A Detailed Look At U/M Turnovers

Submitted by Enjoy Life on October 29th, 2009 at 7:39 PM
EDIT: I changed the TO numbers for M in 2001 per the comments below. The MGoBlue site apparently has bad data for 2001. This actually reinforces the conclusions, though.

I posted a board topic on U/M turnovers a few days ago. I'm also in the midst of taking a look at turnovers throughout the FBS and doing an analysis of what the heck TOM (turn over margin) means.

There has been a lot of stuff posted about the turnovers M has been experiencing and what the cause might be. I decided to look at RR's experience with turnovers at WVU and compare that with turnovers at M before RR and during RR.

Here are the turnovers for RR at WVU:

WVU

INT

FL

Tot

Opp Int

Opp FL

Opp Tot

TOM

2001

19

13

32

11

13

24

-8

2002

9

6

15

19

15

34

+19

2003

8

12

20

21

15

36

+16

2004

11

11

22

16

9

25

+3

2005

7

10

17

17

14

31

+14

2006

8

9

17

16

8

24

+7

2007

6

15

21

16

18

34

+13

Average/Game

0.8

0.9

1.7

1.3

1.1

2.4

+0.7


Here are the turnovers for M for the same time period:

U/M

INT

FL

Tot

Opp Int

Opp FL

Opp Tot

TOM

2001

9

13

22

11

7

18

-4

2002

7

10

17

16

10

26

+9

2003

10

9

19

14

7

21

+2

2004

12

9

21

16

11

27

+6

2005

8

10

18

12

11

23

+5

2006

8

4

12

12

14

26

+14

2007

14

13

27

14

15

29

+2

Average/Game

0.8

0.8

1.6

1.1

0.9

2.0

+0.4


Here are the turnovers for M for the RR era:

U/M

INT

FL

Tot

Opp Int

Opp FL

Opp Tot

TOM

2008

12

18

30

9

11

20

-10

2009 (8 Games)

10

8

18

7

4

11

-7

Average/Game

1.1

1.3

2.4

0.8

0.75

1.55

-0.85


Here is the summary:

Average Per Game Summary

INT

FL

Tot

Opp Int

Opp FL

Opp Tot

TOM

WVU

0.8

0.9

1.7

1.3

1.1

2.4

+0.7

U/M (2001-7)

0.8

0.8

1.6

1.1

0.9

2.0

+0.4

2008/09

1.1

1.3

2.4

0.8

0.75

1.55

-0.85


My conclusions:
  • Turnovers are NOT a result of the RR "system".
  • Turnovers were very high in RR's first year at WVU but came down after that.
  • So far TO Lost has been unusually high for M in the RR era and TO Gained has been unusually low.
  • There are still 4 games to go this year. Last year TO Lost for the last 4 games was 6 and TO Gained was 5 for a TOM of -1.

Comments

Ichiro

October 29th, 2009 at 7:59 PM ^

Unfortunately, it looks like we're still in "Year 1 @ WVU" rebuilding mode with regard to turnovers. BUT, this year, our TOM looks worse because our opponent's number of turnovers is pretty low. Scaling up by 50% gives 16.5 for the entire year--much lower than any UM or WVU opponents during 2001-07.

jlvanals

October 30th, 2009 at 10:09 AM ^

1. Michigan's defense sucks and can't create turnovers because they're too busy getting burned. (Please god help us)
2. Our offense still can't hold on to the damn ball whether young or old. I don't know who to blame about that, but there are a myriad of factors contributing, not the least of which is two frosh QBs, but its not like Minor and Brown haven't had key fumbles this year. Regardless, if this situation does not improve next year, it will be difficult to write it off as three straight years of bad luck.

raleighwood

October 29th, 2009 at 8:19 PM ^

This is a good analysis. However, I'm not sure that you can totally discount the system being the cause of the turnovers. If you're not running out of the shotgun every snap, you won't get as many balls snapped past the QB.

The analysis doesn't indicate how many offensive plays were run. The previous incarnation of the Michigan offense was more of a ball control, grind it out, attack. They were usually the winner of the TOP battle. The current offense is more designed around the "quick strike". I'd be willing to bet that Michigan ran more offensive plays (and possibly by a significant margin) from 2001-2007 than WVU did. Couple more plays with the fact that Michigan had fewer turnovers (picks and fumbles) during this time and you might get a different picture on a per play basis.

