Projecting the 2013 Big Ten Basketball Race: January 16

With four conference games in the books, it's time to look at the race for the title in the toughest basketball conference in the land. As always, these are based on two variants of the Bradley-Terry method, with ratings here. Record probabilities are based on individual game probabilities; outright, shared title, and top 4 (bye in the conference tournament) probabilities are based on 100,000 simulations of the remainder of the season.

(Note: The ratings are through games of January 13 and therefore do not include the Wisconsin-Indiana game; however, Wisconsin's win is taken into account when simulating seasons. For ties for the last bye, each team is given partial credit for a top-4 finish accordingly - for instance, if there is a three-way tie for 3rd through 5th in a simulation run, each of those teams gets credit for 2/3 of a top-4 finish since two of those three would get byes.)

Basic Method

The conference breaks down into four approximate tiers:

• Minnesota, Michigan and Indiana are the big contenders (prior to their home loss to Wisconsin, Indiana was about even with Minnesota at the top of the projections). Minnesota has a slight advantage due to schedule difficulty - they don't play at Michigan or home against Ohio State.
  
• The fight for the fourth bye appears to be primarily between Michigan State, Ohio State, and Wisconsin, with about a 1 in 3 chance that one of the top 3 manages to slide out to leave another spot. In the ratings, Michigan State has a clear advantage but also the most difficult schedule in the conference (all four of the single-play games are against the bottom half) to bring them closer to the pack.
  
• Illinois and Iowa are potential bubble teams, though Iowa will probably need to exceed these projections by at least a full game, maybe two, to make it.
• Nebraska, Northwestern, Purdue, and Penn State bring up the rear, with Penn State well behind the others.

Margin-Aware Method

Same tiers here (though the odds of one of the top three slipping far enough to have to play in the first round of the conference tournament are much smaller), but Michigan State and Ohio State approximately swap places in the ratings (28th and 20th instead of 19th and 26th), with the result that MSU drops to the back of the second tier. Iowa's chances look a little better (20% or so, if we assume they make it half the time at 10-8 and always at 11-7; while that sounds pessimistic, a look at Iowa's non-conference schedule and the fact that their single-play opponents are MSU, OSU, Michigan, and Illinois suggests that 9-9 won't be enough). Illinois and Nebraska drop a game back as well, and Penn State ... well, ouch. A 1 in 12 quadrillion chance of winning the rest of their Big Ten games is the worst number I can remember seeing at this point in the season.

Michigan State's Remaining Games

In schedule order:

@ Penn State: 76% win probability by the basic system / 86% by margin-aware / +9.5 points margin
Ohio State: 69% / 66% / +3.5
@ Wisconsin: 51% / 39% / -2.5
@ Indiana: 24% / 9% / -12
Illinois: 71% / 80% / +7
Minnesota: 51% / 42% / -1.5
@ Purdue: 76% / 71% / +5
Michigan: 51% / 38% / -2.5
@ Nebraska: 65% / 72% / +5
Indiana: 53% / 40% / -2
@ Ohio State: 38% / 23% / -6
@ Michigan: 22% / 9% / -12
Wisconsin: 79% / 80% / +7.5
Northwestern: 89% / 93% / +13.5

Breaking it down by difficulty (and including games already played in the category in which they would fit based on current ratings), we have:

Prohibitive Favorite: (>90% probability, averaged between the two systems): Purdue (W), Nebraska (W), Northwestern
Heavy Favorite: (75-90%): @ Penn State, Illinois, Wisconsin
Slight Favorite: (60-75%): Ohio State, @ Purdue, @ Nebraska
Toss-up: (40-60%): @ Iowa (W), @ Wisconsin, Minnesota, Michigan, Indiana
Slight Underdog: (25-40%): @ Ohio State
Heavy Underdog: (10-25%): @ Minnesota (L), @ Indiana, @ Michigan

Getting anything from the underdog categories is a bonus, more so if you believe the margin-aware data (road trips to Indiana and Michigan slide into the "prohibitive underdog" category and Ohio State to "heavy underdog"). Of interest is that in the remaining toss-up games, we're a very slight favorite in all four in the basic method but more significant underdogs in all four in the margin-aware. Winning two of the remaining toss-ups and six of the remaining seven games where we are favored would get us to a respectable 11-7; to win the conference, we would probably have to win everything in the "toss-up or better" categories or pull off a major road upset.