Published by Jim Dahl on 01 Feb 2011 at 03:24 pm
…but could also fall to #7.
I predicted back in December that I would next revisit the Pairwise Ratings (PWR) in February, and felt particularly compelled this week to follow through with that promise.
Predicting UND’s PWR as of Feb. 8
Bye week is always an interesting week to do a 1-week prediction, because it reveals how much our PWR can change without even playing. UND will have the same set of wins and losses a week from now as today, only PWR’s interpretation of the strength of those wins and losses will change.
UND is currently #3 in the PWR (actually, tied for #2 with Denver, who wins the RPI tie-breaker).
After next weekend (I projected out to next Tuesday to include the first round of the Beanpot), UND has over a 70% chance of being either #2 or #3.
#1 is actually possible (7%) as is #7 (.03%).
What to cheer for
You can follow along on the UND PWR Details table
The biggest games for UND are in the WCHA.
- Minnesota beating Minn.-Duluth. Even a split would be enough to give UND the TUC point and the overall comparison, letting UND climb one spot.
- CC beating Denver. This one is defensive — if Denver swept, they would get the TUC point from UND and the overall comparison, dropping UND a spot. A split should be enough to prevent that.
- BU winning. It’s difficult to decipher when a team plays two different opponents, and it’s hard to figure out why Lowell matters, but it seems to a little. BU beating BC is important as defensive measure — a BC win over BU would be enough for BC to take the RPI point from UND and the overall comparison, dropping UND a spot.
No one has actually asked yet, but someone is probably wondering…
How can UND get the #1 spot, I thought UND couldn’t take comparison with Yale?
I don’t think UND can take the comparison with Yale this week, overcoming Yale’s advantage in both RPI and TUC seems far-fetched. However, if UND took the comparison with Minn.-Duluth this week (see above), it would only be losing one comparison. If Yale lost a comparison to someone else this week, the two would be tied. UND would then need to surpass Yale in RPI to take the tie-breaker. Since UND is playing, the only way for that happen is for Yale to lose (probably both).
I just stumbled on this, how is PWR being predicted?
I simulate the outcomes of the remaining games in the season a million or so times, using KRACH to predict the likelihood of each team winning in each iteration. That’s enough simulations to ensure that even the most unlikely possibilities occur at least once. For each simulated season I calculate what PWR would result from that set of outcomes. I then use the aggregate results of those simulations to assign a likelihood to a particular outcome, that is, if UND finishes #3 in PWR in 370,000 of 1,000,000 simulations, I say that UND has about a 37% chance of finishing #3.
Current PWR rankings (SiouxSports.com)
Explanation of how PWR mimics NCAA tournament selection (CollegeHockeyNews.com)
USCHO Bracketology Blog (USCHO.com)