My issue with the current turnover problem is that so many of them have been unforced. Odoms fumbled punts in three consecutive games last year. How does this happen? D-Rob throws picks in roughly 20% of his attempts (much higher percentage if he's throwing more than eight yards down field). How does this happen? These are coachable, correctable things. We should hopefully see these stats skew back into a more favorable direction as the players get more experience.

raleighwood

October 29th, 2009 at 9:16 PM ^

Four more plays per game, 12 games per year. That's 48 more plays per year. In other words, more than half of an extra game's worth of plays during the course of a season.

So, Michigan had ~10% fewer turnovers while executing ~6% more plays. You can draw your own conclusions from that.

ajscipione

October 29th, 2009 at 10:36 PM ^

I believe Brian indicated that most of the M turnovers this year were by Tate and Denard. Therefore, it's a combination of youth/inexperience and learning the system. I believe that this number will drop significantly in the next couple of years. On the defensive side of the ball, the low number of turnovers caused by the M defense is again, youth and lack of experience also combined with the fact that this is Greg Robinson's first year as DC. The turnover ratio should gravitate more to 1/1 in the upcoming years.

jmblue

October 29th, 2009 at 11:09 PM ^

Are you sure about some of those Michigan numbers? I remember Navarre throwing a lot more than five interceptions in 2001. I think he had more than seven in '02, also.

tybert

October 29th, 2009 at 11:22 PM ^

Didn't Navarre throw like 3 or 4 picks in that game alone. He was awful. I do remember some picks from that game at the game at MSU when Spartan Bob saved MSU.

We also lost at UWash, Navarre had a pick-six in that game.

Yes, we had a lot more INTs in 2001, good catch.

Navarre was a lot better in 2002.

tybert

October 29th, 2009 at 11:19 PM ^

The MSU game, I think 2 TO's, forced 3 (+1). Sloppy conditions, forced pass at the end. Not a game I worried about for TO's. Both teams made mistakes.

The Iowa game, I was there for that one. Iowa is living off TO's this year. Both INTs were rookie mistakes, forced into coverage. One fumble slipped out of QBs hands, can even happen to SRs. Minor fumble, held up by D and stripped. He was trying for extra yds. Matthews fumbled punt, simply a unforced screw-up. BTW on the Matthews fumble, we roughed the punter, so Iowa would have had the ball at the 35 anyway, got it at the 15 instead. We got one unforced pick-six so we ended the game at (-4).

The PSU game was butt ugly no matter how you look at it. D Robs pick was a silly pass. Dump it off, scramble, or throw it away. Better yet, if 3rd and 9, bring in Tate. Tell D Rob that 3rd and 9 is Tate down unless D Rob starts making better decisions. Brown's fumble was a killer. Just got the ball away blocked kick. Ball stinking security! A SR should know better. D Robs fumble, thrid time he has been hit and stripped. Again, two hands on the stinking ball! Late pick was OK. 4th down forced pass, so what.

I'm more concerned that we get no pressure on the QB (except for Graham), can't cover a lick, and get ZERO TO's from our D. The worst of all worlds. Other team's O's have NO FEAR of the UM D except for Graham and Warren. I'd rather play soft, drop into coverage and avoid the big plays if we don't ever get TO's anyway.

We're a HIGH-RISK, LOW-REWARD defense. GERG better fix it, even if he has to play prevent D for the rest of the season.

bigge1014

October 30th, 2009 at 3:07 AM ^

The unforced TO's are definitely what have killed us. The hits they have fumbled on were not all that hard. At this point I dont think Denard is strong enough to hold the ball vs hitting/stripping LB's twice his size. A year with Barwis should correct some of the ball dropping.

BlueTimesTwo

October 30th, 2009 at 3:08 PM ^

One of my biggest frustrations in the TO department has been turnovers following drops by receivers. At least two of Denard's bad 3rd down INTs were after 2nd down drops by receivers that would have resulted in 1st downs (one bobble by Odoms and a drop by Koger). When Denard is actually on target (or even close), the receivers need to make those catches. The drops don't excuse the picks that followed (3rd and long should be a Tate-only down), but more surehanded play by the receivers would place us in fewer 3rd and long situations.

swarwick33

October 30th, 2009 at 8:01 AM ^

I think we can attribute the uptick in turnovers to three things.

1) Mike Hart almost never lost a fumble. When he was healthy and they could keep him on the field, losing a fumble was incredibly rare.

2) We have had a complete inability to not fumble the kickoff or a punt. This has gotten a little better this year, and I am sure that last year was a fluke, but still it has to be included.

