Where will he end up ??

I think it's somewhere in between. There is a lot of variation, but that's where the mass modelling comes in.

I like the concept of being able to look at chances taken from an area and seeing the average number that would be converted, then being able to see where one player performs above or below average.

Fair enough, everyone has sockets and you can't predict everything, but I don't like to write it off. If it's something BFC can use to gain a slight material gain, I'm all for it.

Plus, it annoys me that a lot of people who are vocal in their dislike come across as 'it's just stupid numbers', without having actually looked at what it is. That's not having a go at you, you obviously know what you're on about!

 

I work in MI as an analyst, and I don’t gamble at all (my jobs to analyse what’s happening, not what might happen next).

But some of the best analysts I work with have serious gambling habits, because if you’re serious about the science, you can outdo the bookies, who have to weight probability against the market too.

Back to Xg, it’s not intended to be a prediction, but it measures probability based not only on player behaviour, but as data grows, on how team behaviour influences player behaviour.

If you look at our Xg from our last promotion season, you’ll see it accurately described our outcome. If you just look at our ‘shots on target’ number, it would suggest we underachieved. But the reality if you’d watched us was that tons of our shots were from well outside the box, ergo were rarely going to go in. Shots on target = an actual count that is less useful than Xg

 

That spreadsheet built the team in 6th place in the Championship for a fraction of the cost of the 7th 8th or 9th team.

Of course it’s not 100% accurate (no prediction can be). But you can’t deny it’s bloody good.

 

There's no exact science to signings, the most successful manager of all time spent a fortune on Juan Sebastian Veron.

The spreadsheet has identified more successes than failures. Problem was the rate of selling them.

 

My issue with XG is not so much about it's value as a tool for the club to be using to analyse performance and help improve, it's when it gets quoted by people in the wrong context and you get fans of other clubs using it to claim they 'deserve' something more than they got, rather than accepting that maybe they're just **** at converting what should be gilt-edged chances.

You need to be careful how you interpret the data though. Story I heard somewhere was that Alex Ferguson sold Jaap Stam because the data told him Stam was not making tackles, but he figured out too late that Stam hadn't needed to make the tackles because his positioning was so good he generally got to the ball ahead of anyone else. Could be wrong, I think it was Stam he said that about.

 

I think I'm with you there. Those people are in the same group as one who just write it off, and in my mind are highly likely to crack out things like 'teams like Barnsley'!