FanPost

SPAM Week 5 Recap, Projections Update, Week 6 Predictions

If you're not familiar with SPAM, click here for a brief explanation and the pre-season predictions.

You can find SPAM's full team record analysis for all 130 division 1 teams here, updated after week 5. Note: This link will be replaced after week 6 is done, so if you're coming back to this later, it'll be current for that week.

PRIOR SPAM UPDATES:

HOW DID WE DO?

The law of averages holds that, basically, everything eventually returns to level. If the coin is fair and it comes up heads 10 times in a row, it'll eventually come up tails enough times to offset that and get back to 50/50. That's why I run one million simulations of each team for SPAM. Even at 100,000 simulations, there is enough variation in randomness and trend that results are inconsistent.

This principle was on display this week. Last week SPAM went just 13-18 in Planned Sick Days' pick 'em game on Yahoo, but this week SPAM went 19-11 for a whopping 63.3% accuracy rating. It's not just me, though. Other mathematically-based predictive rating systems also had great weeks. I saw Bill Connelly on Twitter talking about his S&P+ analytics doing an absurd 80%+ (his math is much better than mine). So, there's a "there" there, somewhere.

SPAM went 6-1 in the Big 10 again, whiffing only on the Minnesota game. I also had commented that -11.3 line SPAM calculated for Wisconsin seemed "way too tight." Hah! I should have known. October Northwestern creeps closer.

SPAM is now 43-8 picking Big 10 games straight up, for an 84.3% accuracy rating. It's going to get harder, since there are fewer games and missing just one can drag my average down.

SPAM also predicted 4 upsets this week, including Baylor over Iowa State and Toledo over BYU. The other two picks missed (North Texas and Boston College). Concerning Iowa State, I wrote this last week:

I look at Iowa State's schedule and don't see a game they couldn't win except perhaps @Oklahoma, but SPAM is less rosy and has the Cyclones projected at 6-6 (26.8%) or 7-5 (23.8%). 5-7 isn't out of the picture (19.3%). Considering the Cyclones are already 2-1, that means only 4 wins left in the season? You'd think they'd beat TCU and Kansas, and I like their odds against Texas, Oklahoma State, Kansas State, and Texas Tech, but the math tells a different story.

Math can be misleading, but it doesn't lie.

IOWA SEASON PROJECTION UPDATE

Here's your updated win percentages.

Opponent Margin Win%
Miami (OH) WIN (1-0) WIN (1-0)
Rutgers WIN (2-0) WIN (2-0)
@Iowa State WIN (3-0) WIN (3-0)
Md. Tn. State WIN (4-0) WIN (4-0)
@Michigan -1.8 (last week: -1.5) 42.8% (last week: 43.5%)
Penn State 1.9 (last week: 2.2) 57.5% (last week: 58.0%)
Purdue 12.7 (last week: 11.3) 78.6% (last week: 76.3%)
@Northwestern 6.3 (last week: 5.0) 67.0% (last week: 64.2%)
@Wisconsin -1.3 (last week: -2.5) 44.0% (last week: 41.2%)
Minnesota 11.0 (last week: 9.9) 75.7% (last week 73.9%)
Illinois 15.9 (last week: 14.9) 83.7% (last week: 82.1%)
@Nebraska 6.6 (last week: 4.4) 67.5% (last week: 63.0%)

The numbers are basically settled and they tell us exactly what we already knew: Michigan, Penn State, and Wisconsin are hard games. Northwestern and Nebraska are tricky road games. Iowa should breeze past Purdue, Minnesota, and Northwestern. Michigan's tune-up vs. Rutgers widened its margin ever so slightly, while Northwestern's throttling of Wisconsin's offense dragged the Badgers back to earth. Nobody else did anything of note.

