xG Discussion

[Guest Cujek]

35 minutes ago, daventry_fox said:

I agree, Its just a fad. You don't need statistical data to know whether player X should have scored a certain chance or not

 

And I don't know about anybody else but I certainly don't need statistical data to tell me not to scream shooooot at Johnny Evans when he's 30 yards out with the ball

I was possibly a bit blunt.

 

but what annoys me is the Amerification of our sports, not everything can be boiled down to a stat.

 

 

[Foxxed]

1 minute ago, Cujek said:

I was possibly a bit blunt.

 

but what annoys me is the Amerification of our sports, not everything can be boiled down to a stat.

 

 

Ugh. Imagine it becomes like baseball, a sport devoid of entertainment whose fans wank themselves to death over a decimal number.

[AKCJ]

xG is bollocks for a few reasons.

 

1) It doesn't take the shot taker into account - Messi is going to gobble up the chance that Josh Low would miss.

 

2) It doesn't take the goalkeeper into account - Ter Stegen is going to save the shot that Chris Weale would let in.

 

3) Every penalty is given an xG of 0.76.

 

4) It completely ignores chances where a shot wasn't taken.

[Guest Cujek]

1 minute ago, Foxxed said:

Ugh. Imagine it becomes like baseball, a sport devoid of entertainment whose fans wank themselves to death over a decimal number.

Exactly.

 

or...

 

Maddison has carried the ball for 13 yards on average this match.

 

Awful.

[Guest]

48 minutes ago, daventry_fox said:

I always use Spain as the example their passing stats were through the roof both passes made and completed. Then when you actually watched them play you realise just how missleading those stats were. As most of them were 2 yard passes that actually lead to absolutely nothing or even lead to them making any progress up the pitch. 

You've inadvertently just explained one of the reasons why xG is useful. As you say, number of passes attempted and completed doesn't tell you how useful or incisive the passes were. A team can very easily play a lot of passes but have it result in nothing, like us under Puel. Similarly, just because you have a lot of shots or shots on target, doesn't actually mean you're creating a lot of good chances. If you lose a game having attempted 20 shots then it's easy to look at that and say "we battered them, look at all the shots we had", but if they're all Wilfred Ndidi 35-yarders into the stands or straight into the keeper's arms, it's not really a true reflection of how the game went. Nobody is saying it's a flawless metric or it's the most important thing ever invented in football, it's just fairly useful for assessing chance creation and conversion and it patently isn't just a useless waste of time or nerd fad.

[egg_fried_rice]

21 minutes ago, StriderHiryu said:

That said the xG stat specifically does confuse me. As @messerschmitt points out, Burnley had a higher xG in our game against them, but that didn't feel indicative of the game at all. It's not the first time I've seen such an occurrence, though at the same time it's been shown that over a course of the season it's quite reflective of a team's performance. 

 

16 minutes ago, messerschmitt said:

Yeah, thanks I see what your saying, it did make me think that if your XG was less than the opponent you didn't deserve to win the game or were lucky. Like you say XG doesn't  take account of intent to make the XG of a chance larger and losing it. In the first half a Burnley player had what I think was a shot, it was blocked and took at least 2 deflections off Leicester players before falling to Chris Wood who tried to slide it under Kasper. I guess this would count as two XGs in one attack, where as in the second half we had a good passing movement where eventually Ricardo ended up in the six yard box and tried a pass where we had two players at the far post who would have had a tap in but the pass was blocked so no XG.

@StriderHiryu It may not have felt indicative of the game as a whole because that's not what it measures - if you put all bias to one side and purely look at the quality of the chances where shots were taken, I would suggest Burnley's were superior.

 

@messerschmitt Yep - if you take the Wood chance which Kasper saved as well as his 'shot' which VAR ruled out - they would both have pretty high xG stats because of the position they occurred.

 

 

6 minutes ago, Foxxed said:

It's a load of old geezers going "Oooh I reckon they should've scored there" and "their attackers aren't getting in good positions" but using numbers instead isn't it.

Yes! Well sort of, but based on a consistent statistical model rather than subjective opinion.

 

 

7 minutes ago, Foxxed said:

Ugh. Imagine it becomes like baseball, a sport devoid of entertainment whose fans wank themselves to death over a decimal number.

 

7 minutes ago, AKCJ said:

xG is bollocks for a few reasons.

 

1) It doesn't take the shot taker into account - Messi is going to gobble up the chance that Josh Low would miss.

 

2) It doesn't take the goalkeeper into account - Ter Stegen is going to save the shot that Chris Weale would let in.

 

3) Every penalty is given an xG of 0.76.

 

4) It completely ignores chances where a shot wasn't taken.

3 minutes ago, Cujek said:

I was possibly a bit blunt.

 

but what annoys me is the Amerification of our sports, not everything can be boiled down to a stat.

