Emi Martinez

[Footy-Vill]

Emi Martinez last 5 penalty shootouts 24 penalties 10 Saved
Won all 5
No body does it better!

[chrisw1]

I know XG isn't popular here, but I think Martinez is a major contributor to us outperforming our XG by so much this season - 12-13 points by the look of this chart.  Before I saw this I'd have guessed that compared to an average keeper he's probably saved us 8-9 points this season.  Probably not too far wrong?

[pablo_picasso]

But what is that based on?

If it is ANY data that has to have human input at the start to work out the math, then it is subjective & therefore completely pointless.

One only has to look at where we have been in & around 3rd-4th in the overall form table from when Emery first joined to know that we are where our performances & results dictate where we "should be".

If some random data inputter named Colin isn't knowledgable enough or intelligent enough to research Aston Villa Football Club properly, then we end up with the "Villa are an over-performing mid-table side" agenda talking point & then other equally wilfully ignorant morons parrot that garbage based on xG, because "look, maths that I don't understand either".

xG belongs in the bin.

[pauliewalnuts]

If it is ANY data that has to have human input at the start to work out the math, then it is subjective & therefore completely pointless.

This is it. People assume it is mechanical, scientific fact, when it isn't, it involves subjectivity.

[Bad English]

It's maffs. Can't be having math.

[chrisw1]

You can argue that XG is irrelevant all you like, but it is a stat used worldwide as a performance indicator and I'd imagine clubs data analysts look at it when assessing players even if you don't. 

But that's not the point in this instance.  We are performing above XG and as you point out have been for some time.  I'm just observing that one of the reasons for this probably is that we have the best keeper in the world.

[Bad English]

To me, the great thing about Xg or xG, whatever it is, is that you can be poring over it like a 13 year-old who has found some hedge porn or you can, understably, completely ignore it and simply enjoy/suffer the football. ;-)

[pauliewalnuts]

But that's not the point in this instance.  We are performing above XG and as you point out have been for some time.  I'm just observing that one of the reasons for this probably is that we have the best keeper in the world.

Exactly, and that's why I take XG with a pinch of salt - it ignores facts like the difference between having the world's best goalkeeper in goal for us, and having Olsen there.

Not (really) saying XG is bollocks, I just don't see it as some scientific, infallible measure like a lot of YouTube football hipsters do.

[pauliewalnuts]

To me, the great thing about Xg or xG, whatever it is, is that you can be poring over it like a 13 year-old who has found some hedge porn or you can, understably, completely ignore it and simply enjoy/suffer the football. ;-)

And besides, all the really cool kids have moved on from XG now and onto the world of financial reporting and commercial revenue.

[pablo_picasso]

You can argue that XG is irrelevant all you like, but it is a stat used worldwide as a performance indicator and I'd imagine clubs data analysts look at it when assessing players even if you don't. 

But that's not the point in this instance.  We are performing above XG and as you point out have been for some time.  I'm just observing that one of the reasons for this probably is that we have the best keeper in the world.

And I will argue its pointlessness until the day I die. 😉

It's flawed data because the initial data is completely subjective. The fact that they then spin the numbers through a bingo wheel adds the grand total of fuck all to those subjective numbers.

I would agree & say that having one of the best players in the world in his position would add to where we are in the table. Having a few more should push us up the table even further.

I don't need xG to tell me that though.

Neither do I need xG to feed the narrative that Villa are "over achieving".

Because we aren't.

Ever since Emery arrived we have been between 3rd & 4th in the overall form table, so thats where we should expect & deserve to be.

Im not trying to be difficult, I just fucking despise xG & think it is a worthless tool.

[cdward]

, I just fucking despise xG & think it is a worthless tool.

That's funny, i feel the same way about SG

[paul_e]

The problem with xG isn't the stat itself, it's that so many people don't really understand what it's designed for so they use it to make points that they shouldn't.

That graphic for example will be them rescoring every game in line with xG and creating a new table.

The purpose of the stat is nothing more than to add a little additional qualitative data on top of the shots and shots on target stats. It should be used for things like asking if we making the most of the chances we are creating or if we're giving away too many good scoring chances. Both of those are questions that are difficult to answer (statistically) with the more traditional stats. It shouldn't be used to say 'we should have won today, look at the chances we created' or 'player x is better than player y because of xG/xA' and it absolutely does not gve you a 'real' table as some people try to suggest.

It is not 'completely subjective' though, it's based on fairly simple probabilities. What it represents is: across thousands of games players shooting from that position, using that foot and with a ball that is bouncing/on the floor/whatever they have scored 14% of the time. Is doesn't care about ability, footedness, the quality of the keeper, etc because that's not how it was ever designed to work. Which leads to the most important part of the whole thing that people get wrong far too often, being above xG doesn't mean that you're lucky, it means that you're above average.

On this season it shows that, despite scoring a lot Haaland has had a very poor season, and I reckon most people who've watched them regularly would acknowledge that, whereas Watkins and Palmer have both been exceptional both in terms of outperforming on xG and xA and just by eye for anyone who's watched them play. This is what really matters, xG, at it's best, should reflect what fans could tell you about specific players already, it's not trying to do anything more than that and most of the complaints are caused by people trying to flip the stat to give themselves a moral victory of sorts.

[pablo_picasso]

There are some fair points within that, so thanks for explaining it fairly simply.

One thing I would say is when they are looking at teams "over-performing" using this tool, it is absolutely subjective & pointless.

But like you say, they are using it incorrectly in this instance.

And I can understand it being used in the way you described in your "14%" example.

Tells us what our eyes see, but as we cant be at every game, it could be a useful stat for scouts, etc.

But for a table to prove a club like Villa are over performing?

Pointless...

[Axl Rose]

Good fucking lord.

[chrisw1]

Good post Paul.