Is it an either-or choice?
I love it when we play terribly and nick a 1-0 win from our only shot on target.
The results will show we took 3 points. The stats will show we didn’t play very well.
Perfectly possible to recognise both.
4 Likes
Also this forum talks as if all of sudden we’re considering XG for the first time. We were even evaluating it under Ainsworth. We continued with much higher emphasis of importance under Bloomfield. Every well-run club is evaluating the data. The importance is interpreting it effectively/correctly in conjunction with your own eyes.
This forum doesn’t talk. Certainly not with one voice. People have opinions, you might have a different one. Could always move on.
3 Likes
Thanks, @Barneysmith for pointing me towards those highlights. We certainly played them off the park that evening.
1 Like
Or a mightily fine music combo
1 Like
My only issue with xG is that some posters seem to think that it’s incontrovertible evidence on a par with DNA.
Other than that, each to their own.
2 Likes
If the data finds us The Player who will score 30 goals next term, The Player at the back who will prevent 97% of goal attempts being successful and a little box to box midfielder to kick stylish opposition players up in the air all over the pitch.I’m in.
4 Likes
Fair enough.
My view is that xG js a good tool for conducting a dispassionate analysis of performances, strengths and weaknesses.
Although I imagine relatively basic numbers like that aren’t really of much practical use to the team and management and are a bit of a red herring in regards to how clubs might use data.
My assumption is we’ll be using data predominantly as a recruitment screening tool to help identify undervalued targets that outperform in certain characteristics.
I should think if that’s done well then it’s possible to massively over achieve.
My concern is that it must be very easy in football to either a) record data poorly or b) analyse it badly or c) look at completely the wrong data in the first place (perhaps all three).
Games like baseball and cricket can be reduced down to relatively few actions and measurable qualities. Invasion sports must surely be much harder to breakdown in that way. And I imagine it’s incredibly difficult to adjust for the relative strength of each country and division - if there’s a centre back in some far flung destination that wins 100% of their headers is anyone controlling for the height of the strikers in that division, the direction and quality of the balls faced and so on?
2 Likes
My feeling is that Dodds needs some confidence in his players. Could it be that coming from a massive club like Sunderland to “little” Wycombe, Dodds has felt,at least subconsciously, that to get a result Wycombe have to be ultra defensive. Once he has the confidence that 11 v 11 we are as good as anyone in this league, then we might see some more attacking football.
It still amuses me, in a way, that people seem to think we’re only just now “using Data” when it comes to recruitment.
Like somehow scouts and head of recruitment teams weren’t going out, looking at players in games and training and assigning factual information against those players to help make informed decisions.
There’s only two main differences between “data Dans” approach and the old system: Scale and Personnel. And both of those two things are linked.
Previously, each club would have their own recruitment team with a finite number of scouts on the books, meaning they could only see a limited number of players and rank them against their own personal metrics.
With data dans approach there is the option of an almost infinite number of scouts all pouring information into one centralised database. Scouting becomes a freelance enterprise and rather than massive scouting teams across the globe you have a smaller specialised in house team that can accurately pin point specific players to go a see.
Rather than casting a net and hoping to catch a big fish - they can put a rod in the water with food for the specific fish they want to catch.
I don’t want to really boil it down to a game, but you really need to have a look at the idea that is under the hood of Football Manager.
For pretty much every team across the entire professional footballing world they have information and data collected by a huge team of scouts and analysts so that the game and the players are as close to real world as possible. For the longest time (they might do still) they relied on fans and volunteers to do it.
You take that basic idea turn it professional and have trained scouts and data analysts using a real world model and you have things like Wyscout or whatever Dans company is called.
You still have the problems you mention on how good the data is but with more people pouring more data into the system it (should) only get better and more accurate.
4 Likes
How long would it take for an infinite amount of scouts with infinite data to sign a couple of Danes nobody had heard of?
12 Likes
If a Dane makes the first XI in a forest, and nobody is around to see it, did he play at all?
11 Likes
Sounds brilliant. And any doubters will surely be won round by looking at the roaring success of our January 2025 transfer window.
5 Likes
Who ever claimed every transfer window will be a roaring success?
Sport at an elite level, and even league one level is about marginal gains. If it makes transfers 1,2,3,4% better in any category upto and including better players, more options on players, or even just a cheaper process, they are all positives and may give an advantage.
What I am 100% sure of is the old system wasn’t perfect and neither will this one be.
1 Like
I don’t know.
But if an infinite amount of monkeys with an infinite amount of time will write Shakespeares Hamlet, I can only hope Data Dan will find his own Prince of Denmark. Eventually.
Let’s not undersell how bad January was. In terms of value for money it must be our worst ever. We broke our transfer record twice on players who aren’t good enough to play.
19 Likes
Dan’s just bought a cottage in Little Kingshill as the data said it was a Hamlet in Buckinghamshire.
2 Likes
Now, to be fair, there are two giant asterisk against those two. One (apparently) had a serious but not life threatening illness when he came to us that affected him massively and the other got injured so I don’t think fair to write them off completely.
Sure, it doesn’t look great on either of them but it’s rather peak football fan to slag them off when it might be out of their, or the clubs hands, if they couldn’t physically play for us.
Personally, as with Dodds Tbf, I’m giving them an actual proper preseason and about 10 games before I start jumping up and down.
2 Likes
Alas, poor Kingshill!
If the ‘data Dan’ approach is to screen and identify targets using one or more databases and then go and watch those players, surely that’s quite different to watching games and then looking at the numbers behind the players the scouting team like the look of?
One is data first. The other is data second.
I could be completely wrong about the way we’re applying data to player recruitment of course. As far as I’m aware the club have given very little detail at all.