This thread is telling. Not just because of
uTurn 's analysis but the reactions to it.
Some stats, on their own, tell no story. I'm not particularly fond of xG for example. A team which takes a relatively early lead in a game, especially away from home, and holds that lead for a long time is likely to have a lower xG than it might have simply because that team doesn't need another goal but the opponent does. So the final xG "result" is skewed by necessity. Also, a cross which flashes across the six yard box which someone just fails to get a touch to is classified with an xG of 0.00, which is ridiculous.
However, when you take all this data in the round you can get a reasonable picture of where you are; and we are as lousy as the stats say. Anyone with any kind of objectivity can see that, and unless performances improve we're going to be flirting with relegation if not wedded to it.
Equally, Charlton's early season promise from those numbers looks like it's probably a mirage with points attached, so we'll see in due course if they also sink to where they seem to belong.
But some are suggesting we ignore these dry numerical warnings, plus the evidence of our own eyes because of what? Misty water colour memories?
The quote "Lies, damned lies and statistics" suggests that statistics can be used to frame an argument.
Its interesting to read this thread as I think there is a valid point being made in amongst what
RandomBlade has said, but also misses some key points.
The main point i would make is that we shouldn't take stats and data in isolation, as
RandomBlade alluded to, our eyes will tell us what stats can't always do.
In football and at United we've used data for years, probably decades, its just that now we have far more data than we've ever had and whilst it is useful, we have to be careful in how we use it. Adkins was using the GPS vests before Wilder came in and he was the first real user of data / stats in his interviews. Wilder v1 would often talk around it but didn't explicitly mention data. Wilder v2 mentioned data a lot more and v3 its often referenced.
The James Bord experiment was less about data being the problem, but the use of data in recruitment without having the football experts in control.
I think even in game or post game, stats become a part of the story, they should never be the story and i agree with your xG point, i think its less relevant than many want it to be. If you want to use xG:
United have scored 9 goals from roughly 9.8 xG. So we're not massively underperforming our attacking chances.
Conversely, we've conceded 11 from roughly 12.7 xGA. So the goals conceded aren't some freak statistical bad luck either.
The stats from 8 games so far can't be taken in isolation, with an overall snapshot
8 games | 2W 3D 3L | 9 points | Scored 9 Conceded 11
Summary
Performance: mixed
Chance creation: below where you'd want it
Defensive control: below where you'd want it
Efficiency: surprisingly good in several games
Quality of opposition: reasonably demanding (5 top 10 sides)
Main concern: the underlying numbers say this isn't sustainable if the chance creation/defensive balance doesn't improve. But you don't need numbers to see that when we have eyes.
The numbers say we're a limited, reactive side.
The reality is that we have already played some good opposition and we have been competitive enough in some games to collect results.
Norwich was the only game we've lost by 2 goals, both Stoke and Wolves we conceded the winning goal late in the game.
Its not a good look, we're not doing great