Spurs in full cry here.

Lonella! Bye Bye Yeah, Yeah.

Well, well worth the wait.

Yeah, absolutely.

We've got a bunch of questions, some good ones.

Some from a week ago, so apologies that it's taken us this long to ask.

Cameron Williams emailed in to podcast at theexchange.com.com. He says, which advanced football stats are best for analysing a predicting player productivity? In other words, if you were asked

to analyse player X, what would you first look to gauge how good they are relative to others?

Nathan, I mean, you are literally asked to analyze Player X, kind of what you do on the side.

So what's advanced stats to you tend to use?

It kind of means what you mean by advanced, right?

Because if you're opening that up to something that's like progressive past distance and stuff like that, those are all very good.

But if you're using advanced to sort of mean predictive as in not what they're actually so much doing as to what you might expect them to do in the future, then expected goals and working backwards from there are really the only thing on their level.

All attempts at sort of doing really clever stuff like that with defensive actions and passing actions are still falling short.

And that's because, you know, football is a shooting game and the goals are the most important thing, obviously.

But it's just, there are so many, many more moving pieces in every other aspect of the game.

And you can do a lot to get around that.

And you can do a lot by helping with the eye test, of course.

But really, and you can do a lot working backwards from expected goals.

There are models I've seen and I've used that are built entirely backwards from expected

goals where you say, okay, here is the contribution towards expected goals that this player

is putting in through their tackling.

...