Quote:
Originally Posted by Mathgod
5 on 5 xGF% numbers through 19 games. I think the numbers pretty much speak for themselves.
|
Yup, if you don't mind numbers lying through their teeth by pretending to be significant.
The data aren't accompanied by any information about linemates, quality of opposition, usage, score effects, or any other kind of context.
A player who is put on the ice in a defensive role with fourth-line plugs is not going to have the offensive numbers that he would have in a top-six role. (Yes, I am talking about Monahan.) But you can tell something reasonably meaningful about his defensive contribution by looking at xGA and adjusting it for time on ice, because quality of opposition varies from shift to shift a lot more than quality of linemates.
So let's look at the numbers you gave, and zero in on xGA/60:
Code:
Blake Coleman 1.79
Oliver Kylington 1.82
Matthew Tkachuk 1.88
Nikita Zadorov 1.96
Elias Lindholm 1.97
Christopher Tanev 2.06
Andrew Mangiapane 2.12
Sean Monahan 2.13
Johnny Gaudreau 2.14
Mikael Backlund 2.17
Milan Lucic 2.17
Erik Gudbranson 2.22
Trevor Lewis 2.24
Rasmus Andersson 2.30
Noah Hanifin 2.34
Tyler Pitlick 2.35
Dillon Dube 2.75
Defensively, Monahan is in the middle of a pretty tight pack. His xGF/60 (2.03) is cratered partly because he's a month behind the rest of the team in getting his game together, but also because he has been given fourth-line duties with wingers such as Lewis (1.65) and Lucic (1.93).
The only really bad xGA/60 in the lot belongs to Dube, but he looks pretty good on xGF% (53.86) chiefly because, though not contributing so much offensively himself, he got a lot of minutes playing beside a red-hot Mangiapane. Since those two have been split up, his defensive shortcomings have become much more obvious, because there's no glittery offence from a linemate to mask them.
Your numbers that ‘speak for themselves’ would lead an incautious observer to believe that Dube has been clearly better than Monahan, where in fact Dube has been the weakest link on the team defensively and not a big offensive contributor.
In fact, numbers never speak for themselves. They require analysis, and the message they deliver depends entirely on who is choosing the numbers and what information that person chooses to leave out. Remember the remark that Mark Twain attributed to Disraeli, ‘There are three kinds of lies: lies, damned lies, and statistics.’