I am a little surprised that Nebraska's line moved so much. OSU's offense is good, so Nebraska should not have been heavily penalized for losing. What this says is that even considering how good's OSU offense is, the Cornhuskers' defensive performance was just unforgivably awful. And if you watched Nebraska Twitter last night, they agree: it was. Ohio State's players looked 6 inches taller and 30 pounds heavier across the field. Their domination looked utterly effortless. It was men vs. boys out there. Time was when Nebraska was on the other side of that equation. Was the Mike Riley hole really this deep?

Both Michigan and Wisconsin look beatable. Penn State was the best looking of those teams this week, but the margin in all three of these games is essentially home field advantage. If Michigan and Wisconsin were playing in Iowa City, Iowa would be a slight favorite. Likewise if Iowa was playing Penn State on the road, PSU would be favored. We're about to learn a lot.

UPDATED SEASON RECORD PROJECTIONS

Record Freq.
0-12 n/a
1-11 n/a
2-10 n/a
3-9 n/a
4-8 n/a (last week: 0.1%)
5-7 0.2% (last week: 0.9%)
6-6 1.8% (last week:4.1%)
7-5 7.8% (last week: 12.2%)
8-4 19.5% (last week: 23.3%)
9-3 29.6% (last week: 28.3%)
10-2 26.1% (last week: 20.9%)
11-1 12.5% (last week: 8.7%)
12-0 2.4% (last week: 1.5%)

The numbers are moving in the right direction. Standing at 4-0 with the nation's #3 scoring defense (the top 4 are Wisconsin, Iowa, Penn State, and Ohio State!), and an unusually potent (for Iowa) offense, the probabilities are creeping into those upper echelons. A bowl bid is all but a certainty and 8-4 is looking like the floor now. We also see the odds of 6-6 and 7-5 seasons vanishing and 10-2 and 11-1 ticking up. If Iowa can win in Ann Arbor, you'll see these jump again by quite a bit.

DIVISION CHAMPIONSHIP ODDS

Big 10 West
  1. Wisconsin (47.1%) (was 40.7%)
  2. Iowa (30.3%) (was 28.9%)
  3. Minnesota (11.0%) (was 8.9%)
  4. Nebraska (4.2%) (was 6.8%)
  5. Purdue (0.7%) (was 5.5%)
  6. Northwestern (0.3%) (was 1.1%)
  7. Illinois (0.2%) (was 0.7%)

Big 10 East
  1. Ohio State 65.1(%) (was 49.8%)
  2. Penn State (19.3%) (was 22.3%)
  3. Michigan State (6.1%) (was 9.5%)
  4. Michigan (3.2% (was 5.5%)
  5. Maryland (0.3%) (was 5.8%)
  6. Indiana (0.1%) (was 0.7%)
  7. Rutgers (n/a) (was 0.05%)
Nothing interesting here other than perhaps that Maryland and Michigan has flipped spots in the East (reshuffling the deck chairs on the Titanic, and Ohio State is the iceberg). The West is a two-horse race between Wisconsin and Iowa until somebody else beats them. Wisconsin has a win over a Big 10 West opponent in Northwestern, which basically kills Northwestern's chances. This moved some probability in Minnesota's favor, since they also beat a Big 10 West opponent. The fact that Iowa has no Big 10 West wins yet and is still projected as #2 is a testament to the team's performance. Nebraska is hanging on there for basically the same reason.

LOOKING AHEAD IN THE B1G
  • Nebraska (-3.2) over Northwestern
  • Maryland (-6.8) over Rutgers
  • Michigan (-1.8) over Iowa
  • Penn State (-14.3) over Purdue
  • Wisconsin (-30.2) over Kent State
  • Minnesota (-8.1) over Illinois
  • Ohio State (-11.0) over Michigan State
IOWA STATE

As discussed above, SPAM predicted a Baylor upset over Iowa State and it happened. The Cyclones are still projected at 6-6. The Cyclones are favored (-2.9) over TCU.

Unless otherwise expressly indicated by BHGP editors, this FanPost is strictly the viewpoint of the author and is not endorsed by BHGP in any way.