Blue Blooded
Most people rejected his message
First of all, I can pick out minuscule samples where DeAngelo has dominated top guys (like the Kane example) and where Pionk has been crushed. In fact that is way more common than the other way around. E.G. DeAngelo has been a +19 RelCF% against Panarin this season and +15 against Barkov, Barzal and Giroux. In these microsamples you are going to see every outcome under the sun, what matters is the aggregate.I like TDA and don't take this as a knock on him. He was matched against Kane, Strome and Caligula against Chicago and did well.
Against Boston:
View attachment 178721
Against Hurricanes:
View attachment 178723
Yeah, its mind-boggling how well they did above, isn't it?But it doesn't matter who you play against right? A math professor in Vancouver said so (with total disregard of a ton of relevant factor speaking in the other direction). Its only a coincidence that TDA's CF% without the top opposition is 5x as good against Chicago as with them almost twice as good as against Carolina? Right? Purely a coincidence?
Secondly, nobody is claiming that "who you are playing against doesn't matter". Nobody is disputing that in the specific minutes you play against top competition, such as the Bergeron line, you will have worse metrics than when you play weak competition - that is a ridiculous strawman.
What has been said is that the impact of QoC on your overall on-ice metrics is significantly lower than the impact of QoT and is in fact almost, though not entirely, negligible. This is because on the whole your matchups aren't as specific as one might think. My view is that unless you do multivariate projection models that put a specific number to weight the QoC on relation to other metrics you might as well disregard it on the aggregate.
It has also been observed that players who have better metrics against poor competition usually have better metrics against top competition. I did some work on this way back in the summer of 2013 when most of this board was adamant that Girardi was a legit top-pairing defenceman and that Strålman, seen as a 3rd pairing guy, never could handle the top guys like he could. I dug into the data and showed that Strålman was significantly better than Girardi against every tier of competition, because even though he was on lower pairings he still had accumulated a ton of minutes against top competition - just like every defenceman. In fact in the specific case of Girardi it turned out that he was comparatively worse against top competition than against weak - funny that. Most players' competition curve looks the roughly the same, though Strålman did have a flatter curve and Girardi a steeper one. Unfortunately I can't replicate the work now because hockeyanalysis.com is no longer around and they were the only ones to present the data in a manner that such studies were feasible.
I also had a post earlier this year, which was a response to one of your posts, where I clearly showed that Skjei-DeAngelo had been used as our primary defensive pair for a stretch and that they performed significantly better than Staal-Pionk. A post which you conveniently ignored and never responded to.
Last edited: