Data analysis

AyeAyeAye said:
If I’m deciphering correctly you think input is the boring stuff which is right, although how that applies in the case of football players I don’t know. However it appears we don’t now have the in house experts.

You said it better than me.
Click to expand...

Two different sets of people. Data scientists handle the input and develop the algorithms to process the data. Very mathematical and pretty much exclusively IT skill based. Data analysts take the output from the algorithms and use it internally in two ways. Firstly, existing player performance to tailor individual training and monitor fatigue, and on other players to assess potential recruits to fit the desired playing shape and style.

As I understand it, we've outsourced the former, and kept the latter.
 


If the club needs data scientists, they need look no further than any 11year old kid who plays Fifa. The game is riddled with more real data than people could ever need, and the kids would have little trouble in assembling the greatest squad of players from across the continent for next to no money.
 
Lord of the Dance said:
Does this mean he’s getting the blame for the shite signings this year or that we’re doing away with the Data-based recruitment?

Or has he just been offered a shit load money elsewhere?
Click to expand...

Probably the latter. I think football clubs, particularly Championship ones, underpay by the industry average for a competent data scientist by quite a bit.

People will do it for the entry on their CV then move onto better money I’d imagine.
 
HopeAndGlory said:
I’m really cynical about all this data led stuff. I know there’s evidence that it helps but a pair of eyes does the job better for me.
Click to expand...
How do you know which players to watch though? What are you comparing them with and what is the benchmark?

Back in the day when we only signed predominantly British players it was fine, you could send a couple of scouts out to watch a League One game on a weekend, League Two, maybe a game in Scotland and a few reserve games during the week. They'd maybe get a tip off from a contact to go and watch a certain player. But they can't be in two places at once and you end up signing players on the back of a couple of decent performances or off a hunch and you're limiting yourself to a very small pool of players.

Plus we all have biases and eyes do lie to you, look at the scouts on STID writing players off because they wore gloves. With data you can build a profile of the type of player you need and filter players who fit that. It isn't just picking players with the most goals and assists. There's all sorts of things you can track about your own players in terms of recovery, fatigue and fitness.

But it isn't about replacing one or the other, it's just another tool to help identify potential targets. And it's only intended to present information,what Speakman etc choose to do with it is another matter.
 
Maybe realised when preparing his presentation for the next manager he sorted the candidates in descending order by mistake? Would explain why we appointed Beale?
 
Chewit said:
Nowt wrong with 1-2-3 I was using that at work in 1983


There will be some industrial espionage going on there with mags
Click to expand...
A senior consultant at a big 4 firm is a very junior level. Tongue in cheek, but I reckon he could easily be bought as he won't be rolling in it!
Weren't the data team supposed to be building tools for us to use? I seriously doubt this is the case, but if they did their job well enough wouldn't they make themselves somewhat redundant, at least in terms of building stuff for harvey et al to use?
 
molthemackem said:
How does it go again, "you only get out what you put in"
Click to expand...
Garbage in, garbage out.
Grumpy Old Man said:
Two different sets of people. Data scientists handle the input and develop the algorithms to process the data. Very mathematical and pretty much exclusively IT skill based. Data analysts take the output from the algorithms and use it internally in two ways. Firstly, existing player performance to tailor individual training and monitor fatigue, and on other players to assess potential recruits to fit the desired playing shape and style.

As I understand it, we've outsourced the former, and kept the latter.
Click to expand...

Logon or register to see this image
 
Last edited:
Boring Old Man said:
A senior consultant at a big 4 firm is a very junior level. Tongue in cheek, but I reckon he could easily be bought as he won't be rolling in it!
Weren't the data team supposed to be building tools for us to use? I seriously doubt this is the case, but if they did their job well enough wouldn't they make themselves somewhat redundant, at least in terms of building stuff for harvey et al to use?
Click to expand...

First time I heard that last point… but it would make sense
 

Back
Top