This post is about 'possession adjusting' defensive stats, but really it's about

- exorcising past demons
- the extent to which making player statistics comparable to each other is a science vs a craft
- extended metaphors

get-goalside-newsletter-archive.beehiiv.com/p/possession...

And as the biggest reminder ever, probably not needed for Stoke fans, but even so:
Head coaches/managers are *highly* situational. The search for the ‘best’ manager at our level isn’t a thing, it doesn’t exist! It’s the *right* person for your process at that time that’s needed.

But how many lessons is it possible to learn?
How many individuals do we give full control of our footballing direction before we find success? What even is success, as apparently midtable after 5 games isn’t it?
If we do find it,what happens when they leave? A dice roll again?"

2/2

It lost me when I renewed last year, and the basic subscription suddenly stopped providing you with basic team rankings and stats, and only gave you 70 minutes of viewed footage and a report a month.
Absolutely mad.

Sorry, bit of a long one! It's been a while and there's lots to say 😅
Data-driven as much as it is frustration-driven, but any comments or questions always welcome, enjoy!

In fairness I'd be very surprised to see him go anywhere. I'm not sure he's got the feet to be in the PL just yet, and it'd take a huge huge bid for Stoke to let him go.

As an first glance, I actually think we’re on a similar points total to what would be expected on the chances for/against, but nothing concrete yet on it.

There is! That's my next plan, may take a bit of time mind, but will hopefully be in the final piece later this week!

As a quick explainer, I essentially take each shot and simulate it 100,000 times, with the xG of the chance taken as the probability of it being scored.

There are big caveats on this, especially in the treatment of lots of shots in one sequence, but it's reasonable for this use case.

Now actual Sandi Toksvig please.