2016-17 WCHA Week 18, Series 55 Prediction: Minnesota State (1623) at Alaska-Anchorage (1350)

2016-17 Week 18, Series 55: Minnesota State (1623) at Alaska-Anchorage (1350)

ResultProbability
MSU sweeps56.94%
Splits13.91%
MSU 5 points (W, OTW)11.78%
MSU 4 points (W, OTL or OTW, OTW)8.81%
UAA Sweeps1.22%
UAA 5 points (W, OTW)1.01%
UAA 4 points (W, OTL or OTW, OTW)0.73%

I’ve begun randomizing the BELOW inputs as previously discussed.  In this case, Minnesota State could be estimated to be as high as BELOW 1663 or as low as BELOW 1583, while Alaska-Anchorage could be as high as 1390 or as low as 1310.

Note that the system gives a healthy respect to UAA picking off a game from the Mavericks.  From my output (this is new):

Minnesota State won on 1st night by:

shootout  4.16%, 3v3:  3.71%, 5v5:  4.14%, 1 goal: 13.91%, 2 goals: 11.08%, and 3+ goals: 45.62%.

Minnesota State won on 2nd night by:

shootout  4.17%, 3v3:  3.69%, 5v5:  4.15%, 1 goal: 13.87%, 2 goals: 11.13%, and 3+ goals: 46.04%.

Alaska-Anchorage won on 1st night by:

shootout  4.18%, 3v3:  3.71%, 5v5:  4.15%, 1 goal:  5.34%, 2 goals:  0.00%, and 3+ goals:  0.00%.

Alaska-Anchorage won on 2nd night by:

shootout  4.15%, 3v3:  3.71%, 5v5:  4.12%, 1 goal:  4.93%, 2 goals:  0.03%, and 3+ goals:  0.00%.

The average WCHA game is decided by:

shootout  8.33%, 3v3:  7.41%, 5v5:  8.33%, 1 goal: 27.78%, 2 goals: 22.22%, and 3+ goals: 25.93%.

In comparison, this series was decided by::

shootout  8.33%, 3v3:  7.41%, 5v5:  8.28%, 1 goal: 19.03%, 2 goals: 11.12%, and 3+ goals: 45.83%.

You can see that UAA is never expected to win by more than one goal, which makes sense given the disparity between the teams.  Does this account for results like UAA dropping a 3-goal-margin bomb on Bemidji at the end of 2016?  Not really.