The numbers aren't complete .. but they are dependable and the extent to which they're not dependable is quantified (ie we know the margins/variation over time)
The alternative is putting the prime colours on your face and face planting it onto the canvas and hoping its a picture
Dependable until you start to pull at a thread a little.
How's his record vs the top half of the PL?
How's his record per 90 if you remove the goals in cup completions and look at PL only, the one competition think it's fair to say is what we care about most.
The goals against Prague and Norwich and Tolouse are lovely and all that, but I'd swap them out for even 1 in those big important games vs City, United or Arsenal.
How dependable is a goals per 90 stay when it doesn't take in to consideration who the minutes and goals are against?
Goals per 90 doesn't apply context to the degree of difficulty some 90s possess vs others.
Darwin's league goals have come vs
Newcastle x2
West Ham
Forest x2
Bournemouth x2
Burnley X2
Brentford
Sheff Utd
Now you need to beat the shite to win the league. These goals have merit, I'm not arguing they don't.
But I don't subscribe to the opinion that goals from a player vs that calibre of them are worth the same as a player who is able to score against the top 6 clubs, let alone the top 10 (which only includes WHU and Newcastle from that list).
The gp90 and perhaps your own opinion seems to be that one goal is worth the same amount as any other, and empirically that's true.
The position I'm coming from is, are they really?
Its harder to score vs the top teams, that's why they tend to concede fewer goals. Therefore extrapolating that his gp90 are strong and therefore that can be projected forward to *every* 90 he plays is disingenuous isn't it?
A 90 vs Burnley, a team he loves playing seemingly, is not the same as a 90 vs Arsenal, or Madrid or City.
Therefore, as a statistic, what's it's
true value?
If I was presented with his gp90 vs various tiers of teams then id be more interested to hear about it.
It's easier to project down, if you're good at scoring against good teams then it stands to reason you'll be good against bad teams.
I don't agree it projects in the opposite direction however. Which using gp90 as a means of assessment seems to, as all 90s are created equal?