3) We have started Threet, Sheridan, and a freshman in Forcier. We have also seen significant playing time from another freshman (Robinson). Not having a proven field general usually leads to turnovers in the form of interceptions, bad snaps, and QB fumbles.

While we may never have another Mike Hart, in terms of ball security, the problems on special teams and at QB are on the mend. This should lead to a lower TO per game ratio in the future.

The thing that makes me more nervous is the fact that the defense does not seem to be able to force that big TO when we need it.

maizenbluenc

October 30th, 2009 at 9:11 AM ^

... the defense and the special teams.

I think the defense and special teams are more focused on angle to the ball and making tackles this year. At this point, lets just let them work out those kinks, and add stripping the ball or interceptions next year.

Go Blue!

NoNon

October 30th, 2009 at 9:40 AM ^

...go down as long as they learn to HOLD ON TO THE DAMN BALL on special teams.

Any type of spread comes with an increase in turnovers as opposed to a DeBored, ball-control offense, that's just common sense.

The question is which would you prefer to see. I liken it to the Florida Capital One Bowl where no one thought we had a chance in hell to win...Carr opened up the offense for a last hurrah, and despite an uncharacteristic amount of turnovers that game (as opposed to the rest of the season)...we threw the Gators off and won.

Raise your hand if you think Michigan could have won that game by playing as we did all year but never turned the ball over?

If we play like that when Rich Rod gets his system installed completely, I'll live with the turnovers.

jamiemac

October 30th, 2009 at 10:51 AM ^

I backed down a bit in the podcast, but I think this team, on the whole, is much better at ball security than last year's. And, its not even close.

Yes, we have 2 frosh QBs taking every snap. Wow, what a shock that they throw picks. I will take their picks and other TO's over last year's put the ball on the ground every fourth play and fumble every punt and kick approach.

Speaking of which.....only one fumbled punt and no fumble losses on KOR this whole season......if you tell me that back in July, I take it to the bank.

Brown has lost one fumble, IIRC, this year and Minor has only lost one.....didnt we all spend time last fall talking about Minor putting it on the carpet like every fifth carry or something like during the first half of the season.

It has taken 2 elite defenses to turn this team into the 2008 team as far as fumbling, bumbling, stumbling goes. Yes, those games count. And, yes, those games show that we have a long way to go to close the gap.

But, its night and day, AFAIAC, this year and last year with turnovers.

Shit, our starting QB still has twice as many TD passes as INTs.

(also, every stat I threw up there is going by memory. Please correct me if I am wrong. kthnxbai)

colin

October 30th, 2009 at 4:45 PM ^

and the TO/g in '08 was 1 TO every 4 games worse than '09. The talent level is obviously significantly different, which is probably where the seeming disparity comes from. The cost of a turnover for a team that doesn't expect to score a lot of points per drive is lower for a team that actually can put the ball in the endzone. The opportunity cost for the latter is much higher; the more talented the team, the more costly the turnover. So a turnover this year is more likely to feel backbreaking since we have a competent offense.

WanderingWolve

October 30th, 2009 at 10:51 AM ^

I think this is linked to our youth in the specialty areas. I would like to see a breakdown over the last two years of who gave the ball away for us. Minor and Brown have more fumbles than any of us would like to see (I wonder how many fumbles they have compared to other senior running backs in the country). Did anyone watch Va. Tech and UNC last night? UNC's game-winning field goal came off of recovering a fumble by Va. Tech's freshman RB. I guess the guy's a home-run hitter and was so used to flying through the hole untouched he never learned to secure the ball. I think he learned his lesson last night. Once again, we have to endure the learning curve on Saturdays.

IPKarma

October 30th, 2009 at 11:18 AM ^

TO stats? I mean, some plays in the game are simply going to result in a fumble. Others are due to players not protecting the ball.

Unfortunately, IIRC Michigan's TO woes are still primarily bad offensive plays.

jg2112

October 30th, 2009 at 11:53 AM ^

The vast majority of Michigan's turnovers came against Penn State and Iowa, two teams which have, according to numbers, pretty darn good defenses.

I'm not sure why this is a surprise.

Brown and Minor have one fumble apiece, both of which were forced by the defenses. I'm not sure why this is a quibble.

The QBs are true freshmen. In ANY OTHER YEAR, the mistakes you see them making are ones they would be making on a practice field, not in front of us on national television. Again, why is this a surprise?

Plus, Jamie Mac is right - Mathews has fumbled a punt. That IS IT for the special teams. How can anyone complain about that after what happened in 2008?