A Blowout Upset

When the line for Saturday’s Utah-Washington game at Century Link Field came out last week, Husky fans were surprised to see their team as a slight underdog despite a superior record (5-4 vs. 4-5) and home-field advantage. Based on the numbers, the line made perfect sense.

In one sense, Washington and Utah are mirror images. The Huskies’ record dramatically overstates the team’s performance, while the Utes have played better than their sub-.500 record. Jeff Sagarin’s rankings for USA Today allow us to make this comparison because Sagarin includes two different rankings: One, ELO Chess, that takes into account only opponent, location and win/loss (the BCS uses this version to discourage teams from running up the score) and another (Predictor) that uses margin of victory rather than win/loss (and is more accurate).

Washington has one of the largest differences between these two ratings of any major-conference team. Here are the 10 teams whose ELO Chess score most exceeds their Predictor score:

School              ELO  Rk   PRED  Rk      Diff
--------------------------------------------------
UMass              55.1 156   40.7 201   -14.4 -45
Miami (Ohio)       65.3  87   54.3 149   -11.0 -62
Central Michigan   64.6  92   54.5 146   -10.1 -54
Notre Dame        101.5   1   91.6   5    -9.9  -4
Eastern Michigan   58.3 141   49.0 172    -9.3 -31
Ohio State         90.5  11   81.5  26    -9.0 -15
Buffalo            63.0 107   54.1 150    -8.9 -43
Wake Forest        66.3  81   57.6 130    -8.8 -49
Tulane             58.1 142   49.4 170    -8.7 -28
Washington         85.5  15   76.8  41    -8.7 -26

Based on wins over two top-10 teams and a number of “good” losses, the Huskies are an elite team when just wins and losses are taken into account. However, blowout losses at LSU, Oregon and Arizona and few large margins of victory means Washington’s point differential is much more average considering the schedule. Not bad, mind you–the Huskies have improved from last year, when they ranked 50th–but not as good as the record and a strength of schedule ranked second in the country would indicate.

By contrast, Utah has played better than its record, getting blown out just one time all year and pounding Cal (by 22) and Washington State (by 43) at home. The Utes aren’t quite in the top 10 of the opposite ranking, but they are close, ranking 13th:

School              ELO  Rk   PRED  Rk       Diff
---------------------------------------------------
BYU                72.3  51   83.8  18    11.5   33
Florida State      76.4  34   87.3  12    10.9   22
Fresno State       71.0  57   79.9  27     8.9   30
North Carolina     68.3  73   77.1  38     8.7   35
Utah State         74.9  40   83.2  21     8.3   19
Arizona State      75.9  36   84.2  17     8.3   19
Alabama            93.1   9  100.4   1     7.3    8
Boise State        72.3  52   79.4  29     7.1   23
Georgia Tech       66.7  80   72.8  56     6.1   24
Clemson            77.5  31   83.6  19     6.0   12

Utah               72.7  50   78.1  33     5.4   17

Factoring in home-field, Predictor showed Utah as a slight favorite entering the game–which makes it all the more remarkable that Washington not just won but dominated, 34-15. The Huskies outgained the Utes 437-188 and passed for five times as much yardage (277-55). Given the level of competition, the same method I’ve used to evaluate games in the past–actual differential minus expected differential based on opponent and location–shows this as Washington’s best performance of the season and one of the top three of the Steve Sarkisian era.

All three share the same basic characteristics–games that should have been nearly even, but were won by the Huskies in lopsided fashion. The second should be familiar–last year’s Utah game. Washington was actually a 10-point underdog entering that game, per OddsShark.com, though this method shows the Utes a lot worse because they lost starting quarterback Jordan Wynn to injury during the game and struggled to score thereafter. Based on season-long performance, Utah and Washington were essentially similar, but the Huskies rode turnovers and a big second half from Chris Polk to a 31-14 road win.

The other game was the finale of the 2009 season against No. 19 Cal. Again, based on the whole season, the Huskies and Golden Bears were basically even. This method shows Washington as 3.5 points better by virtue of home field; they were actually favored by six because of Cal’s poor finish to the season, which continued during a 42-10 Husky win, tied for the biggest margin against a Pac-12 foe in the last decade.

If you want to put the 2010 Holiday Bowl–which ranks fourth–ahead of those regular-season games, I won’t blame you. By any measure, though, Saturday’s win was one of the most complete the Huskies have had in a long time. Coming just three weeks after we were musing about a lowpoint of the Sark era, that’s quite an encouraging change.

Advertisements