Expectations: Week 2 Total Wins Update

Preamble

Granted, after week 2, the statistical sample size remains small, and the quality of competition lacks the robustness required to keep games from lapsing into garbage time by early in the third quarter, which further aggravates the lack of data that makes up the advanced analytics. Nonetheless, the Vegas bookies aren’t going to stop setting lines and taking your money, because it’s their job; and your favorite fancy-stats analysts will still put out equivalent point-based ratings because it’s their job to help you give less of your money to the Vegas bookies. The wheel in the sky keeps on turning, and the publication of the statistics such as they are goes on. Coming up with more reasons for ignoring these early season statistics is left an exercise for the reader in the comments below. Here you will find further ruminations on said statistics into still more statistics as a means for enabling further discussion, jumping to conclusions, flying off of the handle or goading your rival.

Previous diaries presented overall as well as in-conference win probability distributions for all Big Ten teams based on relative expected points ratings from Bill Connelly (S&P+ preseason on SBNation), ESPN (FPI) and Ed Feng (Power Rank). Last week bore witness to the shift from the more subjective preseason elements that are re-evaluated after each season to the more objective factors that are updated with each game played. Now we can press on with the first iteration of the in-season ratings. 

Schedules, Spreads & Win Probabilities

Since we remain in the midst of the glorious non-conference segment of the season, it seems reasonable to stick with analyses of the overall schedules.

B1G East Schedule Rundown

The table of schedules below shows the overall schedules for all seven teams in the B1G East with attendant point spreads and win probabilities for each game based on the Bill Connelly’s S&P+ weekly ratings. The last table simply shows a rank-ordering of the B1GE teams based on their  expected in-conference win totals, it’s not a projection of divisional standings based on projected wins, losses, and tie-breakers.

What was particularly notable last week in the S&P+ results - that was, U-M being favored in all of its games - was apparently short-lived, as once again no team in the B1G East can make that claim. The love affair between S&P+ and our beloved Wolverines does go on - just not so hot and heavy as last week. Michigan and Ohio State are underdogs in one game apiece, Penn State in three (including its loss), and Michigan State in four. What is especially nice is that both OSU and MSU are underdogs in their games coming up. The rub is that OSU's only underdog matchup is in a non-conference game, and their being favored over U-M although not good, does not entirely kill U-M's chances for the B1GCG. This will be re-examined next week after the non-conference schedule is wrapped up (with the exception of MSU, who has an Oct. 8 matchup with BYU).

Of the other three Rutgers remains absolute cannon fodder, but Indiana is favored to win six total games, and Maryland, five, so either may well be bowl-eligible at the end of the season.

Honestly, six of seven teams in the Big Ten East becoming bowl eligible would be something quite significant toward laying claim to the title of most powerful division in all the land.

In the aggregate, Michigan still looks to be the team to beat with about 10.2 expected wins, edging Ohio State for the top spot by just under 1.2 wins - down from 1.4 wins last week. Michigan is still the only team expected to have a double-digit win total.

Here is a link to a similar table of schedule win probabilities based on FPI Ratings. Ed Feng is being a slacker, so there are no results from Power Rank.

The FPI results differ to some extent, most notably that the spreads of expected win totals are not as wide, which suggests more closely contested races. All of the “contenders” are underdogs in at least one game. U-M and OSU are both underdogs in a single game. PSU is an underdog in four games; MSU, five; and Maryland, six. Interestingly, FPI result now show Maryland edging Michigan State in total expected wins, both within 0.1 of 6 wins. Michigan remains the only team expecting double-digit wins.

B1G East Expected Overall Wins

The bar plots below show the expected total overall wins distributions for teams in the B1G East, in alphabetical order. Noted above each bar is the probability for that number of wins (you may need to click & embiggen to read it). The bar with the highest value is the most likely outcome (the mode). Also marked on each plot is the expected overall win total (the mean). The last line plot is just an overlay of the same data from the other seven bar plots.

