Quote:
Originally Posted by Bingo
While true and factual I think the comparison is better served with expected metrics. If a trio is stoned repeatedly or has had subpar goaltending when on the ice you may miss a better combination.
In this case though it still holds.
Ryan version - 269 minutes xGF% 50.81%
Bennett version - 37 minutes xGF% 50.68%
Pretty tight, but the Ryan version with the edge.
The big difference then, as expected is puck luck with the Ryan version posting 1.05 PDO compared to the Bennett version 0.925. That's a huge discrepancy.
Bennett version has had terrible goaltending at .783 save percentage when together.
All this says is that they have two options, and Bennett has done quite well at center. If they think Ryan would be better anchoring his own line at center on the fourth it's a pretty good option.
|
xGF% is a great stat and does a great job of attempting to extrapolate an expected goal rate. That we agree on. However, for me the delineation of the quality of shot is still a little basic and doesn’t fully represent the true quality of a shot.
I think xGF% is good to use when lines have a small sample size but when there is a large, say 269min over various goaltenders in all kinds of different situations and games, you might as well use actual counting stats as I believe they tend to balance out the inadequacies of advanced stats.
What you’d need to validate the numbers for Lucic-Bennett-Dube is 200 or so more minutes of playing time together. If it was the beginning of the year, I’d be all over that. Given there are 19 games left in the year, it makes perfect logical sense to go with Lucic-Ryan-Dube for the stretch run and keep Bennett as an option for the 3rd line based on performance.