Data driven managerial appointments

according to their CEO , about half way in article , “The league table can lie,” Acheson asserts

That settles the debate. If Big Data fella has ran the numbers and the simulations 10,000 times and comes up with that , then that's us told. someone get Woody on the phone
 
The clue is they scanned 10,000 coaches we can’t do that under our current policy ( as it stands)
 
Data is a funny thing. In my professional experience I have found that people that have created a data model will be too enthusiastic about it. They will have a narrow focus and won't consider any variables they haven't already thought of. That means it can be good for a specific question but it could also be terrible.

With something like a manager attributing a style of play, like that Edwards one above, doesn't take into account that things are different for different matches, which players are available etc. You only need to look at Wilder's tenure with us. We have played very differently at times, some games very slow patient build-up and working an overload to create chances and other games we have played long ball. Look at 2 players playing the same position (last season because it was consistent), Dijksteel at RCB and McNair at LCB. They played very differently but Wilder would've loved a left footed version of Dijksteel and the same for the LWB. Then look at the difference between Mowatt and Tav. Do you adjust what you believe is the manager's style because it is rare that a manager gets to play a team that fully suits exactly the way they want to play.

That article talks about the model. How would Wilder be judged? Did he improve us? We are way below where we were when he took over but do you give more weighting to his recent seasons or do you give his Alfreton success equal weighting? Warnock improved us after Woodgate but does that mean he was a good manager or would the stats just benefit him because Woodgate was so poor so his baseline was set lower? You basically have to know a teams true position with the average manager to know whether a manager has over/under performed but there is no such thing so it's all based on assumptions.
 
Back
Top