Expected Goals

[SouthStandUpperTier]

It’s the 7th November so my expected shits for the month so far is about 5 but as far as I can remember I’ve only had one. Looking forward to a flurry of shits over the next couple of days do get my xShit back in order.

You're obviously attempting your shits from too far out and from too acute an angle. Get closer to the toilet bowl my friend.

 

[Fox92]

25 minutes ago, Poznan34 said:

So, say, we've peppered a team's goal with shots but not scored and the other team has had one shot from 30 yards out and scored and won 0-1, how do you feel at the end of the game? Is it "Oh well, the other team deserved the win" or is it "we were unlucky, we should have scored a couple there, and their's was a bit of a fluke"?

 

Do you ever say "he should have scored that" when a player misses from 6 yards out?

 

That's basically expected goals. It's just putting a probability of a shot going in based on various factors that we all subconsciously think about when watching a game. Distance from goal, the angle of the shot, the proximity of defenders. 

 

If we continue to take higher quality (nearer and more central) shots then we will score more goals in the long run. Over a season or so, the teams with the highest expected goals will have the most goals. 

 

As others have mentioned, football is very low scoring which means there's a lot of luck involved. This means, in the long run, expected goals is a better predictor of future performance than just looking at results. 

 

However, expected goals isn't a great name, think of it as "chance quality."

 

(Also, I can see how if you're just not into football stats, it can be dull! That doesn't mean it's not useful though!)

I like stats (well, some stats), I've even got some of Opta's books.

 

If I say "he should have scored that" - as I do - then it counts as a miss. This "expected goal" thing is shite imo. If he misses, then he misses.

[skmanuk]

17 minutes ago, Rogstanley said:

It’s the 7th November so my expected shits for the month so far is about 5 but as far as I can remember I’ve only had one. Looking forward to a flurry of shits over the next couple of days do get my xShit back in order.

School must be out!

[boots60]

95.4% of all statistics are bollox

[stripeyfox]

I wish they would publish the expected goals on the ticket website. This would help me decide whether it is going to be worth going or not.

 

What a load of shit!
 

[Gerard]

@Wookiehas absolutely sodomized this thread. 

 

Expected goals is the most relevant stat in football, far more important than possession, shots or shots on target. It's amazing how many people have completely dismissed it as irrelevant.

 

A good analogy would be poker, some times you can do everything perfectly and still lose to a bad beat. This happens plenty of times in poker and football but if you keep winning the goal expectation stat or making the right play in poker you will win more than you lose long term.

 

 

[stripeyfox]

3 minutes ago, Gerard said:

@Wookiehas absolutely sodomized this thread. 

 

Expected goals is the most relevant stat in football, far more important than possession, shots or shots on target. It's amazing how many people have completely dismissed it as irrelevant.

 

A good analogy would be poker, some times you can do everything perfectly and still lose to a bad beat. This happens plenty of times in poker and football but if you keep winning the goal expectation stat or making the right play in poker you will win more than you lose long term.

 

 

It's meaningless. 

 

Expected goals - what a crock!
 

The random element of football, the break of the ball, the rebound, the mistakes are what makes it great. 

 

It may help slightly in predicting the outcome, but not much. 

 

It is a bit "Spursy" for me.

 

I'm out!
 

[Gerard]

Just now, stripeyfox said:

It's meaningless. 

 

Expected goals - what a crock!
 

The random element of football, the break of the ball, the rebound, the mistakes are what makes it great. 

 

It may help slightly in predicting the outcome, but not much. 

 

It is a bit "Spursy" for me.

 

I'm out!
 

 

That's because you don't fully understand it. Whilst I agree it's subjective there will be a huge number of sample size to determine the chances of any shot and it will be quite accurate. I would be worried if we were 7th in the table but had lost the expected goal stat in every game we played as it would suggest we were one of the poorest sides in the league and when we had a bit of mean reversion we would plummet down the table.

[stripeyfox]

5 minutes ago, Gerard said:

 

That's because you don't fully understand it. Whilst I agree it's subjective there will be a huge number of sample size to determine the chances of any shot and it will be quite accurate. I would be worried if we were 7th in the table but had lost the expected goal stat in every game we played as it would suggest we were one of the poorest sides in the league and when we had a bit of mean reversion we would plummet down the table.

You're right - I don't understand it!
 

[sdb]

I like this sort of thing. I'd imagine that stats look good party because none of our midfielders would ever dream of shooting (Gray and Mahrez are forwards ?). ..

[SouthStandUpperTier]

Expected goals is the most relevant stat in football

No, the actual scoreline is the most relevant stat in football.

[Poznan34]

If anyone who doesn't understand wants to try to, rather than rejecting it out of hand: 

 

 

[SouthStandUpperTier]

Expected Goals doesn't take into account that if a player misses a chance he would usually expect to score, or vice versa, then the game moves on in a completely different reality to what should have happened. The stats accrued after that incident wouldn't happen because it would be a completely different game. Different action, different tactics etc.

 

 

[MC Prussian]

The whole thing about "expected" goals is a lot of tosh. Also, expected goal difference is a similarly wacky term. It's getting silly now with statistical analysis of football, it's merely a playing field for the bored.

 

Why bother about the creation of "high-quality chances" when you can't convert them?

Ultimately, goals scored tells the story, not the amount of goals we could've scored.

