We start off this week with a look at Eric T's new way of looking at possession rates: balance corsi. Here's how he explains the new metric:
I took the 100 closest comparisons in usage (50 above and 50 below) and averaged their results to see what I should expect from someone with that role.
With "role" meaning zone start ratio. Tim went through and performed the analysis for every player in the league and then did it again for relative corsi rates, which technically corrects for strength of team effects (the difference between on-ice and off-ice corsi rates for each player). Fortunately, he supplies a spreadsheet of his data at the end of the post, which has led to some team specific analysis in both Calgary and Toronto.
Balance corsi is an effective way to compare players to a "league average" possession performance in the context of zone start usage. That said, one must also take into account other contextual factors when looking at corsi, including quality of teammates and the difficulty of the opposition.
Speaking of player quality, Scott Reynolds of the Copper and Blue has been comparing the leagues worst contracts in an on-going series recently. His latest battle is between the Brian Campbell albatross ($7.14 million per year) and Scott Gomez ($7.357 million).
It's a tough contest given the length and expense of both deals, but Brian Campbell currently leads the CnB poll with 65% of the votes.
Finally, this talk at TED by Oxford mathematician Peter Donnelly shows how unintuitive things like stats and probability are for humans. To the degree that, even when we try to use and understand things with stats, we end up screwing it up.
Donnally doesn't specifically talk about hockey, but his points are relevant to evidence-based analysis in sports nonetheless. He shows now only how stats can be misunderstood or abused, but also why fans may generally not initially take to such methods.
Finally, Dave Shapiro of Blueshirt Banter celebrates the season of unheralded rookie Mike Sauer. The 24-year old defender won't be mistaken for a Calder candidate or anything, but it seems he's done a capable enough job as a middle rotation defenseman for the Rangers this year. Look out for that team high +18 rating howeverit probably has a lot to do with the unsustainable .946 ES SV% behind him this year.