04 Sep 2024 18:11:31
Apologies Eds. Long term reader who never really posts. I have been listening to the Ian Graham book 'How to Win the Premier League'. Not sure if any of the Eds have come across it or read it. What's your opinion on the book and data analysis in general. Some of the stuff in the book seems to echo or say things Ed001 said years ago about Rodger's and there is a bit about Liam Ridgewell that actually kind of reminded about Ed001 talking about Skrtel. General if you looked at certain numbers and stats it would paint Liam Ridgewell as a far better defender than he actually was in reality. I think I remember Ed saying something similar about Skrtel, correct me if I am wrong Ed!
{Ed001's Note - I haven't read the book yet but I would like to, he has sold it well in interviews. I love data and stats analysis but I hate the way it is misused so much. You are right, I did say something similar about Skrtel and data in general. It is understanding the data and putting it into context that makes it useful. Unfortunately it is usually misinterpreted or used in isolation with no context applied.
It was N'golo Kante and Gana Gueye that should be used as the ideal example of misinterpreted/misused data without context applied. There were a number of pundits claiming Gueye was better than Kante simply because he made more tackles. They even ran polls about who was better and it was a close call in most of them. But if you looked at the figures overall and applied context, it is clear Kante was on a whole other level. They never even applied any basic context like possession. Gueye was playing for a team that had much less possession, so there were more opportunities for him to make challenges.
It is very difficult to really use data anyway, simply because there are not enough games in a season to really build up enough data. It leaves too many anomalies with so few occurrences. You need to be repeating actions thousands of times to really get a true picture. That is why teams like Brentford and Brighton, who were built on the back of an owner's data analysis company, only actually achieved long term Premier League stability when they added back scouting on top of the statistical data. It is still not fool proof, but it gives you a better chance of weeding out those anomalies and recruiting the right players.}
04 Sep 2024 22:42:10
Gueye was better than Kante? Such people should lose their jobs.
Also, weren't the Everton fans calling Gueye shortly after he arrived "Baby Kante"?
{Ed001's Note - I don't know, I don't remember that nickname.}
05 Sep 2024 00:27:17
The Everton fans I know would never use the name Baby Kante ?.
05 Sep 2024 22:09:06
I think you would enjoy the book Ed. I am currently listening to the audiobook.
It has a chapter on Brighton and Brentford. It is quite well balanced. He does talk about where the data got it completely wrong. He also explains all the flaws and issues with it as well with lots of examples throughout the book.
{Ed001's Note - I will have a look for a copy then, cheers mate.}