Possession = Success?
- Thread starter Frogfish
- Start date
- Latest activity Latest activity:
- Status
Someone also needs to tell Rodgers that it's not all about possession !
Just 6% greater chance of winning with the possession stat in your favour.
Too simplistic and misleading - which is true about almost all use of statistics in football ever.
Once a team goes ahead it tends to concede possession.
Someone also needs to tell Rodgers that it's not all about possession !
Just 6% greater chance of winning with the possession stat in your favour.
1. That's a bloody confusing chart to read.
2. So let me see if I have this right - the picture says:
Possession Win% Draw% Lose%
More 41 24 35
Less 35 24 41
So, okay, 6% doesn't look like a lot on the face of it, but if you look at the expected rewards (values):
Possession Expected_Value (i.e. points)
More = .41x3 + .24x1 + .35x0 = 1.47
Less = .35x3 + .24x1 + .41x0 = 1.05
So, E(More) - E(Less) = .42, i.e. if you have more possession in every game, you can expect to win .42 points more per game. That is 15.96 points over a 38-game season. Not sure if I'm remembering my probabilities lessons and/or if I'm applying the stuff correctly.
Someone also needs to tell Rodgers that it's not all about possession !
Just 6% greater chance of winning with the possession stat in your favour.
Rodgers used to say 79% of the time the team with more possession won.
And the possession myth is something I've been as arguing about for a couple of years at this stage.
I'm not sure I'd agree with your interpretation either. The better team usually wins the game and the better team will usually have more possession.
I don't think the second leads to the first.
Yes. Mathematically you're correct. Assuming the original percentages are correct and from a reliable source.
The team that takes the lead tends to concede (some) possession. You would get a more accurate picture of the importance of possession if you looked at which team dominated possession at 0-0 and correlated with the end result.
Good players play more accurate passes and will tend to be in better positions to receive a pass. Good teams tend to consist of good players. Greater possession can be seen as simply a symptom of being "good at football". Deliberately setting your team up to achieve greater possession is not going to make the players better, and might be counter-productive.
With all statistics in football context is everything. However, if you run a newspaper ignoring context is a great way to generate clickbait.
Good players play more accurate passes and will tend to be in better positions to receive a pass. Good teams tend to consist of good players. Greater possession can be seen as simply a symptom of being "good at football". Deliberately setting your team up to achieve greater possession is not going to make the players better, and might be counter-productive.
With all statistics in football context is everything. However, if you run a newspaper ignoring context is a great way to generate clickbait.
It's never an element that can be considered in isolation. While not wanting to sound like a 'all stats are crap' advocate, it's all over-complicated. Possession is a good thing if you don't want to lose. Possession is a great thing if you want to win and know how to use it to attack and score goals. The irritating thing about the use of possession stats is that on its own it doesn't tell you much at all. Speed of passing from one player to the next while in possession is more revealing, as is movement in different areas of the pitch. Last season and in pre-season this time, the team uses possession much like the great teams of the late 70s and 80s did after getting safely ahead of the opposition: lots of backward passes, at a slowish tempo, with few attempts at killer through balls or crosses. The problem is we play like that when the scores are level. Rodgers' sides only really thrive when players release the ball very, very quickly and the tempo is ultra-high.
I don't even know what sample sizes they used, so it's just all a mental maths wank.
Bang on.
Strange thinking - 10 superstar defenders is never really gonna win many games.
Fucking madness. It would never work.
- Status
