Rich

They've just been relegated, they've got next to no leverage.

Hope you win!

Slim runs about four hours behind the rest of the world these days, like.

That's great to hear, good luck with writing it all up!

Nah, still think he has a big role to play. Along with Miggy, he's always been the biggest weakness on the "strongest XI" teamsheet for this year, though.

Think this was always the case, tbh. Joelinton and Bruno are nailed-on, and Shelvey was probably ahead of Willock before his injury.

A fair few journalists have said we've asked to be kept informed on Broja, Pulisic, Hudson-Odoi, and Gallagher from Chelsea.

Re: Martin Hardy in The Times saying a move for Gallagher is "progressing". Based on everything I've read about / listened to on Chelsea lately, they aren't willing to loan him out again unless it's a deal with a guaranteed purchase at the end of it. Assume that's what we'll be working on, if true.

Why would Chelsea do that deal, though?

I'll make one for your mam.

He clarified that two Tweets down. How does that make a news story?

The fact they're advertising it so publicly points to the fact that it wasn't planned, but no idea what the story is behind him leaving.

They look trash but they broke at least one of our signings in Jan, if I remember rightly.

You just have to brush it off. Dementors gonna dement.

Advert for new Head of GK coaching now on the website too.

Ah, it would be the same aye.

I'd always take that approach, too, because when it does come off, as with Trippier, Bruno, and hopefully Botman, we'll be laughing. While we'll still all be happy enough if the others we bring in improve us as much as Burn and Targett did and continue to do.

No, never. The data itself would tell you whether a player was in form or not, as you could look at similar instances for Torres ~2/3 years prior where he'd probably be outscoring his xG in most situations. His regression was quite easy to see in the data. We would also produce match reports on hundreds of games a weekend where this sort of stuff would be discussed, to add some context to everything like you just did above.

Aye that was always a really tricky situation, because you're getting into the realms of player positioning and anticipation, which is a nightmare to quantify. If the player got close enough (again subjective) then it might be assigned a score. Our approach towards the end of my time was to start grading chance-creating passes like the one you mention as their own significant action, and trying to apply that as a weighting on a final xG score, and it also had the handy effect of allowing us to better judge the value of creative players who don't necessarily score many themselves.

This was another thing where I felt like we did it better than others, actually, it's a great question. Back then Opta's model and most others would count that Lingard effort as double. However, we used phases of play and only took the best chance from a single phase of play so as not to give double credit for someone missing and then getting the rebound. Again, it was a big step forward for the model when doing that, and I think a much more common sense way of handling those instances. If he scores the first, then the second never happens, so you can't give double xG credit IMO. Because we were using professional analysts who could read the game quite well, we would also apply a bit more subjectivity to your second example so as to better reflect the xG with an adjustment, but other models absolutely didn't do this, and probably still don't. It was always tricky to know how far you could take that, though, because you're getting into the realms of trying to predict player actions and also "knowing better" than the players. Current models would play it safe and just assign a score to Bowen, ignoring Lingard, if they even assign a score at all if no shot is taken. Some models rely on a shot being taken to assign a score at all, which is wild to me. So Bowen could dribble clean through on goal, fall over for some reason, and no score would ever be assigned because he didn't actually shoot. Added: We would assign a score to the "best" point on the pitch when he still had full control of the ball, and it would then go down as a "miss" to the minus value of that score on his individual metrics.

This is fair, like, and it was much the same in January when we thought we were getting Diego Carlos/Botman, Digne, and someone like Isak then ended up with Burn, Targett, and Wood. It was fairly underwhelming but thankfully turned out very well. It's also why I'm still pretty relaxed about everything, though. Only Wood has been a duffer so far, individually speaking, and we've now brought in seven players. If they can keep that sort of record up when not being able to get the #1 targets, for whatever reason, then we'll keep progressing nicely. I don't think we'll be signing anyone without having done plenty of research into them and feeling like they'll improve the squad or first team, it just won't happen under Howe and the others. Pedro will have always been on the list.

But aye, sorry to clog the thread, just had a bit of time while eating lunch and love a bit of xG chat. To sum up, the current models that everyone's being fed by the media are way, way off the stuff being used behind closed doors by clubs/bookies/gamblers for their "edge". That edge (and the sums it involves) is the reason why progression is so, so slow publicly with xG. The current models have got so many glaring weaknesses that they can't be relied upon as a single measure of team or player performance, but from everything we get given by the likes of BBC and Sky, it is still, by far, the best individual measure of judging how a game went without seeing the game for yourself. And it's still definitely a useful barometer to bring into any debate around the quality of a team or player for that reason.

Opta's public model, aye. Not the one they sell to all of the clubs, bookmakers, and professional gamblers for hundreds of thousands of pounds each per season. It was ~0.8 at the time for an average open-goal effort from inside the six-yard box, obviously meaning only 4/5 were scored.

They absolutely did, aye. I just mean publicly, there's not enough data out there to do so still today for us plebs. The likes of Brentford will have so, so much more data. I was working with Matthew Benham's right-hand man for most of my time being involved in data analysis and models. We basically had Brentford's base model from ~2010 as our starting point, then spent the next few years making tweaks and improvements. SmartOdds (Benham's company) were no doubt doing the same with a team 10x the size, of course . When we got access to more video and when our team was big enough, we built our own model with everything we had learned. We got it to a point where it was really, really good considering the limitations of the time. The amount of data I used to have to wade through was glorious. It's amazing how quick that knowledge goes when you no longer have access to it all, sadly. No regrets at all, but I'd love to know where we'd be at if I hadn't stepped away in 2017.

This is it. And this was the issue back in 2014, and still doesn't appear to have been properly rectified in the public domain. At that time the biggest step forward was adding active attacking/defensive bodies between the ball and the goal (in a sort of cone shape from ball to both posts) to adjust the base level xG for the position on the pitch. I believe we were the only people using this metric back then, and it made our model miles better. From there we introduced "pressure" metrics to add another layer of context in terms of how "free" the player was when taking the shooting action (be it header or shot), which was another step forward in helping to explain how an uncontested shot from 12 yards on a counter-attack was different to a contested shot from the exact same position that might occur on the second phase of a corner kick or other set piece. The thing that tickled me the most was how a completely open goal from inside the six-yard box was nowhere near as close to 1.00 as you would expect. Largely because those situations were quite rare compared to other shots, but also because people missed them with such frightening regularity at all levels.