Form Table according to xG

All advertisments are hidden for logged in members, why not log in/register?

beatific blade

beatific blade

Active Member
Joined
Jan 4, 2014
Messages
1,073
Reaction score
2,321
Location
S10
Form table below according to xG - data from Understat.com - last 6 match weeks.

Note our two defeats have come against two of the three teams at the top of this table. Also note that we have won the xG battle in 5 of those 6 games - i.e we've created quantifiably better chances and restricted the opposition to enable us to win games.

We are absolutely fine. Nothing to worry about. If anything, we've got better.

Beans Roygbiv

1578308913249.png

UTB FTO FVAR
 

Only the bedwetters are predicting doom and gloom after losses away to Man City, Liverpool and a shit showing against AFC Fylde with our second string.

One thing the xG table tells me though is that we need someone who can put the ball in the net but that's quite obvious really.
 
derekacorah said:
Only the bedwetters are predicting doom and gloom after losses away to Man City, Liverpool and a shit showing against AFC Fylde with our second string.

One thing the xG table tells me though is that we need someone who can put the ball in the net but that's quite obvious really.
Click to expand...
No ones predicting doom and gloom ,so why make things up ?
 
beatific blade said:
Form table below according to xG - data from Understat.com - last 6 match weeks.

Note our two defeats have come against two of the three teams at the top of this table. Also note that we have won the xG battle in 5 of those 6 games - i.e we've created quantifiably better chances and restricted the opposition to enable us to win games.

We are absolutely fine. Nothing to worry about. If anything, we've got better.

Beans Roygbiv

View attachment 67994

UTB FTO FVAR
Click to expand...


Brilliant.
 
I still don't fully understand why so much weight is put into XG stats.

From what I know, every opportunity on goal is allocated an XG value of between 0.0 and 1.0. These are then tallied to create an overall XG value for the match (2.3 v 1.2 for example).

What gets me though is that what if you have a useless team that does nothing but have 25 yard pops at goal all game that get nowhere near goal? Let's say they have 20 of these 0.1 efforts, that would give them an overall XG for the match of 2.0. And what if the better opposition creates a smaller number of better quality chances but ends up with an overall XG of 1.9.

On paper does the XG of 2.0 always trump the 1.9? Because any team no matter how dire can have long range punts at goal. But the XG stat would be very misleading in this scenario.

Hope someone can shed light on this for me cos it's doing my nut in.
 
guancheblade said:
I still don't fully understand why so much weight is put into XG stats.

From what I know, every opportunity on goal is allocated an XG value of between 0.0 and 1.0. These are then tallied to create an overall XG value for the match (2.3 v 1.2 for example).

What gets me though is that what if you have a useless team that does nothing but have 25 yard pops at goal all game that get nowhere near goal? Let's say they have 20 of these 0.1 efforts, that would give them an overall XG for the match of 2.0. And what if the better opposition creates a smaller number of better quality chances but ends up with an overall XG of 1.9.

On paper does the XG of 2.0 always trump the 1.9? Because any team no matter how dire can have long range punts at goal. But the XG stat would be very misleading in this scenario.

Hope someone can shed light on this for me cos it's doing my nut in.
Click to expand...

The xG part already answers your question. The key is making sure that the 25 yard pop is assigned the right xG value. In your example, it has a score of 0.1. That suggests that one in every ten of those ‘pops’ results in a goal. If that’s ‘true’, then twenty of those chances in a game would be quite good. In reality, I expect that a 25 pop would probably only have a value of 0.03 or so (depending on the angle), so you’d need about thirty, just to get 1 xG.

I think you’re underestimating the complexity that goes into assigning the xG scores for each chance. They should reflect the real likelihood of a goal.
 
Geordie Blade said:
The xG part already answers your question. The key is making sure that the 25 yard pop is assigned the right xG value. In your example, it has a score of 0.1. That suggests that one in every ten of those ‘pops’ results in a goal. If that’s ‘true’, then twenty of those chances in a game would be quite good. In reality, I expect that a 25 pop would probably only have a value of 0.03 or so (depending on the angle). I think you’re underestimating the complexity that goes into assigning the xG scores for each chance.
Click to expand...

I'm probably playing Devil's Advocate more than anything... that the worst players/teams can make terrible attempts on goal but have enough of them they could lead to a misleading stat (i.e. that they deserved at least a goal despite their play not being up to that standard).

Additionally, you could have a team that creates loads of guilt edged, high xG chances but have a striker who can't hit a barn door with a fist (sound familiar?)... xG interpretation would say that team is showing all the signs of performing well. But if they persist with the wasteful striker, then xG indicator doesn't really prove anything.

In fact, if xG could factor in the player's previous xG for that kind of xG chance, that would be a more nuanced statistic.

