Brentford vs Aston Villa pre match ramblings

[Pat Mustard]

It's the amount of sheer smug wankery that's on display in those Tifo things that pisses me off as much as anything else.  I don't need some tosser in a £400 T-shirt (to paraphrase the greatest take-down on a similar theme) telling me I've been watching football all wrong when it's clear that his analysis is so wide of the mark he makes Aleksandar Tonev look like William Tell.  The fact that they could watch the Newcastle game and still not grasp that the numbers they are looking at might be misleading (one of them even admits he didn't watch much of the game) invalidates 99% of what they are saying for me.

As someone else has said as well, if we aren't the second or third best team in the country in the last 18 matches then I'd like to know who is?  The point isn't who is the best team at their absolute peak as we can all admit that there are much stronger squads at a number of clubs, but we are deservedly 2/3 in the form table over half a season now - that's not just a lucky streak.

Oh, and we will stuff Brentford 3-0 this Saturday.

[ChicagoLion]

Compared to the likes of Steve Bruce, who probably takes his laptop to PC World so they can show him where to plug the mouse in again.

Excellent - actually LOL

Made me laugh as well.

[LeeB]

Compared to the likes of Steve Bruce, who probably takes his laptop to PC World so they can show him where to plug the mouse in again.

Excellent - actually LOL

"Ahve tried everythin' pet, put wor boots on, rolled up wor sleeves, nothin'"

[DennisHodgetts]

Can you clean up my history Pet while i'm here?

[CT Villan]

As with any predictive modelling, it is limited by the quality of the inputs and the logic programmed into it. Crap in can equal crap out - just ask the Meteorological Office. xG will not be used in isolation by Emery's team but is merely one point in a battery of tests/analyses.

Clearly though we have just been very, very lucky    ...reeks of panic from Sky and their beloved six that we are about to displace an incumbent.

[hipkiss92]

xG is neither bullshit, artificial or perfect and no one is trying to use it to replace the players (which is a particularly bizarre comment). It's data, nothing more and nothing less.

An important thing about data is that it's always there, it exists whether you like it or not, it just needs to be collected and analysed to see if it can be useful. Useless data generally stops being collected and certainly doesn't evolve into a multi-million pound industry that major clubs across the world rely on.

The other thing about data is that it's value relies on the quality of the analyst(s) using it and very few people who work in the press or who have football podcasts have ever properly worked with football data to create real world benefits, unlike the people in the clubs who actually understand it and how it can be used.

There's a comment about hitting the post counting more for xG and, without wanting to be twat or belittle anyone, that's exactly the sort of thing that shows how poorly understood xG really is.

Every single shot has a chance to be scored but clearly a wild swing from 40 yards is less likely to be than a tap in to an open goal inside the 6 yard box. All xG is really doing is using historical data to say how likely (between 0 and 1) each shot is to result in a goal. The value doesn't change if a goal is scored, if it hits the corner flag or if the goalkeeper pulls off the greatest save of all time.


It doesn't even care who the ball is falling to, so a shot by Haaland gets exactly the same score as a shot by a 17 year old centre half on his debut, all things being equal, but clearly Haaland is more likely to score.

Looking at an individual chance is ok, mostly in a bumping the confidence of someone who thinks he's missed a sitter aspect.
Looking at the average score for chances is very useful (the higher the better because it means you creating high quality chances).
Looking at an individual players score over a season is useful for tracking confidence and efficiency, it can also highlight areas where they do well and others where they could do with more training/practice.

Looking at the total score for the team over the season is of little value.
Looking at the total score for the team in a specific match is of even less value.
Using either of those to claim a team has over or under-achieved is utterly pointless.

However with that I'm going to try to stop talking about it because there are clearly people who aren't interested and we have bene babbling on for most of the week at this point.

Though again, indicative of the quality of chances you're creating as a team, even if you're not finishing them. Can be used to show if a team over a sample of say 10 games is actually underperforming or if they are playing well but just not converting their chances.

XG over a season tends to correlate strongly with league finish if I'm not mistaken?

[Sexual Ealing]

I know some of you accuse other accounts on here of being run by multiple posters. It feels like a number have handed their login details to Richard Keys, Neil Warnock and the editorial board of TalkSport. The proper-football-man-it's-all-bollocks-just-stick-it-in-the-onion-bag vibes are strong here today.

[Monty]

'The onion bag' is my favourite cliché that I absolutely hate.

[Stu]

I know some of you accuse other accounts on here of being run by multiple posters. It feels like a number have handed their login details to Richard Keys, Neil Warnock and the editorial board of TalkSport. The proper-football-man-it's-all-bollocks-just-stick-it-in-the-onion-bag vibes are strong here today.

I guess. But I have some sympathy for the idea that the current trend of stats-based analysis doesn't tell the entire story of the reality - and Danno made a really good point earlier in the thread when he said that anyone involved with stat analytics as a job shouldn't really be putting things down to luck; either the stats are right or you have an opinion.

I'm pretty cynical anyway and get the feeling that a lot of podcasters etc don't really know what they're talking about when it comes to stats and regurgitate figures which are, at best, abstract to what happens on the field.

[Stu]

'The onion bag' is my favourite cliché that I absolutely hate.

Get it in the mixer

[Sexual Ealing]

I know some of you accuse other accounts on here of being run by multiple posters. It feels like a number have handed their login details to Richard Keys, Neil Warnock and the editorial board of TalkSport. The proper-football-man-it's-all-bollocks-just-stick-it-in-the-onion-bag vibes are strong here today.

I guess. But I have some sympathy for the idea that the current trend of stats-based analysis doesn't tell the entire story of the reality - and Danno made a really good point earlier in the thread when he said that anyone involved with stat analytics as a job shouldn't really be putting things down to luck; either the stats are right or you have an opinion.

I'm pretty cynical anyway and get the feeling that a lot of podcasters etc don't really know what they're talking about when it comes to stats and regurgitate figures which are, at best, abstract to what happens on the field.

No doubt the broadcasters, such as they are, are at times approaching the data all arse about tit, but to dismiss the entire concept as nonsense is prehistoric.

[danno]

'The onion bag' is my favourite cliché that I absolutely hate.

Get it in the mixer

Top Bins

[Jon Crofts]

A routine 0-2 away win. FTF.

[Toronto Villa]

Press conference

https://www.avfc.co.uk/video/0_0zl69eb5

Fucking hell, he's never completely happy and always wants improvement. Don't rest on your laurels and keep driving forward.

[Monty]

A routine 0-2 away win. FTF.

Can someone explain this comment pls