Do you think stats ruin the game?

B_G said:
Does xG take into account the player that has the chance?

I.e. if Haaland or Lee Dong Gook had the same chance in the same place would the xG be the same or is the the xG higher in the scenario that Haaland has the chance because he is much more likely to finish?
Click to expand...

No.

What it will show is that Haaland will score more than the xG of his chances suggest he should. So it'll demonstrate his superior finishing. The xG won't change dependent on player though.

I assume xG itself is worked out based on how many times similar chances have been put away across all matches, by all players. So kind of an average. (I think).
 
  • Like
Reactions: B_G
B_G said:
Does xG take into account the player that has the chance?

I.e. if Haaland or Lee Dong Gook had the same chance in the same place would the xG be the same or is the the xG higher in the scenario that Haaland has the chance because he is much more likely to finish?
Click to expand...
No it doesn't. One of the weaknessess in the xG stat. Only 1 provider, Opta, which is also an issue.
 
  • Like
Reactions: B_G
festa5 said:
No.

What it will show is that Haaland will score more than the xG of his chances suggest he should. So it'll demonstrate his superior finishing. The xG won't change dependent on player though.

I assume xG itself is worked out based on how many times similar chances have been put away across all matches, by all players. So kind of an average. (I think).
Click to expand...
You are correct Festa in that Opta analysed hundreds of thousands of chances before compiling xG version 1.0. It was still awful though.
 
festa5 said:
No.

What it will show is that Haaland will score more than the xG of his chances suggest he should. So it'll demonstrate his superior finishing. The xG won't change dependent on player though.

I assume xG itself is worked out based on how many times similar chances have been put away across all matches, by all players. So kind of an average. (I think).
Click to expand...
So xG is really only as good as your finishing then? And that is where the problem lies when quoting xG isn't it?

When you have players like Akpom, Archer, Ramsey and even Forss playing, their finishing was superb for most of the season and I'm guessing we will have scored above our xG through the season? With the players we've replaced them with, we just don't look to have that finishing ability.
 
'Assists' is one bit of Americana that we didn't really need. As someone pointed out here recently, Chuba A would typically start a move and then pop up to score off another player's assist. I remember Bernie Slaven doing similar.
 
Personally I largely ignore them. The xG nonsense in particular relies too much on interpretation to be a meaningful stat. Is this chance like that chance? Well it's slightly wide, a different keeper, a different ground, etc.

Might be some use to coaches in a macro sense (i.e. looking at performance over a whole or part season) but when you look at a single incident or even match?

Assists. Is a hopeful punt upfield that results in a goal an "assist"? A deflected pass?

They seem to appeal to a certain kind of fan. Whatever rocks your boat but I can happily ignore them by and large.
 
B_G said:
So xG is really only as good as your finishing then? And that is where the problem lies when quoting xG isn't it?

When you have players like Akpom, Archer, Ramsey and even Forss playing, their finishing was superb for most of the season and I'm guessing we will have scored above our xG through the season? With the players we've replaced them with, we just don't look to have that finishing ability.
Click to expand...
I think part of problem is people not really understanding what it's trying to show.

xG is supposed to be a measure of how many, and how good the chances are that you create (and concede). But finishing is obviously a big part of the game.

So yeah, you're right. I'm pretty sure once Carrick took over last year we scored above our xG rate. What it shows this year is that we're missing too many chances. On the goals scored front at least. That doesn't mean we're unlucky, it means we aren't taking chances.

On the goals conceded side, you can probably argue a bit more that there's an element of "being unlucky" involved if you're conceding more than xG suggests you should. But there are so many factors involved it's hardly an exact science.

As laughing implies, it's a far from perfect tool. But, it does give an indication at least of how many chances you create/concede and how good they are. Although you don't really need it to know our finishing has been abysmal this year!
 
Stats have a place, they paint a picture but don’t tell the whole story. How reliable is the data? How good is the person receiving it in assessing its importance to the wider picture? How key is each stat and how do they relate to improvement or lead you down blind alleys? I don’t need an xG for Lath to know whether he should have scored more than he has and why he hasn’t, my eyes tell me.
 
