The Scuzz Model: Post Mortem
It was a familiar ending for NU fans this past Saturday, as the Houston bowl did not go in our favor. From this fan’s perspective, the loss didn’t hurt nearly as much as some previous games (UCLA ’05 and Missouri ’08 are still the most painful for me)… I can’t tell whether this game was fundamentally less painful or if the ongoing streak has just numbed me (pun intended). I know that the worst I felt on Saturday was during the 3rd quarter, when it was 30-7… maybe somehow the disappointment of the final score was mitigated by the fact that the Cats finally played well on offense in the 4th quarter.
The Scuzz model was downright unconvinced the Cats had a chance in this game. The model liked A&M to win by 15 and assigned 32 confidence points to this game in the ESPN bowl challenge. I’m not trying to assign blame, but if you breakdown NU and A&M’s performence by efficiency, the Wildcat offense did not live up to expectations, while the D performed better than anticipated. I’ve tried to highlight this in the graphic below:
Each individual unit’s efficiency rating (purple and burgandy) is geared to show performance against an average team… since A&M’s offense is above average, and NU’s defense is below average, a much higher efficiency output was expected by the model (light green bars). NU’s defense held A&M’s offense under the model’s expectation and below A&M’s average season output; the NU offense did not rise to the expected level of output, though did outpace what A&M’s defense allows on average (dark green bars).
Now, one might argue that A&M was just clock killing in the 4th, and that lowered the result, but after NU got within 1 score, A&M had one drive that resulted in points – that wouldn’t have much of a negative effect on A&M’s overall efficiency output.
I highlight this as a lot of the post-game commentary I’ve heard focused on the D, rather than the offense’s inability to attack for 3 quarters. As I’ve said on the podcast most of the year, I believe this team wins with offense, not defense.
Speaking of painful bowl losses… I had to feel for Stanford the other night. I felt like I was watching the ’08 Alamo and ’09 Outback bowls all over again, where the Cats’ played for a FG rather than going for the TD. Brutal. Something for all us fans to remember the next time NU’s kicker missed a “chip shot” at a critical moment… this happens in college football all the time (and in the NFL too… coaches are way too reliant on kicking the FG in those situations).
The Scuzz Model is leading the pack in the West Lot Pirates ESPN bowl pool, and has reached the top 15 in the Lake the Posts pool. The model tripped up on both Wisconsin and Stanford, but had either of those gone the other way, another 99th percentile finish could’ve been in order. Instead, the model will probably finish in the low-to-mid 90s (currently 94.7% in the confidence pool, and 95.0% straight-up).
The picks I posted last week only went 2-4, but the model is winning at a 62% clip against the spread:
Scuzz Model Bowls – Straight Up: 23-7
Scuzz Model Bowls – ATS: 15-9
Here’s picks for all the remaining games:
West Virginia +3 over Clemson
Pittsburgh -3.5 over SMU
Arkansas State -1.5 over NIU
Kansas State +8.5 over Arkansas
Alabama +1 over LSU