Monday, January 19, 2009

PairWise Update and Mailbag


Despite a disappointing finish to Saturday's game against CC, UW took a step forward in the PWR for the weekend as a whole. UW is up to 17th in the latest iteration of the rankings, and still holds a slim but significant comparison win over CC.

At this point, UW's RPI is still very fluid, as are many teams around them in the rankings. Simply winning games is the single biggest thing UW needs to do, regardless of the opponent.

For example, UW is currently losing the comparison with Alaska. Fairbanks has a slim edge on UW in the Ratings Percentage Index component (RPI), beating the Badgers .5286 to .5280. A win over anyone with a pulse puts UW up on UAF there. However, UW's also winning the Common Opponent criteria (COp) - yet UAF has not yet played NMU, thus UW's 0-2 record against the Wildcats is not factored into that comparison. When it does, UW will take a hit.

This kind of scenario repeats itself all throughout the PWR, which makes the week to week fluctuations so large. There are also several teams that have yet to play 10 games against Teams Under Consideration (TUC), but will pass that threshold soon. This, too, will shake up the rankings considerably.


I've recieved a few questions on potential impacts to UW's PWR:
If N. Michigan goes on a tear the rest of the season that would help our RPI correct?
The short answer is yes, but the total answer is a bit more complex.

The RPI is a rather simple calculation (and an inherently flawed one, if you ask me), that multiplies a team's winning percentage against the winning percentage of their opponents, and then again against the winning percentage of their opponents' opponents. Each of those three factors is then weighted. Basically, the formula is trying to see how good you're doing, how tough your opponents are, and what kind of schedule your opponents have played.

The current formula is Pct-OPct-OOPct 25-21-54. Previous iterations have had different formulas, such as 25-50-25. There are numerous problems with this, but we can discuss those later.

The net result is that NMU's record will factor into the least important part of the RPI, the 21% weighted factor for their winning percentage. For them to improve, they will certainly increase UW's RPI, but not by a whole lot.

The more interesting question is what that does to the PWR. So long as NMU wins, they can't hurt UW's PWR standing. However, if they win enough to enter the top 25 of the RPI and become a TUC, then UW will be saddled with an 0-2 record against all other teams in the TUC category. NMU would have to go on quite a tear for that to happen, as they currently sit at #42 in the RPI.
Also, it would be to UW's be interest if UAA played well and became a TUC by season's end, correct?
Absolutely. Adding UW's 4-0 record against the Seawolves would be an excellent way to boost UW's TUC comparisons. UAA currently sits at #32 in the RPI with a .4927 - #25 St. Lawrence is sitting at .5182. That's a lot of ground to make up, but it can be done.

Any other questions? Leave them in the comments, and I'll try to give them a good answer.