 

 

@Foxxed @Cujek No one forces you to take any notice of these stats, nor is this a stat which claims to have boiled down the essence of football. When I'm at the game, the thought of xG never even enters my mind. It just provides interesting context after the fact and can be used as just one of a wide ranging group of metrics for more in-depth analysis on the team's performance

 

@AKCJ It's only bollocks if you expect it to measure those things. It's just a stat which measure one particular metric. The best players consistently outperform xG. It is in fact a great metric for separating the likes of Messi and Low - Messi scores over xG because he's better than average - Low scores less than xG because he's worse than average.

The best xG models do actually account for goalkeeper positioning and you could look at an expected saves metric (and how much a goalkeeper outperformed that) if you wanted to see how the quality of keeper would impact a shot.

Why shouldn't every penalty have the same xG? Given what I've stated above - it's the same quality of chance for everyone and the stat by itself does not claim to account for the quality of the keeper or other factors such as the occasion.

[egg_fried_rice]

I think the name is less than ideal. xG is just based on the historical average of that sort of chance. A better name might be 'historical average conversion rate'. The single players involved don't really matter compared to the overall because it is based on an average - obviously Low v de Gea is less likely to be scored than Messi v Weale. But add those in to the historical average and you can see why an xG amount is determined for a shot taken under any particular circumstance.

[AKCJ]

19 minutes ago, egg_fried_rice said:

 

@StriderHiryu It may not have felt indicative of the game as a whole because that's not what it measures - if you put all bias to one side and purely look at the quality of the chances where shots were taken, I would suggest Burnley's were superior.

 

@messerschmitt Yep - if you take the Wood chance which Kasper saved as well as his 'shot' which VAR ruled out - they would both have pretty high xG stats because of the position they occurred.

 

 

Yes! Well sort of, but based on a consistent statistical model rather than subjective opinion.

 

 

 

@Foxxed @Cujek No one forces you to take any notice of these stats, nor is this a stat which claims to have boiled down the essence of football. When I'm at the game, the thought of xG never even enters my mind. It just provides interesting context after the fact and can be used as just one of a wide ranging group of metrics for more in-depth analysis on the team's performance

 

@AKCJ It's only bollocks if you expect it to measure those things. It's just a stat which measure one particular metric. The best players consistently outperform xG. It is in fact a great metric for separating the likes of Messi and Low - Messi scores over xG because he's better than average - Low scores less than xG because he's worse than average.

The best xG models do actually account for goalkeeper positioning and you could look at an expected saves metric (and how much a goalkeeper outperformed that) if you wanted to see how the quality of keeper would impact a shot.

Why shouldn't every penalty have the same xG? Given what I've stated above - it's the same quality of chance for everyone and the stat by itself does not claim to account for the quality of the keeper or other factors such as the occasion.

Well yeah, but that's precisely what the media make out.

 

The media will give just the xG figure and make out that one team deserved to win more than the other. That's why it's bollocks.

 

If you take it as it is and look into xG and see who had chances and where then it's a valuable tool. But the media rarely, if ever, does this.

 

Sky and MOTD are terrible for it.

[egg_fried_rice]

Just now, AKCJ said:

Well yeah, but that's precisely what the media make out.

 

The media will give just the xG figure and make out that one team deserved to win more than the other. That's why it's bollocks.

 

If you take it as it is and look into xG and see who had chances and where then it's a valuable tool. But the media rarely, if ever, does this.

 

Sky and MOTD are terrible for it.

I don't necessarily agree that is what happens from all media outlets, but I can understand some consternation if that's how it comes across.

 

Media types and twitter types who do use it this way without any additional context are ill-informed. That's why stats will never be able to replicate watching the game. 

 

Thankfully, xG will never replace Goals as the One True Metric.

[KingsX]

42 minutes ago, AKCJ said:

xG is bollocks for a few reasons.

 

1) It doesn't take the shot taker into account - Messi is going to gobble up the chance that Josh Low would miss.

 

2) It doesn't take the goalkeeper into account - Ter Stegen is going to save the shot that Chris Weale would let in.

 

3) Every penalty is given an xG of 0.76.

 

4) It completely ignores chances where a shot wasn't taken.

 

xG models are proprietary, and always being tweaked.  I do know that in some models, your (1), (3) and (4) are not the case.  Probably the keeper bit too, though I'm not sure.

[Stadt]

Wait til you luddites here about expected possession value, golly gosh is it gonna fry your vacuous little brains

[StriderHiryu]

1 hour ago, egg_fried_rice said:

 

@StriderHiryu It may not have felt indicative of the game as a whole because that's not what it measures - if you put all bias to one side and purely look at the quality of the chances where shots were taken, I would suggest Burnley's were superior.

 

Thanks! This is actually the best and most succinct description of the statistic I have seen yet, and it makes it a lot more palatable to understand. I genuinely appreciate this response!

[worth_the_wait]

2 hours ago, BrummieFOX said:

So you genuinely don't think stats can tell us anything about a team, their performance, weaknesses/strengths? 

I'll tell when I realised all these silly stats are rubbish.

 

2015-16.  Title season.  

We had absolutely mullered Swansea 4-0 and Everton 3-1.   Got knows, how Everton managed to get a goal that day.   And if Vardy hadn't been demob happy, he'd have smashed in another penalty.

We dominated those games.  Totally.  From start to finish.  Over every square foot of that turf.

