Returning to last year's sporadic diary, here is a preview of Indiana by the numbers.  If you are new to By The Numbers or, more likely, don't remember details will follow at the conclusion of the post.

Rush Offense vs Indiana

Michigan is currently rated +6 on run offense, good for 5th nationally and tops of any team having played 3 games vs FBS competition.

Indiana, despite their impressive yards allowed and traditional rankings, have only established an opponent adjusted rush defense of -1, 87th overall.

Factoring in projected carries for Michigan, the running game should be worth 10 points above average for Michigan this weekend.

Currently, Michigan's RB have been sharing the value:
Carlos Brown, +3/game
Brandon Minor, +2/game
Michael Shaw, +0/game

Pass Offense vs Indiana

Michigan's as necessary pass game is rated 38th nationally overall. It factors out to be a solid but not spectacular +2.

The Hoosier's pass D ranks out about as average as average can be with a flat 0 rating and a 60/120 national rank.

That puts the expected value for the game when the ball is in the air for Michigan at +2.

Individually, Tate has posted a +5/game (individual stats don't count sacks and include rushing) while Denard is still in the positives at +2/game.

Hemingway remains at a +6/game
Koger is the only other qualifier at +3/game.

Rush Defense vs Indiana

Like the Hoosier pass D, Indiana's rush offense has been pretty middle of the road, to date. Through 2 qualifying games, the rating is -1, 62nd nationally.

Michigan's disappointing run D finds itself smack dab in the middle of the pack, as well. Rating +1, 50th nationally.

Projected overall advantage, +2 points to Michigan.

Indiana RBs:
Trea Burgess: +2/game
Darius WIllis: +1/game
D McCray: -2/game

Pass Defense vs Indiana

To date, Indiana's passing game hasn't been anything to get too excited about.  Ranking 88th nationally, their -3 rating must go up against Michigan's 15th ranked pass D, producing an overall 8 point advantage for the Wolverines.

Indiana QB:
Ben Chappell: -2/game

Indiana WR:
D Belcher: +2/game
Tandon Doss: +2/game
T Turner: +1/game

Special Teams Ranks

FG/PAT: M 27 vs Ind 108
Kickoff: M 32 vs Ind 62
KR: M 12 vs Ind 3
Punt: M 40 vs Ind 116
PR: M 62 vs Ind 39

Penalties

M -3 pts/game Ind -1 pts/game

Prediction?

Michigan 39 Indiana 12

By The Numbers Background

All down/distance/yardline situations are assigned a value.  A 1st and goal at the 1 is worth about 6.5 points on average.  A second and 4 from midfield is worth just under 3 points.  Every play changes down and distance and yardline and therefor changes the value.  That value is then added up and assigned to the players who either rushed, passed or caught for it.  These numbers are then added up for the season and compared against how other teams fared against the same unit. A great example of this is Indiana's run defense.  The raw numbers look great but the teams they did it against weren't any good so the value is greatly diminished based on competition.

A couple notes and changes.

• Games against non-FBS (1AA) schools are excluded and have no factor.
• Garbage time plays are either diminished or dropped altogether.  Plays count for half if the lead (in possessions) plus the quarter is 7 and they don't count if the difference is any great than 7.
• Sacks don't count against QBs but they do count against a team's passing numbers (not as rushes).
• Fumbles are considered random events and excluded from all calculations, interceptions are included for both players and team ratings.
• Defensive players are rated on two metrics, quantity of negative plays (plays reducing the offenses expected value) and quality of negative plays, total value taken away from the offense.
• All data comes from play by play data found on the NCAA's official website.
• Receivers are only rated on passes they caught, not incompletions targeted at them, therefore their ratings always seem higher than RBs and sometimes even QBs.
• Interceptions factor in returns, so Denard's throws last week weren't terrible because they were like punts, downfield with little or no return.  Picks with long returns produce seriously negative values.