[Albion] The xG comparisons to last season thread - includes graphs!

[HP Seagull]

I know not everybody thinks xG is a useful stat, so for those who don't like it - look away now! I know it's not a perfect science but if you're interested keep reading...

First off let me say I'm no mathematician or stat anorak so if anyone has any suggestions on improvements please let me know!

The table uses "xGoals" to create an "xResult". For example in yesterdays game against Arsenal the xG was Arsenal 0.97 - 1.53 Brighton. If we round the xGoals to the nearest whole number we are able to get a result - in this case it would round to Arsenal 1 - 2 Brighton. In this case xGoals nails the actual result but it's not always like that....

For that reason I've also included ΔGF (The difference between our xGoals and our actual Goals) as well as ΔGA (The difference between our xGoals Against and our actual Goals Against)

Points and xPoints shows how many points we've gained in that match according to actual Goals and xGoals.

ΔPoints shows if we've gained more or less points than our xResult indicated. For example the West Ham game earlier this season we had an xResult of Brighton 1 - 0 West Ham, but the actual result was Brighton 1 - 1 West Ham meaning we lost 2 points from the xResult.

And finally taking a page out of [MENTION=3734]Giraffe[/MENTION]'s wonderful Tracker we have the points tally shown in CUM Points and xCUM Points.

I'm not usually very good at explaining these things, so the best way is to probably just look at the tables below and you'll be able to see what I've been doing!

All xG data is from www.understat.com


Season 2018-19

View attachment 117905
View attachment 117907






Season 2019-20

View attachment 117906
View attachment 117908

So what does this tell us so far? Well it really shows what I'm sure most of you remember from last season that when we had that run of 3 wins on the bounce last season we got very lucky against West Ham, Newcastle and Wolves. It shows that at this stage last season we were 10 points better off that our play "according to the xG stats" deserved. We'd scored 1 goal more than our play "deserved" and we "should" have conceded 6 more. This season it shows we are more or less doing as our play deserves, we our just 1 point behind our Cum x Points. The surprising thing is that our xG against Watford show we should have drawn the game 1-1, but I remember the game being much more one sided than that. Ah well, like I said before it's not a perfect science but I find it interesting all the same. I can imagine it's the sort of thing Tony looks at and is one of the reasons he decided a change was needed in the summer - we got very lucky with some of our results last season and only stayed up by the skin of our teeth (and Cardiff's ineptitude)

Viva la Potter revolution!

I'll try and update this thread periodically through the season.

Great post - very interesting. Thanks!

 

[Dolph Ins]

Great post. I must admit I am surprised how closely xG and actual results match.

 

[BN41Albion]

Great post. I must admit I am surprised how closely xG and actual results match.

Same, i expected us to be below our xg. I suppose it shows that despite our dominance in some games we ended up not getting 3 points in (Newcastle springs to mind!) we didn't actually create many chances. Hopefully that'll be the area that we continue to improve in over the season.

 

[Perkino]

Same, i expected us to be below our xg. I suppose it shows that despite our dominance in some games we ended up not getting 3 points in (Newcastle springs to mind!) we didn't actually create many chances. Hopefully that'll be the area that we continue to improve in over the season.

It's all about the decision making, we will improve in that area the more often we find ourselves attack during matches. Creating better chances will lead to better movement off the ball and subsequently a larger amount of goals

 

[rogersix]

A bit rude?



As did I. Thanks @Ⓩ-Ⓐ-Ⓜ-Ⓞ-Ⓡ-Ⓐ

he's scotch

 

[rogersix]

How come I can't @ Zamora?

lose the space

 

[Surport Local Team]

Was it palace away last season when our goal expentace was 0.33 at a guess, but won with two wonder goals! Uta

 

[Ⓩ-Ⓐ-Ⓜ-Ⓞ-Ⓡ-Ⓐ]

21 Game Update

2019-20







2018-19





3 points worse off than our xG stats suggest we should/could be, at the same stage last season we were 9 points better off than our xG would indicate.

Just another way of highlighting the progress made this season

 

[Moshe Gariani]

BHA 1.57 CHE 1.57

Amazing exact draw on xG.

Reflects other match stats (and indeed the match itself) which were remarkably even.