Umm... no, that is not how you interpret data.
The Dunning-Kruger effect.
Data is analysed by the majority of incidents, and not by exceptions. Exceptions are always disregarded.
So by your logic, if a player doesn't score in the majority of his appearances, the games were he did score are exceptions and the games where he didn't are "the majority of incidents" So we can selectively leave out up to 49% of games in an average?
If you were neutral towards James, you would take all his games and goals into consideration.
He is quite unproductive. 3 goals in his first 4 games... and then 3 goals in the next 44 games. I believe that's a very clear difference and perhaps proves that his development hasn't been satisfactory.
Sure. I never argued to the contrary.
The reasons I took it out is because it will not give you a true representation of the type of player he is.
I get it, you don't like James and took out some of games where he scored to make the stats look worse.
His first 4 games would make him seem like the next Ronaldo - at 0.75 goals per game. But we clearly know that is not the case - this is an exception.
It's not an anomaly, it's a low sample size. Your bias really shows when you keep calling a good run "an exception".
If only there was a way to get an average that included the good and the bad...
You only need to look at the vast majority of the games rather than these outliers,
They're outliers to you, because he scored. The point of using an average is to get a ... well, average.
because his first 4 games are massively skewed in his favour.
Skewed in his favour as opposed to ... your bias of thinking he's shit/should be sold?
That fact that a good run happened early on, rather than in the middle or towards to end is irrelevant and can never be discarded in an
average.
His average goals per game for the next 44 is .06. Massive difference.
Wrong, it's not an average when you're selective about what you include. It's only an average goals per games you felt you wanted to include, which is only a useful number for you and means diddly squat to anyone else.
Since James was a youngish player new to the league when we bought him, I think it's fine to look at his last season to get an idea of how productive he currently is:
-James rarely plays the full 90 minutes. This makes looking at his GA per appearance worthless. He has about 7 goals/assists in 1500 minutes this season. Which is one goal contribution every 217 minutes. A bit less than a goal every 3 games worth of minutes.
-James rarely gets a run of games. Even experienced players like Mata, Matic, Telles, Bailly struggle to perform when they haven't played in a while. Let alone younger players like James or Van de Beek. James only had one real run of games from February 'till April. During that run he made 3 goals an had an assist.
You cannot use an unrepresentative stat for a squad player, say goals per appearances, then take out a selection of games because it pleases you (to make James' contribution as small as possible)
What you're doing is the equivalent of excluding January to show how hot the average temperature in a year is.
I think there's a big difference in how we perceive players.
I think he isn't good enough and should be sold for a good fee. You think he has 0.06 goals per game and might as well be replaced by a bucket of cow dung.
Massive difference, agreed.
