Slow starts and fast finishes

In a first look at the PWR a couple weeks ago, I suggested that the superficial similarities to past seasons were deceptive and that this season was different. Since that claim, the Sioux have extended their unbeaten streak to 9, compelling me to hurry up and prove my point before it starts to look even sillier.

At first glance, this season did indeed seem to have one of the slow starts for which Hakstol is becoming famous:

12-31-2005 13-8-1
12-31-2006 10-10-1
12-31-2007 9-7-1

What, then, made me claim the Sioux weren’t following the same pattern? I was looking at the Pairwise rankings. In each of 2005 and 2006, the Sioux would have been #16 or below if the season had ended at some point in January. In 2007, the Sioux never dipped below 9th, a guaranteed NCAA tournament spot. Here are the charts:

2005 PWR2006 PWR2007 PWR
Or in a more interactive format (hover over a datapoint to see the weekend’s results):

2007-08 PWR
2006-07 PWR
2005-06 PWR

Because I had looked at PWR first, I was a little surprised when I dug deeper and found the win-loss table for the first half of this season to be so similar to 2005 and 2006. The similarity in game results but wild difference in PWR seems to come down to who the Sioux played and lost to.

Though the Sioux failed to sweep any opponent in Fall ’07, they also escaped unswept. In contrast, the 2005 and 2006 campaigns featured sweeps of lesser opponents and getting swept by greater opponents. In 2007, the Sioux played tough opponents and managed to stay just above .500 against them, also allowing the Sioux to stay above .500 in TUC and most COP comparisons.

Every Sioux opponent in Fall 2007 is currently a TUC (in fact, every past or future opponent but Alaska-Anchorage and Bemidji State are TUCs, and AA is on the cusp at #26). So, to further try to compare the level of last Fall’s competition to that of ’08 to date, I used KRACH to predict game outcomes. The Sioux should have won 63.5% of their game in Fall 2007 but 73% of their games so far in 2008(*). Though it doesn’t diminish the stark contrast between a 9 game unbeaten streak and last Fall’s sequence of splits, the Sioux have certainly played easier opponents in the former.

FYI — for its remaining series, KRACH predicts a 69.9% winning percentage for the Sioux, though that average is really tugged by its outliers: 52% chance of defeating Denver vs. 88% vs. Bemidji St.

(*) I’m conveniently ignoring that the Sioux splitting against the Fall opponents raises those opponents’ KRACH while their defeating Spring opponents lowers those opponents’ KRACH

2 thoughts on “Slow starts and fast finishes”

  1. Hah. It was an homage to the annual December thread in which someone notes disappointment in Hakstol’s sub-par team.

    It was also written BEFORE the current hubbub or suspension. In light of current events, maybe I should stop trying to make jokes… 😉

Leave a Reply