Does xG and statistics take into account skill and technique of a player? An extremely skillful player could have **** stats, or a player with great stats could be a really **** player??
 
Stats allow you to measure reality rather than perception, to remove inherent biases. They're a great tool for coaching and analysing the game. If that isn't your bag, then no one is forced to look at stats, and you can carry on with less knowledge. Personally I love learning, love understanding and analysing, it's the way I'm wired but that isn't for everyone.
 
I think the ratio on penalties is something like expected to score 3:4 so penalties will somewhat distort xG surely? I presume they will have stats with and without penalties though. xG highlights issues good and bad, but you need your eyes, brain to interpret its relevance, reflect on before judging its relevance. For example, I would expect it easier to beat Lumley in an identical one on one more often than I would say Ederson.
 
People used to use Shots and Shots on Target as the stats to show whether a team had a good game or not but that didn't take into account anything to do with how likely it was to score. xG is just a better version of that but still not perfect. One of the biggest issues is it doesn't count chances unless the shot is actually taken. If a player gets 1 on 1 with a keeper and has a very good chance to score but tries to take it round them and is tackled then there is no shot so there is no xG.

Assists is big because people play fantasy football. It's not surprising though that some of the best passers/chance creators are the best players so it does tend to correlate. De Bruyne is always up there on the assists charts for example even though he's in a team of other players that can also create chances but he does it better.

Data is an indicator. It needs knowledge and skill to collect it, analyse it and to understand/use it. The same person doesn't need to be the one doing all those things. Good managers/coaches/recruiters will know which stats are important. Barlaser was a top assister for Rotherham but that stat on it's own tells you very little. Watch all his assists and you see they were mostly from free kicks and corners and that can be down to the person attacking the ball as much as the person delivering a cross. Put him into a team like ours with less height and nobody any good at scoring from corners and his assists drop. I hate watching those penalty takers with the funny hops and jumps and I prefer an old school put your laces through it sort of penalty taker but the stats will tell you which one results in more goals. Feelings can be wrong and stats can tell you the real story better at times.
 
BoroMart said:
Wouldn't make them right and they'd be less effective in the modern game for it.
Click to expand...
I dare say they'd probably fail.

But if they were managing in the modern game they'd probably use them. They might scoff openly at some of them, but they were all winners who would seek any advantage to be successful.

Ridiculous to compare eras anyway. The information simply didn't exist when they were managing.

You'd be giving an advantage away to your rivals by not using them. It's fine margins in elite sport.
 
festa5 said:
So yeah, you're right. I'm pretty sure once Carrick took over last year we scored above our xG rate. What it shows this year is that we're missing too many chances. On the goals scored front at least. That doesn't mean we're unlucky, it means we aren't taking chances.

On the goals conceded side, you can probably argue a bit more that there's an element of "being unlucky" involved if you're conceding more than xG suggests you should. But there are so many factors involved it's hardly an exact science.
Click to expand...
This all makes sense and basically they way I thought it worked. I see a few people quoting xG to say we are playing well and should be winning games but that argument surely can only be used short term. After then, it is exposing that we are too toothless in front of goal.
 
davidmills said:
Wouldn't have thought so a great manager would be a great manager in any era
Click to expand...
This maybe your opinion but I disagree. Brian Clough would not have a clue how to manage todays players. They would down tools on him.

I dare say Alex Ferguson would struggle.

The way players need to managed and the style of the football being played has changed massively. I don't think Brian Clough could adapt to the man management side.
 
The modern world is dominated by data collection/analysis and with AI that is only going to increase. Companies like Amazon have gone from nothing to a global giant because they know how to harvest data and use it to their benefit.

I've no doubt you can be a successful football manager/club without relying on data analysis, but if that is your choice, you will be at a disadvantage to those that do exploit the data.
 
You must log in or register to reply here.
Back
Top