i'd like to hope that in some sense we can consider denard a freshman QB. ordinarily he would've redshirted last year and had that happened, this year we would have said "well freshman QBs turn the ball over."
Turnovers – The Year In Review
With a turnover margin (TOM) of –9 for the year and a national ranking of #110, the only conclusion that could be drawn is that TOM was a significant factor that caused Michigan to lose more games this year. That conclusion is absolutely WRONG. In reality, positive TOMs helped Michigan win as many games this year as negative TOMs contributed to M losing games!
Situational Analysis of Turnovers: At the end of the 2009 season, my youngest son and I were talking about the effect of turnovers (TOs) on college football games. We both agreed that looking at the total TOs for an entire season was stupid and irrational (sorry, Phil Steele). Intuitively, we also concluded that, even on a game by game basis, merely looking at the end-of-game turnover margin (TOM) without considering the situation that existed at the time each TO took place was lazy analysis with the potential for invalid conclusions. We both thought there would be a significant number of turnovers that would be meaningless (e.g. interceptions of a hail mary pass at the end of the half/game, turnovers during garbage time, etc.) Hence, was born the weekly situational analysis of TOs.
Sample Size: One team for one year is about the most minute sample size available. It is what it is. Most of the conclusions stated below are based on only the 2010 season for Michigan. However, IMHO many of the conclusions are valid across all of college football.
Conclusion #1: TOs Affect Virtually All Games. I do have game-by-game TO data for M for the past 5 years (60 games). Only twice (PSU this year and Minnesota in 2006) were there no TOs by either team. Thus, over 98% of all games are affected to some degree by TOs. Over the past 5 years, M had no TOs in 11 games and the opponents had no TOs in 11 games. A team can expect that it will give away the ball at least once in 80% of the games but can also expect to take away the ball at least once from the opponent in 80% of the games. There were 15 games with a TOM of –0-. About 75% of all games will end with a TOM of at least 1 while 25% of games will end with a TOM of –0-.
Conclusion #2: Almost All TOs Are Situationally Important. This was one of the surprises (at least to me). There were a total of 45 TOs in Michigan games this year. Of these, only 8 were basically meaningless (4 because an M interception was followed on the same play by an M fumble and 4 because the game was in garbage time). Thus, 80 % of all TOs occurred at a time when the outcome of the game was still in doubt. This is probably understated since the probability that an interception will be followed by a fumble on the same play has to be very small (even though it did happen twice this year). Without these 4 TOs, approximately 90% of all TO were situationally important.
Conclusion #3: Overall Turnover Margin for the Year Tells Us Very Little. Here is the TO summary for the year.
Without knowing when TOs occurred during the 12 game regular season, this means virtually nothing. It does not tell us how many games ended with a positive or negative TOM for Michigan. It does not tell us if TOs were significant in any games. It does not tell us if TOs helped win some games, lose some games, or had no impact at all.
Conclusion #4: Positive TOMs Helped Michigan Win As Many Games This Year As Negative TOMs Contributed To M Losing Games. M had 2 games with a positive TOM and the TOs helped win both games (UConn & ND). Michigan had 6 games with a TOM of –0- and won 4 of those games (UMass, BGSU, Indiana, Purdue) while losing 2 (Penn State, Wisconsin). M had 4 games with a negative TOM and actually won one of those (Illinois with a TOM of –4). Of the other 3, TOs were a primary factor in 2 losses (MSU and Iowa) but in one (osu) it is doubtful TOs were a significant factor in the loss. Therefore, if all games had ended with a TOM of –0-, it is likely M would still be 7-5 but with losses to UConn & ND and wins against MSU and Iowa. (And, yeah, that would be a lot better.)
Looking at the game by game analysis, if M had been able to eliminate the TOs in the MSU and Iowa games, an overall record of 9-3 would have been possible even though the overall TOM for the year would still have been –2.
Game By Game Detailed Analysis:
This is an excellent example of the importance of situational analysis of TOs. Based on the final score of 30-10 and a TOM of just +1, the initial conclusion would be that TOs were NOT significant in this game. However, the TO occurred at 2:29 of the third quarter with M ahead 21-10 and UConn going for a 4th and 1 at the M7. UConn had gotten the first down when Floyd forced a fumble at the M3 that was recovered by Ezeh. Without this TO it is likely that UConn would have scored a TD bringing the game to 21-17 with the momentum shifting to UConn. The TO was a significant factor in the Michigan win. This is also the game that M blocked a UConn FG attempt (blocked FGs are not counted as a TO).