What remains noticeable in the S&P+ results even going back to the preseason forecasts is how much higher the peak of Michigan’s distribution is than any other team. The spread is also narrower, but that is less obvious. What this means is that not only does Michigan have the highest expected win total, but it is also the most likely to hit that mark. Also, Michigan has the highest mode of any team at 10 wins, one ahead of OSU follows at 9, followed by MSU at 8, and PSU at 7. Michigan still maintains the best chance of having an undefeated season at 11.1% or about 9:1 odds, followed by OSU with a 2.0% likelihood.

At this point, the distributions of Maryland and Indiana are almost indistinguishable, as are both of their expected win totals, less than 0.1 wins apart.

Here is a link to a similar plot of conference win distributions based on FPI ratings.

The FPI results favor OSU to much greater extent, showing both U-M and OSU with the same mode of 10 wins; U-M tilting to the higher side, and OSU to the lower. A clear separation of nearly 3 games has expanded to next closest contender, Penn State with a mode of 7 wins. Maryland, Indiana and MSU are in a three-way tie for the fourth place mode with 6 wins, with Indiana tilting to the lower side.

B1G West Schedule Rundown

The next table of schedules shows the overall schedules for the B1G West based on the Bill Connelly’s S&P+ weekly ratings. Again, the last table simply shows a rank-ordering of the B1GW teams based on their  expected win totals - it’s not a projection of divisional conference standings.

Since the emergence of game-based S&P+ ratings the contenders in order of overall expected wins in the B1GW are Minnesota, Nebraska, Iowa and Wisconsin. That said, the first three are tightly packed, within 0.25 wins of each other, and less than 0.7 wins separate Minnesota from fourth place Wisconsin. No team is favored in all of its games. No team is expected to have a double-digit win total. Indeed, Minnesota and Iowa are underdogs in 3 games apiece; Nebraska and Wisconsin in 4 games. However, Minnesota is favored at home over Iowa on Oct. 8, which would give the Gophers the inside track for the B1GCG in Indy.

After Northwestern dropped its second game of the season, they are on their way to becoming the doormat of the B1GW. Favored in only 4 remaining games, the Wildcats hopes of a bowl-bid may have evaporated. At this point Illinois is expected to win more games than Northwestern, but is favored in fewer: 3. Meanwhile, Purdue is only favored in one of its games the remainder of the year, its matchup with Nevada this week.

Here is a link to a similar table of schedule win probabilities based on FPI ratings.

FPI also expects the same four teams to have winning records. In order of expected wins, they are Nebraska, Iowa, Wisconsin and Minnesota.  Nebraska and Iowa are essentially tied at the top, followed by Wisconsin, less than 0.3 wins behind. Minnesota lags by another 1.7 games. No team is favored in all of its games. No team is expected to have a double-digit win total. Iowa is an underdog in the fewest number of games: only one! Nebraska and Wisconsin are underdogs in three; and Minnesota, five.

The bottom line remains that the B1GW race is wide open in the statistical prediction sense. The consensus at this point is that Nebraska, Iowa and Wisconsin are all evenly matched teams within about 0.3 games of each other. On the other hand, ranging in expected wins from 6.8 to 8.2 and being an underdog in anywhere from three to five games, it still remains that Minnesota is a riddle, wrapped in a mystery, inside an enigma.

B1G West Expected Overall Wins

The bar plots below show the expected overall win distributions for the B1G West teams, in alphabetical order.

The story here is nearly the same as last week, except now add Iowa to Nebraska and Minnesota, whose distributions are almost indistinguishable (within 0.3 expected wins of each other, and the same mode of 8 wins). The distribution for Wisconsin tilts slightly to the lower side, but still with a mode of 8 wins as well. It appears highly unlikely that any team will have an undefeated season. Minnesota has the best chance of a one-loss season at 3.8%, followed by Iowa at 2.7% and Wisconsin at 1.3%.

Here is a link to a similar plot of conference win distributions based on FPI ratings.

The FPI results tell a similar story, but with Wisconsin and Minnesota essentially swapped - and Minnesota with a mode of 7 wins, instead of 8. The distributions for Iowa and Nebraska are practically identical, and Wisconsin tilts to the lower side.

Yours in football, and Go Blue!