By The Numbers - Michigan State Preview

Submitted by The Mathlete on
For a primer on the stats used here, go to the bottom of http://mgoblog.com/diaries/indiana-numbers

Rush Offense

A look at all of Michigan's performances this year, opponent adjusted.

Game - Rush+
Def Total Off Rush Yards TD
Western Michigan 2.42 50 242 1
Notre Dame 5.03 36 199 2
Eastern Michigan 12.75 37 399 6
Indiana 2.40 47 153 3

Michigan's run game, obviously, hasn't a bad game all year.  Their 5.6 overall average is the #5 ranked unit in the country.

As for State, they too have had generally positive performances from a top ranked (17) unit.

Game - Rush+
Off Total Def Rush Yards TD
Central Michigan 4.05 26 81 0
Notre Dame -0.03 35 149 1
Wisconsin 4.21 49 197 1

Sparty's 2.7 average brings down Michigan's expected performance for the game to a still solid +3 advantage.

Player - RB - Season
Player Team G Value Rush Yards TD Rec Yards Rec TD
Carlos Brown Michigan 4 4.89 9.50 80.25 0.75 37.00 0.33
Brandon Minor Michigan 3 2.08 10.33 59.00 0.67



Pass Offense

Despite more moderate yardage output, Michigan's passing value has still been solid at +3.1 overall, good for 30th in the nation.

Game - Pass+
Def Off Total Att Comp Yards TD INT
Western Michigan 8.39 28 15 197 3 1
Notre Dame 0.63 35 25 231 2 1
Eastern Michigan -2.95 19 9 49 0 2
Indiana 6.52 26 15 216 2 1

The real opportunity opens up against Sparty's pass defense.  Even when adjusting for opponents (like Notre Dame), the unit comes in at 101st in the country, -5.0. 

Game - Pass+
Off Def Total Att Comp Yards TD INT
Central Michigan -6.19 50 36 337 3 1
Notre Dame -1.49 36 26 288 3 1
Wisconsin -7.29 31 19 243 4 0


Michigan year to date has passed less than the average MSU opponent, so if Tate and Denard?! put more balls in the air, this could be an 8 point advantage for Michigan.  Even at their current passing levels, it is a solid 7 point opportunity, the biggest unit advantage in the game.

Player - QB - Season
Player Team G Value Yards TD INT Rush Yards Rush TD Att Comp
Tate Forcier Michigan 4 5.60 167.75 1.75 0.50 40.50 0.50 21.75 13.50
D Robinson Michigan 3 2.19 19.00 0.00 0.67 51.67 1.00 3.67 1.33

Player - WR - Season
Player Team G Value Value+ Rec Yards TD
Greg Mathews Michigan 2 3.13 1.17 3.50 43.00 0.50
J Hemingway Michigan 3 3.01 2.54 2.33 37.00 0.67
Kevin Koger Michigan 3 2.76 2.70 2.67 35.33 0.67

Rush Defense

The fears for this unit have been well discussed and the numbers here bare a muted version of the same story.

Game - Rush+
Off Total Def Rush Yards TD
Western Michigan 2.97 21 45 0
Notre Dame -1.81 29 159 1
Eastern Michigan 0.39 46 192 2
Indiana -2.52 32 209 3

After a solid opening game, Michigan has regressed and fallen into the second half (67th) of the national ranks.

Without Mr. 50 carries a game, the State offensive focus has shifted and the running game is still above average (43rd) but nothing fearsome at this point.

Game - Rush+
Def Total Off Rush Yards TD
Central Michigan -0.26 29 106 1
Notre Dame 1.27 25 105 2
Wisconsin 1.09 19 101 0

As MSU has done all season, this shouldn't a gain or liability for either team all season, just a way to slow down BG and Roh on the pass rush. 

Look for State to have a point or two advantage in the running game, depending on how much they are willing to shift focus to it vs throwing it.

Player - RB - Season
Player Team G Value Rush Yards TD Rec Yards Rec TD
Larry Caper Michigan State 3 1.14 8.00 35.67 0.67

Caulton Ray Michigan State 3 -0.21 7.33 24.67 0.33



Pass Defense

Despite the justified fear that comes from having but a single DB you feel safe with and a marginal pass rush, Michigan's pass defense has produced fairly well.  Ranking 29th nationally, the unit has been worth 3.6 points per game to date, without a single game significantly below average.

Game - Pass+
Off Def Total Att Comp Yards TD INT
Western Michigan 8.75 41 25 251 1 2
Notre Dame 2.41 42 25 336 3 0
Eastern Michigan -0.21 27 16 100 0 1
Indiana 3.72 39 22 258 0 1

The fear returns when looking at what Michigan State has done through the air so far this season (17th).

