“There was a long hard time when I kept far from me the remembrance of what I had thrown away when I was quite ignorant of its worth.”

- Estella Havisham (from Great Expectations by Charles Dickens)

This is going to be a significantly abbreviated Diary Post, since various issues have conspired to throw mutiple wrenches into the usual work flow that precedes this nominally weekly posting. Those issues include it being a foreshortened holiday week with games being played through Monday, and so updated rankings were not available until Tuesday at the earliest. What's more (and unbeknownst to me until I went looking for updated rankings) is that in the time between beginning the write up for the Preseason Diary and now, Bill Connelly left SBNation's Football Study Hall and Football Outsiders to shack up with ESPN. Sadly, Football Study Hall is looking like a cobwebsite these days. Gone are things like the Advanced Statistical Profiles that had provided a fix for stats junkies over the past several years. As such, what had been a reliable set of pages from which to scrape data for further processing is no more. Fortunately, Bill did post an article at ESPN that included a table of his latest S&P+ team ratings. So, to get to this point, it's been a bit of a mad scramble to modify the scraping script to grab data from this new source, and this will only work for this one week. It remains to be seen whether the S&P+ data will continue to be available, and if so, if it will appear in a consistent location week-to-week.

This says nothing about the availability of all of the other S&P+ Advanced Stats, (or the nifty Box Scores) for which Bill is known. Football Outsiders no longer provides any links to S&P+ resources on its front page. As of this writing, this typically rich data set has not appeared anywhere else, and so it appears as though there is going to be a dearth of the metrics we've come to know and love - metrics like Explosiveness, Opportunity Rate, Adjusted Line Yards and Havoc, just to mention of few of my favorites. It remains to be seen whether these issues will continue to subvert progress, and if so, well - I'll just lay the blame on ESPN. What this means is that you can rest assured that ESPN's FPI will remain in good standing, but don't be surprised if the more detailed advanced stats reappear secured behind a pay wall at point in the future.

https://youtu.be/DfHJDLoGInM

B1G Team Ratings

So without further ado, let's look at some of the usual tables and charts, beginning with the ratings themselves:

What you'll notice are some significant movements relative to the preseason ratings, most notable those of Wisconsin (who now is tops in the B1G as per S&P+) and Penn State (tops in the B1G as per FPI, and now second in hte B1G as per S&P+). Combine that with the slide Michigan took (from #9 to #14) in the S&P+ ratings (making them 4th in the B1G overall), and the outlook has soured a bit for the Team 140 campaign. Enough said about that, so moving on...

Turning Spreads Into Likelihoods

The following four tables of tables show the overall schedules for each team in the B1G East & West divisions, including the ratings-based point spreads and attendant win probability (it is left as an exercise for the reader to subtract the OOC probabilities from the overall totals to derive the in-conference estimates of expected wins). A green-yellow-red pseudo-colormap is also applied to quickly give a qualitative indication of where each team faces its greatest challenges, how those challenges stack up, and by comparison, which teams have a more or less difficult row to hoe. Green indicates a favorable point spread; yellow, a competitive matchup; and red, an unfavorable margin. The last table simply shows a rank-ordering of the divisional teams based on their  expected in-conference win totals - it’s not a projection of conference standings based on projected wins, losses and tie-breakers. You may need to open the tables in new windows (or save them to your smart-assed device) to be able to embiggen. Another thing remembered whose worth was undervalued.

What's notable - and this is not always the case - is that the rank ordering of the teams is the same as their orders on an expected wins basis.

E(S&P+)N Schedule Tabulations

Do not be deceived by that last statement, however, because even though it is a fact, it is also a statistic. What's also noteworthy is that according the S&P+ schedule tabulations, there is one and only one team in the B1G that is favored in all of its games, and that team is a State University in Ohio. Lucky for them that Penn State comes to Columbus this year, because if the game were in Happy Valley, the projected outcome would be reversed. The margin for that game is that close. What's also implied in that statement is that Michigan is not favored in The Game, where the margin is less than 2 points. What's worse is that UM is an underdog by more than one score when it goes on the road to Madison and Happy Valley.

FPI Schedule Tabulations

FPI has quite a different take on the B1GE. Similarly to S&P+, FPI projects only one team to be favored in all of its games, and that team is Penn State. Indeed, the only matchup in which UM is and underdog is when it goes on the road to Happy Valley, with a margin of less than one score. But on the bright side, FPI has UM favored to win The Game by more than one score, and is also favored on the road in Madison.

Total Win Probabilities

Below are the composite charts of the total overall and in-conference win distributions for the B1G based on the S&P+ and FPI ratings.

The Contenders are those teams that by one rating or the other, are shown to be within 2 total expected wins of the top rated team. The Bubble consists of those teams within 2 total expected wins of being bowl-eligible (or not). The Bottom Feeders are the teams with few prospects for gridiron glory or post-season hijinks.

E(S&P+)N Projected Win Distributions

In the B1G East, the Contenders are Penn State, Ohio State and Michigan, the Bubble teams are Michigan State, Indiana and Maryland, with LOLRutgerz as the solitary Bottom Feeder. The S&P+ distributions for Penn State and aSOU are nearly indistinguishable, particularly when considering the in-conference characteristics. 

In the B1G West, it's looking like Iowa is the only team able to contend with re-emergent Wisconsin, who despite being an underdog when it goes on the road to Columbus, remains the team that is most likely to go undefeated. The remaining teams are all looking like they are in the Bubble (for now), with Illinois showing a remarkable climb up to middling level, combined with the somewhat stunning backslides of Northwestern and Purdue.

 

FPI Projected Win Distributions

In the B1G East, the Contenders are Penn State, Ohio State, Michigan and Michigan State; the Bubble teams are Indiana and Maryland, with LOLRutgerz the solid, solitary Bottom Feeder. The FPI distributions show some pretty clear separation among the teams, with Penn State clearly at the top. Meanwhile, Michigan is contending in a very close #2 spot, with a mode of 9 overall wins to PSU's 10 wins. In-conference, both PSU and UM have share top mode of 7 wins. ASOU is one win behind UM in both overall and conference modes; MSU, one win behind aSOU. 

In the B1G West, FPI shows Wisconsin at the top with Iowa and Minnesota as the other two teams in contention. The remaining teams are all looking like they are in the Bubble (for now), with Purdue barely hanging on to a glimmer of bowl-eligibility.

Schedules Matter … Big Time

So after summing up all of the projected wins and losses we can have a look at projected final standings. Head-to-head and three-way tie-breakers are taken into account where applicable. A crude strength of schedule is provided for both overall and in-conference games by calculating the average rating of each team’s opponents, and rank ordering them. Only the ranking is given, not the computed values.

So here we can see that both S&P+ and FPI are projecting outright winners in both the B1G East and West divisions - there are no ties in which a team with one or more conference losses gets and invite to Indy. In both cases, Michigan has the toughest overall schedule in the conference, which should help in terms of any potential CFP considerations provided it can get through its in-conference slate. That will be easier said than done, particularly since all of the teams projected to finish ahead of Michigan in both the S&P+ and FPI projections have easier in-conference schedules. Indeed, UM has the 2nd most-difficult conference schedule per S&P+, with only LOLRutgerz having a tougher one.

Yours in Football, and Go Blue!