch. However, this is very dependent on position and team tactics.
As for the more advanced stats, key passes and xA (expected assists) are useful for players in attacking roles and will give you a general idea of a player's ability to create goals for their team.
Another interesting one, Progressive Distance (per 90), would tell you how many yards a player's successful passes traveled toward the opponent's goal (backwards passes count as 0). This is useful to analyze how much a player's passes push their team up the pitch. However, this is very dependent on position and team tactics.
When a player has a 95% passing stat, it might be good but how do we tell which passes were the correct ones even if they went to a teammate?
Its a tricky one to measure on stats alone.
Do the stats count a pass completion if the receiver gets a touch on the ball? Is it measured if the receivers touch is a successful one? If the receiver loses possession due to a bad touch is it counted as a successful pass? If the receiver loses possession because the pass wasnt to the right place to enable clean receiving of the ball does the pass count?
Do the stats measure if the pass is to the right place for a receiver to receive the ball? Do the stats measure if the pass was too powerful to enable good control by the receiver?
Not sure if it's the be all end all, but I'd reckon expected threat (xT) is worth mentioning - if you have an Athletic subscription you can read about it here:
https://theathletic.com/2751525/202...ted-threat-or-xt-the-new-metric-on-the-block/
The idea basically is that the pitch is divided up into discrete zones:
Each zone shows the percent chance you have of scoring each time you're in possession - so when a team starts a possession at their own corner flag, there is a 0.1% chance of scoring a goal, but receiving possession right in front of the opposition goal gives a ~33% chance of scoring. The idea behind expected threat is that you can then look to see who is adding value by moving the ball from zones where you are unlikely to score to zone where you are more likely to score - and this can be broken up between those who add threat via dribbling or passing.
Here's passing xT from last year:
Reckon this holds some water, though it probably favours deep-lying midfielders playing for big sides on the whole.
https://theathletic.com/2751525/202...ted-threat-or-xt-the-new-metric-on-the-block/
The idea basically is that the pitch is divided up into discrete zones:
Each zone shows the percent chance you have of scoring each time you're in possession - so when a team starts a possession at their own corner flag, there is a 0.1% chance of scoring a goal, but receiving possession right in front of the opposition goal gives a ~33% chance of scoring. The idea behind expected threat is that you can then look to see who is adding value by moving the ball from zones where you are unlikely to score to zone where you are more likely to score - and this can be broken up between those who add threat via dribbling or passing.
Here's passing xT from last year:
Reckon this holds some water, though it probably favours deep-lying midfielders playing for big sides on the whole.
You can compare passing accuracy for the same position, but never do it for different positions though.
I agree you can't just use stats. My interest here was finding out which the best ones were.
I hadn't heard of expected threat. It looks useful and broadly corresponds to the players you think are threatening just with the eye test.
Even for the same position a player can be asked to do different things, so you do even need that context, too. For example a DM can be expected to work as a DLP who should advance the ball, or he can be expected to just calm the play and play it safe.
Good points there - distance of the passer from opposing player(s) when the pass is released might also be a good factor to consider.
Thing is the interpretation could vary - figures showing good passes far away from opposing players may indicate that a player is good at finding space, while figures showing successful passes with opposing players close could indicate the ability to play well under pressure.
Another interesting metric could be number of touches before a pass is released
Interesting idea.
Our open play xT from last year:
And our xT from the Leeds game:
Yeah I like the concept a lot because it can be segmented pretty cleanly between dribbling, passing, and receiving. Also it's pretty intuitive and for me lines up well generally by the eye test.Interesting idea.
Our open play xT from last year:
And our xT from the Leeds game:
Would we have to factor in the movement of your teammates at some point?
If they're not making intelligent runs/haven't been coached to do so/don't possess the knack or ability to find space as well as another team's players then it's obviously harder to pass to them.
Formations might be a thing to consider too. Not as many options for a lone front man to pass to compared to someone who has a partner. Same in other positions along the same lines.
If they're not making intelligent runs/haven't been coached to do so/don't possess the knack or ability to find space as well as another team's players then it's obviously harder to pass to them.
Formations might be a thing to consider too. Not as many options for a lone front man to pass to compared to someone who has a partner. Same in other positions along the same lines.
A variable measuring how contested the pass was, both in terms of pressure on passer and receiver
A bunch of those stats (all combined together) provide a pretty decent picture as well:
https://fbref.com/en/comps/9/passing/Premier-League-Stats
Progressive distance combined with progressive passes are an interesting ones for players who play in deeper areas, although you can only compare players in similar roles (obviously keepers & defenders would have an unfair advantage over midfielders, for example).
Progressive distance: Total distance, in yards, that completed passes have traveled towards the opponent's goal. Note: Passes away from opponent's goal are counted as zero progressive yards.
Progressive passes: Completed passes that move the ball towards the opponent's goal at least 10 yards from its furthest point in the last six passes, or any completed pass into the penalty area. Excludes passes from the defending 40% of the pitch
https://fbref.com/en/comps/9/passing/Premier-League-Stats
Progressive distance combined with progressive passes are an interesting ones for players who play in deeper areas, although you can only compare players in similar roles (obviously keepers & defenders would have an unfair advantage over midfielders, for example).
Progressive distance: Total distance, in yards, that completed passes have traveled towards the opponent's goal. Note: Passes away from opponent's goal are counted as zero progressive yards.
Progressive passes: Completed passes that move the ball towards the opponent's goal at least 10 yards from its furthest point in the last six passes, or any completed pass into the penalty area. Excludes passes from the defending 40% of the pitch
At the risk of being controversial, use your eyes.