And our so-called possession stats were apparently 38% and 40%.

A meaningless statistic.

 

And that goes along with so-called "shots" and "shots on target".

Do headers count? Does the goalie have to make a proper save, or just let the ball softly roll into his arms?   Do blocked shots count?

Either way, some of the stats don't add up, no matter how to try to invent ways of counting them

 

And as for "expected goals", well you might as well summarise it at the end of the season as "expected league win", "expected relegation" or "expected FA Cup Quater Final placing".  

 

There's only one match stat that matters.   Ever.

 

[Les-TA-Jon]

I think the most useful thing about this stat is that it tells us how many luddites and dinosaur frequent this forum...

[filbertway]

Funny thing is it's a massive emphasis on stats and data analysis that has got us where we are. If you don't enjoy reading them or don't understand them then that's fine. Don't be silly and try and suggest they don't play an important part in football though, especially at a professional level.

 

Us aside, Brentford are a cracking example of how the use of stats and analysis can get you performing beyond your means.

[daventry_fox]

38 minutes ago, Les-TA-Jon said:

I think the most useful thing about this stat is that it tells us how many luddites and dinosaur frequent this forum...

Really? The most useful thing the stat told me, is people go to a game watch the game and have to wait until they see a bunch of data before they can work out who has or hasn't had a good game. 

 

I'm not saying they're useless as they play a pivotal role in those that decide teamsheets, training, systems/formations, player roles and the like.

Nutrition plays a pivotal role to yet none of us are discussing the players eating/drinking routines. 

 

Fans that rely on stats are basically admitting they don't really understand what they were watching on a Saturday

Edited 24 October 2019 by daventry_fox

[PAULCFC]

3 hours ago, Cujek said:

its a load of bollocks.

 

 

Could have not put it better myself!

[Stadt]

Its bollocks , what matters is the goal ratio at the end off the game !! 

[Guest]

17 minutes ago, daventry_fox said:

Fans that rely on stats are basically admitting they don't really understand what they were watching on a Saturday

Who are those fans? No fans are relying on expected goals to figure out what happened in the match or who had a good game, are they? Why does showing an interest in how good teams and players are at creating and finishing chances mean you somehow understand football less than someone who only cares what the final score is and assigns no importance to anything else that happens in the game?

 

This is going to have to be the last time I engage with this topic, I can feel it taking years off my life span

[Izzy]

2 hours ago, egg_fried_rice said:

A better name might be 'historical average conversion rate'. 

[Leeds Fox]

It is good for looking at the quality of chances, without taking into account who’s taking the chance, and who against. Which are massive factors.

 

The thing that gets me is that people are using it to project how a team will do over the course of a season, which is misleading. 

 

It’s a fad, it’s an easy article and most importantly... it favours the clubs who people want to read about (Man City, Liverpool etc). You could say it’s another way they can slate Man Utd too I guess. 

[Raw Dykes]

1 hour ago, worth_the_wait said:

I'll tell when I realised all these silly stats are rubbish.

 

2015-16.  Title season.  

We had absolutely mullered Swansea 4-0 and Everton 3-1.   Got knows, how Everton managed to get a goal that day.   And if Vardy hadn't been demob happy, he'd have smashed in another penalty.

We dominated those games.  Totally.  From start to finish.  Over every square foot of that turf.

And our so-called possession stats were apparently 38% and 40%.

A meaningless statistic.

 

And that goes along with so-called "shots" and "shots on target".

Do headers count? Does the goalie have to make a proper save, or just let the ball softly roll into his arms?   Do blocked shots count?

Either way, some of the stats don't add up, no matter how to try to invent ways of counting them

 

And as for "expected goals", well you might as well summarise it at the end of the season as "expected league win", "expected relegation" or "expected FA Cup Quater Final placing".  

 

There's only one match stat that matters.   Ever.

 

This is puzzling. I don't know what you expect stats to tell you. We all know the team with the most possession isn't necessarily the better team. We won the league without having much of the ball. Are you saying those stats were incorrect?

 

I've got no idea what the bit in bold is supposed to mean.

[Leeds Fox]

46 minutes ago, daventry_fox said:

Really? The most useful thing the stat told me, is people go to a game watch the game and have to wait until they see a bunch of data before they can work out who has or hasn't had a good game. 

 

I'm not saying they're useless as they play a pivotal role in those that decide teamsheets, training, systems/formations, player roles and the like.

Nutrition plays a pivotal role to yet none of us are discussing the players eating/drinking routines. 

 

Fans that rely on stats are basically admitting they don't really understand what they were watching on a Saturday

 

Yes we are. All players should drink port, chew tobacco and eat pizza pre-match. No wonder we’re 8pts off of Liverpool already. 

[egg_fried_rice]

6 minutes ago, Izzy said:

 doesn't exactly roll off the tongue like xG does it

[Dahnsouff]

1 hour ago, Leeds Fox said:

 

Yes we are. All players should drink port, chew tobacco and eat pizza pre-match. No wonder we’re 8pts off of Liverpool already. 

Mmmmmmmm port