Should analysts be working on training, not only match, data.
We've known, for a long time, that sample size is key to the accuracy of judging a player with data. The more minutes a player has played the more accurate our outcomes will be.
Even then we need to be careful as a player with 900 minutes may have played 10 full games, or appeared as a substitute in 38 games, with the game state determining a lot of the type of actions they will be performing.
And we also always have to take into account the style of play and system the player operates within. A wide attacker's output will be very different in a Warnock team to a Guardiola team.
And also we have the problem that a team plays differently against each opponent. A full back will profile differently if only measured against the top 6 clubs than the rest. Actually, I'm not sure on that point, might be worth looking at.
Oh, and the other problem is we only have data for those who play. So how can I judge the other 45 members of playing staff when I am only getting data for about 15 of the players?
The Yips
One of the oft-repeated stories in football is of the player who is amazing in training but crumbles under the pressure of a big match. I can believe this is possible, but I also wonder if it could just be a post-hoc rationalisation for poor performance. It is entirely possible to lose the flip of a coin 5 times in a row. It is entirely possible for a decent striker to miss 5x xG0.5 chances in a row and just be unfortunate.
So how do we know when it is "the yips" and when it is just bad luck? And can players regain battered confidence?
Some of the cliches we think of are "he just needs one to go in off his backside" and "you only start to worry when you don't get chances, not when you are missing them". Both are actually based in truth.
One of the few advantages of getting a bit older is you can remember the early careers of some of the greats. Alan Shearer was at the point of being dropped by England before Euro 96 after a barren scoring run. He was under huge pressure before the Switzerland opening game and went on to be the top scorer. One goal was enough to boost confidence? Or was it reversion? Who knows?
How transferable are training skills and game skills?
Given all these issues it makes me wonder just how transferable training performances are to games?
If I set up a lot of finishing drills, say 100 "life like" chances for shots at goal from various angles, at Manchester United would we expect Lukaku, Rashford, and the other forwards to be the best, and perhaps the central defenders the worst?
I would.
Would we also expect there to be some significant variation between forwards?
Again I would. I've only seen bits and pieces of professional finishing drills but I would suspect someone like Lukaku would be better than someone like Crouch at picking up the ball on the edge of the area, beating a player and slotting it in the bottom corner.
So there must be some correlation between training performance and matchday performance.
Given that then should we actually be tagging, analysing and recording training footage rather than matchday footage? Sample sizes will be much larger. We can create situations we might not get in matchdays. And we can record data for all our playing squad, not just those who play. If my 18-year-old reserve striker is finishing chances at the best rate at the club I would want it flagged up.
I believe I saw a TED talk on Basketball once which implied they put huge sway on training performance. The player consistently hitting 3 pointers in the gym tended to be the same ones doing it on the court.
As always I suspect some clubs already do this.
But I'd love to see a study that compared training drill skill with on-field performance. Let me know if you find any.
Even then we need to be careful as a player with 900 minutes may have played 10 full games, or appeared as a substitute in 38 games, with the game state determining a lot of the type of actions they will be performing.
And we also always have to take into account the style of play and system the player operates within. A wide attacker's output will be very different in a Warnock team to a Guardiola team.
And also we have the problem that a team plays differently against each opponent. A full back will profile differently if only measured against the top 6 clubs than the rest. Actually, I'm not sure on that point, might be worth looking at.
Oh, and the other problem is we only have data for those who play. So how can I judge the other 45 members of playing staff when I am only getting data for about 15 of the players?
The Yips
One of the oft-repeated stories in football is of the player who is amazing in training but crumbles under the pressure of a big match. I can believe this is possible, but I also wonder if it could just be a post-hoc rationalisation for poor performance. It is entirely possible to lose the flip of a coin 5 times in a row. It is entirely possible for a decent striker to miss 5x xG0.5 chances in a row and just be unfortunate.
So how do we know when it is "the yips" and when it is just bad luck? And can players regain battered confidence?
Some of the cliches we think of are "he just needs one to go in off his backside" and "you only start to worry when you don't get chances, not when you are missing them". Both are actually based in truth.
One of the few advantages of getting a bit older is you can remember the early careers of some of the greats. Alan Shearer was at the point of being dropped by England before Euro 96 after a barren scoring run. He was under huge pressure before the Switzerland opening game and went on to be the top scorer. One goal was enough to boost confidence? Or was it reversion? Who knows?
How transferable are training skills and game skills?
Given all these issues it makes me wonder just how transferable training performances are to games?
If I set up a lot of finishing drills, say 100 "life like" chances for shots at goal from various angles, at Manchester United would we expect Lukaku, Rashford, and the other forwards to be the best, and perhaps the central defenders the worst?
I would.
Would we also expect there to be some significant variation between forwards?
Again I would. I've only seen bits and pieces of professional finishing drills but I would suspect someone like Lukaku would be better than someone like Crouch at picking up the ball on the edge of the area, beating a player and slotting it in the bottom corner.
So there must be some correlation between training performance and matchday performance.
Given that then should we actually be tagging, analysing and recording training footage rather than matchday footage? Sample sizes will be much larger. We can create situations we might not get in matchdays. And we can record data for all our playing squad, not just those who play. If my 18-year-old reserve striker is finishing chances at the best rate at the club I would want it flagged up.
I believe I saw a TED talk on Basketball once which implied they put huge sway on training performance. The player consistently hitting 3 pointers in the gym tended to be the same ones doing it on the court.
As always I suspect some clubs already do this.
But I'd love to see a study that compared training drill skill with on-field performance. Let me know if you find any.
Comments
Post a Comment