Admit it. By halftime you thought the game was over. SPAM knew better, people. SPAM said Iowa would win. O ye of little faith! Hell, SPAM has been telling you for the last two weeks that Iowa was going win. You need to listen. SRSLY.
Last week, Super Prediction Action Machine predicted a 4-point Iowa victory, though I cautioned that it was really a 3-point margin because SPAM rounds off fractional points. SPAM called for a 29-25 Iowa win, and the Hawkeyes pulled it out, 24-21. SPAM IS UNCANNY. (ba doom ching).
Once again I plugged the performance of the Big 10 back into the machine, and here is how SPAM sees the rest of Iowa's season shaking out:
L @Nebraska (28-26) (last week: L 29-26)
SPAM stubbornly insists that Iowa will lose to Nebraska. It's hard to reconcile this, but I think the issue is Nebraska's offensive output and Iowa's still-nagging penchant for error. This is my first season of running SPAM and part of the problem is that SPAM looks at overall statistics and, for the first 5-6 games, any one game's statistics comprises 10-20% of SPAM's sample size. By now, it's about 8%, so recent successes/struggles don't play in much. The Michigan game also screwed it up a bit, as the number of yards of offensive output that translate into points for our opponents dropped after Michigan scored 21 points off less than 200 yards of offense. So, the fact that Nebraska is projected to score 28 on us might be a little suspect. I'm not sure how well the model accounts for things like self-inflicted wounds. If the Hawkeyes don't repeatedly shoot themselves in the foot against Michigan, it could very well have been a 28-7 game in Iowa's favor at halftime. Anyway, the official SPAM prediction is 28-26 Nebraska. That's a lot of Hawkeye field goals.
But, if Iowa pulls off the win at Nebraska, I believe the Hawkeyes finish second in the Legends behind Michigan State. To recap, in that scenario, IA will have beaten Iowa State, Minnesota, and Nebraska on the road, and Michigan at home. Wisconsin is a fish that got away, but by any reasonable standard, that makes for a spectacular season of football, especially considering where this team was at the end of 2012. It also makes my prediction of a 2014 or 2015 Rose Bowl win slightly less idiotic.
Season in Review
SPAM has called Iowa's overall record flawlessly to date (7-4), except that SPAM swapped the win at Minnesota with a win at home against Northern Illinois. SPAM has predicted total points scored to 91.28% accuracy. SPAM's final prediction for Iowa '2013 is 7-5 with a likely bowl win, for 8-5. Not bad!
At season's end, I'll post SPAM's best and worst case scenarios for the next year and we can compare, though be forewarned that SPAM can't account for changes in personnel and what not in the off season, it's pure statistics only.
So I was looking at the algorithm and I noticed that SPAM called only two games wrong: SPAM called for IA over NIU and MN over IA. However, if I remove home field advantage from the calculation, both of those games flip - that is, SPAM calls for NIU to upset IA and calls for IA to win at MN. What's more, NONE of the predicted outcomes of other games change, which means when I remove my HFA factor, SPAM has called EVERY SINGLE IOWA GAME CORRECTLY this year and the points accuracy shoots up from 91.28% to 95.61%.
With that in mind, I re-calculated Nebraska without home field advantage, and SPAM arrives at a tie in overtime. I know, that's not possible, but it's just because I only have SPAM iterate out one overtime series and it says they're tied. I had to reformat the numbers to get a tiebreaking, and it goes to Nebraska. The MPTs for Iow aare 25.19, and for Nebraska, 25.38. This basically means Nebraska is favored in SPAM by 17/100ths a point.
I guess this doesn't really change the prediction but it does tell you how close these teams are.