Back in the summer, I presented my full prediction for the Big Ten race, including percentages for each team to win their Division, the conference, and make the playoffs. After two weeks of action, I finally have new data to input to see how things change. But, this data update comes with a TON of caveats. The only information that my formulas have to work with are the preseason rankings and the result of a one or maybe two games, which I now weight equally. So, things are going to be a bit erratic for a few weeks until more data accumulate. That said, here are my updated odds to win the Big Ten, based on my algorithm:

and here is the update expected win table

A lot has changed since the summer (when Michigan was the favorite to win the Big 10 and was predicted to be favored in all of their games) and a lot has even changed since Week 1. Of course, it is still early, but right now the #math really likes how Ohio State and Wisconsin have started the season, as they are ranked #1 and #3 nationally in my current power rankings. As such, they are both big favorites (77% and 72% odds) to win their divisions.

What is more fun is that my spreadsheet gives MSU the 2nd best odds in the East at 13%, which is right around where we started. Also of note is that Maryland actually has the current 3rd best odds (6%) while Michigan’s odds are less than 1%, their ranking has dropped to 33, and I only have them favored to win 5 games total. In the West, Iowa (24%) is still hanging around and ranked 10th in my power rankings.

Now, a lot can certainly change as teams hit their stride and we learn more about them, but right now I think that these numbers are an accurate measure of how things will play out if each team plays to the level that they have played at so far. Another way to say this is that if you think Michigan has looked shaky and have no chance to beat anyone good unless they start looking a lot better, well, the math agrees with you.

I will say that EPSN’s FPI is not quite as down as the Wolverines as my calculations suggest, but MSU did edge ahead of Michigan this week, so the general trend is still the same.

In addition to those odds, I can also provide an update on my projected spreads for MSU’s remaining key games:

• MSU (-10.9) vs. Arizona State (it opened at -10.5)
• MSU (-11.1) at Northwestern
• MSU (-23) vs. Indiana
• MSU (+11.5) at Ohio State
• MSU (+13.3) at Wisconsin
• MSU (-7.5) vs. Penn State
• MSU (-5.7) at Michigan

While the back-to-back road games in Madison and Columbus are looking tougher by the minute, this analysis shows that MSU could actually be favored in Ann Arbor this year. I hope you all locked in that MSU +14 line that was floating around this summer...

Week 1 Upset and ATS Pick Review

Last week I gave my spreadsheet’s picks ATS and also a few recommended bets based on the data that I have accumulated over the last few years. Overall, my model did great ATS this week at 27-19 (59%), which makes up for a poor Week 1 and brings my total to 46-45. Meanwhile, the FPI is off to a bad start. It went only 20-26 this week and is 37-54 (41%) ATS overall based on my calculations.

As for my suggested bets, if I combine my spreadsheets picks with the FPI picks, and discount the Syracuse / Maryland game where the models were in complete disagreement (the FPI was correct in this case) it was a push for the week (6-6). In detail, those picks were (with the correct one in bold):

• Penn State (-24) to cover vs. Buffalo
• Wisconsin (-34) to cover vs. Central Michigan
• Boise State (-12) to cover vs. Marshall
• Appalachian State (-21) to cover vs. Charlotte
• Iowa (-20.5) to cover vs. Rutgers
• Baylor (-26) to cover vs. UTSA
• UCLA (-7) to cover vs. San Diego State
• VA Tech (-26.5) to cover vs. Old Dominion
• Vanderbilt (+9.5) to cover at Purdue
• LSU (-4.5) to cover vs. Texas
• Cincinnati (+17) to cover at Ohio State
• Kentucky (-14) to cover vs. EMU

Of course, 4 of the 6 correct picks were from my model... just saying’. Overall, this methodology is 14-9 (61%) year-to-date.

As for upsets, here is the summary table for the week:

My model got 2 of 3 this week, bringing the year-to-date total to 3-2 (60%), while the FPI was 2 for 2, making up for its 0-3 performance this week (2-3, 40% YTD). The overall 9 observed upsets was exactly what my simulation predicted, with the Cal over Washington upset being the biggest of the weekend.

National Overview

Finally, after each week, I like to take a look at the actual point differential in each game as a function of the opening Vegas line. I find that this visualization is a useful way to understand at a glance which teams over- and under-achieved for a given week. While the spread is a good predictor on the overall average point differential, the variance in the data is quite large. The standard deviation is actually about 14 points, which is where I set the dotted lines in the plot below.

At a glance, it is easy to see that some teams clearly over-achieved relative to the spread, namely Kansas State, Baylor, Ohio State, MSU, UCF, Missouri, USC, and SMU. All 8 of those teams were over a standard deviation above the mean (aka, the spread). Also of note is that Tulsa was flirting with the dotted line, which shows that they did quite well ATS this week too.

Meanwhile, the two teams that still won, yet were notable underachievers are none of than Michigan and Florida State. Wow, you hate to see that... Then, there were the 9 upsets mentions above all below the x-axis. The most notable of which is the Maryland win over Syracuse. It is notable not because it was a big upset, but because of the margin of victory, which was huge. They are a major outlier this week.

That is all for now. Onward to Week 3. As always, enjoy, and Go Green!