Game - Pass+
Def Off Total Att Comp Yards TD INT
Central Michigan 8.07 27 17 210 2 0
Notre Dame 3.63 39 25 340 2 1
Wisconsin 4.60 48 26 384 4 3

State has put up very solid numbers in all three game, even if Wisconsin's numbers are padded by a couple late inconsequential drives.

The net of all of this is a 2 point advantage for MSU.

Player - QB - Season
Player Team G Value Yards TD INT Rush Yards Rush TD Att Comp
Kirk Cousins Michigan State 3 3.91 222.33 1.33 0.67 2.50 0.00 29.00 17.67
Keith Nichol Michigan State 3 2.40 84.67 1.00 0.67 18.67 0.00 7.67 3.67

Player - WR - Season
Player Team G Value Value+ Rec Yards TD
Mark Dell Michigan State 2 6.59 3.52 5.50 97.50 0.50
B Cunningham Michigan State 3 6.42 3.37 6.00 79.67 1.00
Keshawn Martin Michigan State 2 5.47 3.35 2.50 76.50 1.00
Blair White Michigan State 3 5.07 3.49 4.67 62.67 0.67


Special Teams

The specialists should be a 2-3 point advantage for Michigan.  We are +2 for the year, good for 12th nationally while MSU is 47 at just below 0 points per game. 

Based on this years results, Michigan should have the advantage in every unit with the exception of punt return.

Predictions

Because no one can have enough predictions, I am adding predictions for all Big 10 games this week.

The numbers say Michigan has a 10 point advantage with the ball, and State a 3 point advantage when they are in possession.  Home field is worth, on average about 3 points that is likely offset by Michigan's advantage in special teams and field position.

Michigan 38 MSU 31

Elsewhere in the Big 10:
NW 20 @ Purdue 40
Wisconsin 24 @ Minnesota 25
Arkansas State 7 @ Iowa 43
Penn State 31 @ Illinois 9
Ohio State 27 @ Indiana 17

Comments

colin

October 2nd, 2009 at 3:13 PM ^

regressing your numbers? Michigan may well be better, but I doubt they are 10.5 points better than the line on average (i.e. imagining the teams played numerous games against each other). How much do we really know after 1/3 of the season?

The Mathlete

October 2nd, 2009 at 3:23 PM ^

There is definitely a lot of noise in the system at this point in the season, but it also surprising how quickly the season's data pulls out real trends. Doing this for last weeks games, with no more than 3 games data on any team, I had 17 games that had at least a ten point variance to the spread and went 12-5 against the spread in the those games. Don't expect to get those results every time (otherwise I'd be in Vegas right now) but at the same time the data overall holds up pretty well at this point in the season.

colin

October 2nd, 2009 at 3:19 PM ^

if Michigan's true talent is 7 points better than State, that's an implied chance of winning of about 70%. The line currently has State winning about 60% of the time. That's a huge swing, presumably worth betting on.

RONick

October 2nd, 2009 at 3:19 PM ^

Talking about the pass defense you note there was no game below average. I haven't delved into the numbers, but find it surprising that the notre dame game was not below average at best...

The Mathlete

October 2nd, 2009 at 3:28 PM ^

On an unadjusted basis you are correct, but when saying average, I was talking in relation to opponent. Michigan's performance against ND was the second best performance by a defense against them so far this season. Off Def Value Notre Dame Nevada 16.41 Notre Dame Michigan State 10.92 Notre Dame Michigan 8.33 Notre Dame Purdue -1.36

The Mathlete

October 3rd, 2009 at 10:27 AM ^

Could probably go back in and estimate a value that was lost due to the injuries and lost time. However, since I can't do that for every team in the overall picture, the fundamental numbers will stay the same but I will note any adjustments directly affecting Michigan in my writeups.

IPKarma

October 2nd, 2009 at 3:35 PM ^

Add +7 for MSU's continued woes on the DL, and subtract 7 from MSU for the difficulty in passing in the rain, and you get 45-24. I pray for an Indiana upset of Ohio State Community College.

MichFan1997

October 2nd, 2009 at 4:11 PM ^

this reminds me of fangraphs.com, a baseball site that measures things by runs above and below average. Where did you pull all the info? Keep doing these posts.

Not a Blue Fan

October 2nd, 2009 at 9:24 PM ^

I know you've done your analysis based on the actual numbers, but I don't see any way that Iowa puts up 43 or Indiana scores 17 on OSU. It's just not going to happen. Other than that, good work as usual. Out of curiosity, how are you building and maintaining your database of situational data? Parsing PBP data into a query-able DB? I've considered doing that in the past, but frankly I'm too lazy. Anyway, admirable stuff you've put together here.