For example, a one on one with a regular 0.7 xG. If this chance falls to Sharp, the xG then goes up to 0.9. If the chance falls to McGoldrick, xG falls to 0.2 😝

Know what I mean?
 
guancheblade said:
I'm probably playing Devil's Advocate more than anything... that the worst players/teams can make terrible attempts on goal but have enough of them they could lead to a misleading stat (i.e. that they deserved at least a goal despite their play not being up to that standard).

Additionally, you could have a team that creates loads of guilt edged, high xG chances but have a striker who can't hit a barn door with a fist (sound familiar?)... xG interpretation would say that team is showing all the signs of performing well. But if they persist with the wasteful striker, then xG indicator doesn't really prove anything.

In fact, if xG could factor in the player's previous xG for that kind of xG chance, that would be a more nuanced statistic.

For example, a one on one with a regular 0.7 xG. If this chance falls to Sharp, the xG then goes up to 0.9. If the chance falls to McGoldrick, xG falls to 0.2 😝

Know what I mean?
Click to expand...

But in order to get into a position where the attempts on goal are terrible they'd need to create the opportunity. Even if it's 50 shots from 25 yards adding up to 2xG it's impressive to have been in a position to take 50 shots from 25 yards. The teams analysts presumably would then recommend buying a player who scores a lot of goals from 25 yards out (if they exist) and the gap between goals and xG should reduce.

If they can't buy this player then they need to look at doing something differently in their style of play.
 
I find in most cases it supports what you see in a game. You get the odd one like Spurs v Bayern but even that kind of backed up the feeling that everything Bayern hit went in. I recall the pigs had a spell last season where they were scoring a lot of long shots (from Reach mainly) and the Xg stats showed it was unlikely to be sustainable, that their results would fall away. So it seems to be a reasonable stat to use. It’s never going to be perfect because football isn’t really a science.
 
If McGoldrick's personal xg is available for tbhe season it would speak volumes to guancheblade .

Our man has had over 40 shots at goal and missed plenty of what the P.L. calls ,' big chances missed' , somthing like 18 I think.

Each of those is given an xG rating and a few will be 1.0 or just less. Others will be anywhere from 0.1 upwards.
 
guancheblade said:
I still don't fully understand why so much weight is put into XG stats.

From what I know, every opportunity on goal is allocated an XG value of between 0.0 and 1.0. These are then tallied to create an overall XG value for the match (2.3 v 1.2 for example).

What gets me though is that what if you have a useless team that does nothing but have 25 yard pops at goal all game that get nowhere near goal? Let's say they have 20 of these 0.1 efforts, that would give them an overall XG for the match of 2.0. And what if the better opposition creates a smaller number of better quality chances but ends up with an overall XG of 1.9.

On paper does the XG of 2.0 always trump the 1.9? Because any team no matter how dire can have long range punts at goal. But the XG stat would be very misleading in this scenario.

Hope someone can shed light on this for me cos it's doing my nut in.
Click to expand...

If each individual shot has a 0.1 chance of going in then you'd expect a goal on average every 10 shots.

That's not misleading, it's basic probability exactly the same way as saying that rolling a six on a fair dice happens about every six rolls. Now, probabilities don't work very neatly in practice so you'd see things like someone scoring three 0.1's in a row where another striker might have missed 20 in a row, but over a large enough sample size it'll be about right.

The problem is with your assumptions. Shots from 25yds go in far less frequently than 0.1. And secondly, can you imagine a team allowing someone to have twenty shots from the edge of the box in a game?

The maths is essentially proven. You can demonstrate it with coin flips, dice rolls, or any other repeatable chance event. The only problem that XG has is the samples used to determine the value given to any given chance because, unfortunately, that has to come down to a set of selection criteria that will either be subjective, missing, or very hard to factor in. Minor things like distance from goal and the angle to the goal are easy to do, but you could come up with all sorts of other criteria that will have an effect like weather conditions, pitch quality, pressure from defenders, lighting, and so on. All those things will have some impact on the "true" probability of a shot going in but may be impractical to take account of due to either measurement or sample size issues.

So it's not a question of "can we model shots on goal mathematically?". It's a question of how well we can model it. And XG seems to be a pretty good predictor of results.
 

PokerBlade said:
The problem is with your assumptions. Shots from 25yds go in far less frequently than 0.1. And secondly, can you imagine a team allowing someone to have twenty shots from the edge of the box in a game?

