Jagten

Not that close in the end https://x.com/optaanalyst/status/1746253844739186861?s=46

Def getting on board with this guy being our best player

They’re making us look like a plucky lower-league team who’ve gone too far in the FA cup. Not sure who this says more about.

Given our ambition and FFP constraints we want to buy elite upside optionality. The perceived lower variance outcomes of buying domestic make people feel like we're not getting that (even if they have no opinion of X foreign player). Once a player has been a success to the level we now require domestically, buying the elite upside cheaply is incredibly difficult to do. We need a unique catalyst for PL value, which is rare at the level we're looking for - age, contract, financial trouble, etc.

Solanke is better. Jonathan David is mid as and there's a reason nobody has bought him.

Interesting because a lot of their numbers are actually slightly worse than last season. If I had to bet on anyone to fall off it would be them.

Depends on the model, but not generally. It can additionally be a function of many variables, such as opposition player positions, ball speed, ball trajectory, etc.

He is clearly very good and is much better than Gerrard but even a 38-game season isn't a very big sample size statistically. Union Berlin are a great example of this, who were the underlying-numbers overachievers in Europe last year. Now things look much less great for them.

For sure, but the extent to which we can attribute other injuries to the physical demands of our style is probably non-zero I think Tonali’s greatest asset is out of possession, and I don’t think he’s quite the technician many might hope. What’s odd is that Barella definitely was that person, and I’m not sure Tonali was the next logical step

It was the profiles that were disappointing, more so than the positions. Whilst neither have really played (and what I say might turn out to be incorrect), Tonali and Barnes reflected a concentration of our position in our existing tactical framework, failing to consider that aspects of it might be unsustainable, especially under fatigue (some of which was very unlucky).

Great post. I thought, and still think, we have gone too long Howe and last season's tactical template. Ashworth probably assumes most of the responsibility here, as his job should have been to hedge this (with e.g. the profiles you mentioned).

Opta's xG model parsed without incorrect chance accumulation, for fellow statos. Still all-time terrible

zinger!

Shame our opposition tonight didn't have the same perspective, and then maybe their team would be less good.

I think because most xG is used as the objective function for modelling player performance, with deriving stuff like "on ball value". At a micro level individual chance creation is maybe more important, but there are differing views.

Most public models do this, correct, but for results analysis it should really be one minus sum of prob of not scoring in continuous passage, i.e. two 0.7xg chances in the same passage gives 1 - (0.3 * 0.3) = 0.91xG.

The numbers have, until now, supported him, but that expected goal diff is as bad as it will get for any team in the league in basically any season.

It could have been 10. Probably a record xG conceded this season.

Approx 3.5 xG vs our 0.1, for fans of violence.

Well, that was mostly horrific.

But he was also complicit in his signing at 85, and he’s not worth 40m either. I’d sack everyone involved tbh.

But that’s not inferring chance quality. Anyway, this is all an approximation.

Fair enuf. I think most of our season’s underlying performances have been mostly good (and better than sentiment), but we’re starting to concede a high volume of good goal scoring opportunities.

Maybe of some interest - a ten game rolling (exponential) moving average of our xG for and against over the last three seasons. You can almost see the moment Bruce was sacked.

Fair to question whether there’s value in buying from Arsenal. Your reply was weirdly aggressive.