Re: NHL 2012: The Chase for the Cup
Based on Bradley-Terry rankings (i.e., KRACH) found
here, the updated (after 4/12) most likely outcomes are:
West: VAN in 7 (20.2%), STL in 6 (22.2%), PHX in 6 (18.7%), NSH in 5 (22.1%)
East: NYR in 5 (28.8%), BOS in 5 (22.8%), NJD in 6 (18.5%), PHI in 5 (20.1%)
Overall winning probabilities:
West
VAN 51.3% - LAK 48.7%
STL 58.0% - SJS 42.0%
PHX 65.2% - CHI 34.8%
NSH 73.7% - DET 26.3%
East
NYR 88.2% - OTT 11.8%
BOS 75.2% - WAS 24.8%
FLA 39.1% - NJD 60.9%
PIT 31.0% - PHI 69.0%
As always, KRACH is not a perfect predictive system. In fact, it's not predictive at all; it's retrodictive. Home-ice advantage, momentum, and the like can not be accounted for reliably, so they are ignored.
Notes: Phoenix's win last night made their series the first to have its favorite change. I was curious to see if it were possible that the most likely outcome of a series could be different than the series' overall favorite (e.g., VAN in 7 is the most likely outcome, but LAK is more likely to win the series). It looks like, at least for the VAN-LAK series, if Vancouver's KRACH was just a little lower (or LA's a little higher), such would be the case. I imagine it could only occur if the team more likely to win each individual game trailed the series by one game. Anyone have some other potential mathematical insight? (yes, I know most of you were told there would be no math)