[Woof]

Possibly one of the problems with this stat is the use of the word "expected".  It is the word used in Statistics for mean or average.  All it is saying is that over thousands of different kinds of shots on average  a % result in a goal.  Of course as we all know things vary around the average, but over the course of a season the average scored from the shots you take will tend to get closer to the actual number you score.

[HighPeakFox]

Goodness me, people who aren't me being pedantic about use of language. What is this site coming to? 

 

[HighPeakFox]

You're a little late to this particular party...

[Nod.E]

14 hours ago, Fox92 said:

I see they've added more rubbish to football analysis. "Expected". Have a day off.

 

We've drawn, lost and won games because that's what we've done. "Expected goals" mean fvck all. If you don't take chances then you don't score.

Please. Nuance.

[Guest]

but now we have a run of two top 6 teams and 4 "others" 3 of whom are away from home. According to the 52pts 8th place finish that's only 1 pt from the next 3 games and 6 from the next 6. 

 

Let's hope we can upset that prediction and go for Europe.

[yorkie1999]

Load of baloney. With an almost infinite amount of variables during a football match it's impossible to predict a future result based on a previous result, let alone basing it upon a previous goal or save. Nerdy stuff for nerds. Having said that i did collect a bit of a win with the 2-2 against stoke and the other year did the same with 3-2 against west brom purely basing it on the previous years result. It's a funny old game. 

[RumbleFox]

2 minutes ago, yorkie1999 said:

Load of baloney. With an almost infinite amount of variables during a football match it's impossible to predict a future result based on a previous result, let alone basing it upon a previous goal or save. Nerdy stuff for nerds. Having said that i did collect a bit of a win with the 2-2 against stoke and the other year did the same with 3-2 against west brom purely basing it on the previous years result. It's a funny old game. 

ON AVERAGE though the team that has the most clear cut goal scoring chances will score the most goals over a large enough sample size.  Of course, based on individual matches (or even groups of matches in a smaller sample) all kinds of freak results can happen and football is beautiful for this very reason, it is highly difficult to predict.  If it were truly random though a different team would win the league every year.  Odds show a likelihood of what is expected to happen, if it did not work then bookmakers would not exist as they make a living from it.  I find it difficult to understand (though I am a bit simple) why some people think you either have to believe in stats all the time and they are never wrong or you think they are shit.  The truth is nuanced and more interesting.  Over a large enough sample, given enough matches a picture can be created that allow us to say "x outcome" is statistically more probable than "y outcome".  Of course this does not mean that "x outcome" always wins but that doesn't mean the stats are wrong, just that people have misunderstood them.  Anyway, I realise I am a massive dullard with a tiny, tiny penis so I will stop now.

 

X

[norwichfox]

as Ex England centre forward Benjamin Disraeli  said....there are lies, damn lies and statistics. As written above, what matters is the score at the end of the match, as I see it, this particular statistic will determine your level of luck.

[RoyFox]

Reading this is hilarious. The best coaches use it, Pep more so.

 

Using expected goals allows you to enhance your teams ability to score, as it’s tied into; location of shot, the number of defenders present at time of shot, situation (counter etc) and angle to the goal. 

 

Ever notice that most of the Man City’s chances are gilt edged? Also the ball is never crossed high, but low into strikers? There’s a reason for that.

 

If Gray is coached properly to shoot inside the area rather than taking pot shots from outside (2/3% chance of scoring), he’d score a lot more goals.

 

Thankfully Vardy has the natural instinct to shoot from good locations (he’s one of the highest xG strikers in the league). Okazaki also adopted this at the start of the season (he said in an interview as much), hence his goals have picked up.

 

Expected goals isn’t the ‘be all and end all’,  but it can give you a good guide to how a team is truly performing and whether a current win or losing streak is due to variance (luck). Eg Burnley will regress very soon, their expected goals (for and against) isn’t sustainable, they have been lucky with recent results. 

 

Likewise coaches can use this to analyse expected goals against. Can the team change the positioning on the pitch of the back line? Will blocking in certain zones enhance the teams ability to prevent opponent strikers shooting from high % situations? 

[CosbehFox]

11 hours ago, SouthStandUpperTier said:

Expected Goals doesn't take into account that if a player misses a chance he would usually expect to score, or vice versa, then the game moves on in a completely different reality to what should have happened. The stats accrued after that incident wouldn't happen because it would be a completely different game. Different action, different tactics etc.

 

 

That's any statistic involved with football full stop though. Possession is affected by goals and timings. And that continues into short/long passes. Key passes more likely to occur in an open game rather than a close game. 

 

There's so many elements and variables in Football. It's not possible to 'Moneyball' it like a few American sports. 

 

Expected goals is probably the most intelligence statistic there is for football because there's a context there - i.e. distance of shot, lack of defence. This doesn't exist with possession stats where there's no considered context - e.g. two centre backs passing it between themselves. However because of football's variables its no absolute answer. 

[Ted Maul]

Forget how it’s calculated, it’s incredibly complex- it’s a stat to determine how good a chance is, simple as that.

 

For example, Vardy going through 1 on 1 would constitute a better chance than Gray shooting through a crowd of players from 30 yards. If both efforts are saved by the keeper, under the classic stats, both go down as 1 shot on target. With this, it tells you a bit more about the type of chances being created.

 

It would have been very interesting in our title season- Spurs fans were whinging about possession stats and shots on target figures, but a lot of our chances were real goalscoring opportunities and we restricted a lot of opponents to shots from range.