The maths is essentially proven. You can demonstrate it with coin flips, dice rolls, or any other repeatable chance event. The only problem that XG has is the samples used to determine the value given to any given chance because, unfortunately, that has to come down to a set of selection criteria that will either be subjective, missing, or very hard to factor in. Minor things like distance from goal and the angle to the goal are easy to do, but you could come up with all sorts of other criteria that will have an effect like weather conditions, pitch quality, pressure from defenders, lighting, and so on. All those things will have some impact on the "true" probability of a shot going in but may be impractical to take account of due to either measurement or sample size issues.

So it's not a question of "can we model shots on goal mathematically?". It's a question of how well we can model it. And XG seems to be a pretty good predictor of results.
Click to expand...

We have a bloke in our 5-a-side team who always has a pop from the half way line despite being no good at it. Does the fact that he's no good at long rangers factor into the xG value for those attempts? Or is it the average based on a pool of players?

Additionally, does my 5-a-side team become much better as a result of our total xG in a game being much higher than it would have been without him? Cos believe me our team is useless!
 
if we beat west ham and go 5th on friday on 32 points we will only need 2 wins to match our last prem total

so cant see us do anything but finish safe
the only worry is how high we finish as each place up means more dosh
last year 8th got 25m 5th got 29 m
 
Last edited:
guancheblade said:
We have a bloke in our 5-a-side team who always has a pop from the half way line despite being no good at it. Does the fact that he's no good at long rangers factor into the xG value for those attempts? Or is it the average based on a pool of players?

Additionally, does my 5-a-side team become much better as a result of our total xG in a game being much higher than it would have been without him? Cos believe me our team is useless!
Click to expand...

If the 5 a side league you play in collects stats on xG they are ground breaking.
 
guancheblade said:
We have a bloke in our 5-a-side team who always has a pop from the half way line despite being no good at it. Does the fact that he's no good at long rangers factor into the xG value for those attempts? Or is it the average based on a pool of players?

Additionally, does my 5-a-side team become much better as a result of our total xG in a game being much higher than it would have been without him? Cos believe me our team is useless!
Click to expand...

Let's say I were coming up with a sample from 5-a-side games to create an XG stat for them (god knows why I would). My sample of goals scored from that position would include the guy on your team and the XG for that shot would be an average of every player I could get data for.

The answer to the second question is more complicated. The simple way of looking at it is this: if that guy shoots from there every time then he'll score more goals. The more complicated way is that you have to factor in all the much better chances that you now won't have because he keeps pinging the ball back to the opposition. And the opposition, by virtue of getting the ball more, will increase their XG against you.

And the league above reflects that final part - not just one side's XG but the XG compared to the oppositon. Because you could be a team that scores a goal every game but let's three in.

Mostly I think you have to keep in mind that shots from range don't go in very often. You'd need a LOT of shots before you could reliably score goals that way. I remember last season seeing that the teams that shot most often from range were Fulham and Cardiff. Teams that work the ball into high scoring opportunities do better than those who shoot more from range because you simply can't get enough shots away to compensate for the low probability of scoring.
 
Pedro Verde said:
Xg doesnt win you games
Click to expand...

No...it's just a very good indicator of what your future results are likely to be.

It's like when you go into a casino and hit a number on the roulette wheel. The casino doesn't make money by paying out bets. It makes it's money because the long term odds are well in its favour.
 
guancheblade said:
In fact, if xG could factor in the player's previous xG for that kind of xG chance, that would be a more nuanced statistic.
Click to expand...
If you're interested there's some discussion of why that couldn't work yet in this Tifo football podcast: https://open.spotify.com/episode/0pd4pmVMExgfAAZB1sTWEz?si=UjQcbFrnQdaJSXgu-EpE9w

TLDR basically is that there isn't enough data to build a model that understands the difference between a McGoldrick shot and a Sharp shot. To have a full understanding we'd need to have lots of data on McGoldrick taking those shots hundreds if not thousands of times, with the same for every striker in the league
 
Nesh Blade said:
If you're interested there's some discussion of why that couldn't work yet in this Tifo football podcast: https://open.spotify.com/episode/0pd4pmVMExgfAAZB1sTWEz?si=UjQcbFrnQdaJSXgu-EpE9w

TLDR basically is that there isn't enough data to build a model that understands the difference between a McGoldrick shot and a Sharp shot. To have a full understanding we'd need to have lots of data on McGoldrick taking those shots hundreds if not thousands of times, with the same for every striker in the league
Click to expand...

As a rough comparison, when I was playing poker and tracking my hands the general wisdom was that to get a rough idea of a player's winrate in NLH cash games, you'd want around 30-40,000 hands played. To get a decent idea you'd be looking at the 100k area. To get close to your "true" winrate you might want close to a million. And then the problem is that by the time you've played a million hands of cash poker, you've inevitably learned more, altered your game, changed stakes up or down, and your sample is no longer a decent reflection.

But the other extreme is also true. Sometimes you can see a dozen hands someone's played and immediately pick up on a ton of things they could do better.

Statistical modelling like this is extremely useful, just don't lose sight of the woods for the trees or start seeing the stat in front of you as some gospel truth.
 
PokerBlade said:
As a rough comparison, when I was playing poker and tracking my hands the general wisdom was that to get a rough idea of a player's winrate in NLH cash games, you'd want around 30-40,000 hands played. To get a decent idea you'd be looking at the 100k area. To get close to your "true" winrate you might want close to a million. And then the problem is that by the time you've played a million hands of cash poker, you've inevitably learned more, altered your game, changed stakes up or down, and your sample is no longer a decent reflection.

But the other extreme is also true. Sometimes you can see a dozen hands someone's played and immediately pick up on a ton of things they could do better.

Statistical modelling like this is extremely useful, just don't lose sight of the woods for the trees or start seeing the stat in front of you as some gospel truth.
Click to expand...
Or, when you watch the poker on TV they give mathematical percentages for the likelihood of winning a hand. They’re purely statistical and don’t account for how the players will play their cards but they’re still useful. I wouldn’t necessarily know if an ace/king is better than a pair of sixes but the percentage tells me immediately.
 
BushBlade said:
Or, when you watch the poker on TV they give mathematical percentages for the likelihood of winning a hand. They’re purely statistical and don’t account for how the players will play their cards but they’re still useful. I wouldn’t necessarily know if an ace/king is better than a pair of sixes but the percentage tells me immediately.
Click to expand...

The main thing when you play poker is you realise how hard it is to actually visualise what odds look like playing out. You go through periods where you can't win a hand, times where every draw you've got lands, and times where things seem perfectly normal. And you get people who are sure they're plain unlucky, or that they have a "lucky hand" they always go with, and people who are certain beyond any doubt that the unlikely events they've witnessed mean that online poker is rigged to the core. Funnily enough, when you track a decent sample of hands everything works out to what the maths says it should given a standard deviation or two.

People get hung up on short term results when they're actually fairly meaningless. Football's the same as poker in this regard. You can play poorly and win a game, you can play well and lose a game. But if you play poorly every week then that lucky win won't give you much comfort come the end of the season.
 
PokerBlade said:
The main thing when you play poker is you realise how hard it is to actually visualise what odds look like playing out. You go through periods where you can't win a hand, times where every draw you've got lands, and times where things seem perfectly normal. And you get people who are sure they're plain unlucky, or that they have a "lucky hand" they always go with, and people who are certain beyond any doubt that the unlikely events they've witnessed mean that online poker is rigged to the core. Funnily enough, when you track a decent sample of hands everything works out to what the maths says it should given a standard deviation or two.

People get hung up on short term results when they're actually fairly meaningless. Football's the same as poker in this regard. You can play poorly and win a game, you can play well and lose a game. But if you play poorly every week then that lucky win won't give you much comfort come the end of the season.
Click to expand...
Which again comes down to sample size - the bigger the sample the more likely it is that it will be ‘accurate’.

Stats are simple if you get it, difficult if you don’t. In fact, I’d say 25% of people understand statistical analysis and the other 80% are idiots.
 

BushBlade said:
Which again comes down to sample size - the bigger the sample the more likely it is that it will be ‘accurate’.

Stats are simple if you get it, difficult if you don’t. In fact, I’d say 25% of people understand statistical analysis and the other 80% are idiots.
Click to expand...

I think even when you understand the concept we're just very bad at visualising what that actually looks like. After the early ipods they made the shuffle feature intentionally less random because people complained that it did things like playing the same track twice in a row, playing multiple tracks off a single album, stuff that they thought shouldn't happen, but stuff that happens a surprising amount with a random selection.

Another one that gets people is take these two numbers:
3981462705
9231957774

One of them I got from a random generator on google, the other is one I put together intentionally. When they do this test most people will say the first number is the random one. But it isn't. The first one is much less likely to occur by chance because it contains every digit from 0-9. And when people try to make up a random number they usually do what I did with that one. The second one contains two 9's and three 7's in a row and so people think it's less random when in actuality repeated digits are pretty common.

A strategy in poker is using randomisers. It comes up in theoretically optimal play, but also for deception, that you might want to bluff with a certain holding say 70% of the time (so an opponent might think you're bluffing a lot more than you actually are). And it's basically impossible for a human player to do this well without assistance. Online players at high levels will have a RNG open to help them because otherwise you really have no idea if you're doing it well or not. Even when we understand the concepts we still can't apply them well.
 
You must log in or register to reply here.

All advertisments are hidden for logged in members, why not log in/register?

All advertisments are hidden for logged in members, why not log in/register?

Back
Top Bottom