This game is fairly obvious. Without the 3 TOs, M would likely lose the game. The first TO was an interception by Mouton at the ND40 and returned to the ND31. M scored on the next play to tie the game at 7-7. The second interception by Floyd at the M37 stopped an ND possession with the score 14-7 Michigan. The third by Kovacs was at the ND 35 and returned to the ND25. M did not score but punted to the ND4.
These 5 games ended with TOM of –0-. The UMass game had the first meaningless TOs – the M interception followed by a lost fumble on the same play. The other 2 TOs in the UMass game basically offset one another. The BGSU game was really never competitive and I considered all 4 TOs as meaningless (this is the other game that had an interception followed by a lost fumble on the same play). The Indiana game had 2 huge TOs (M fumble at the goal line and Indiana intercepted in the end zone) but they did offset one another. There were no TOs in the Penn State game.
Although the Wisconsin game had a TOM of –0-, M missed a FG and also failed to recover 2 onside Kickoffs that were very recoverable.
The Purdue game had the next meaningless TO (M interception of a hail mary pass at the end of the half). Therefore, this game really had a TOM of –1 for M. I was fracking nervous for the entire game. There were so many TOs that the game could have gone either way with 2 TOs retuned for TDs, a fumbled punt leading to a TD, and an interception leading to a FG. All this pretty much ended with the TOs not impacting which team won the game. My gut feel is that the game was closer than it should have been because of all the TOs.
I have no doubt that both the MSU and Iowa games were decided by TOs. Two interceptions in the endzone when the receivers were open and a third interception with 12:23 left in the fourth quarter with M trailing by 14 points (31-17) were the primary reasons for the loss. The Iowa game had a meaningless TO at 1:54 left in the fourth quarter and M trailing by 10 points. The other 3 TOs and another blocked FG were key reasons for the loss.
Michigan defied the odds in the Illinois game by winning with a TOM of –4.
The final score in the osu game makes it difficult to conclude that TOs had an impact. Michigan needed a well played game plus a positive TOM to make this game competitive. Neither happened.
I'll take all of his mistakes this year, as, for the most part, they were from a combination of inexperience and aggressiveness.
He'll learn from his INTs, watching film during the bowl practice weeks and off-season.
The big change we need from him is ball security. He almost always carries the ball in his left hand. Teams have figured that out and did a good job hitting/stripping him in games. The Purdue fumble on the 1st drive, the hit from his left side at the 10 in the OSU game, etc. wouldn't have happened if he had switched the ball over. He may not be comfortable with the change in hands, but that can be practiced in the off-season.
One guy who needs to do a better job is Vincent Smith. He's small and coughs it up when hit. We got lucky in the 2nd TD of the Purdue game when Lewan recovered. The fumble at the 14 in the Iowa game and before halftime vs. OSU were big. Both killed drives at a time when we really needed a score.
We definitely started out taking better care of the ball. (Against lesser competition of course...)
I'd be interested to see what this kind of analysis shows with T.O.P. I don't think TOP is something that you can as simply improve like you can with TO margins. But it just seemed to me watching games that we really struggled when we couldn't give our defense time to rest. Wincy game comes to mind especially.
Those are called a turnover on downs for a reason. Also, all turnovers inside the opponents 30-yard line represent a lost opportunity to score points.
IMO, turnovers in the MSU, Iowa, and OSU games was a primary contributing factor in those losses.
Horrendous defense was the primary culprit against PSU and WI.
You have taken a single season for one team (a very small sample size) and then layered onto that a somewhat subjective determination as to what constitutes a situationally important turnover and what does not. The bottom line is that turnovers matter. They actually might be the single most important factor in determining the outcome of a game.
Below is a breakdown of the best and worst teams in turnover margin for the 2010 season. The information is taken directly from the NCAA's website. Unless the numbers for previous years are vastly different (and I doubt they are), clearly turnovers matter.
|100||San Jose St.||-0.54||1||12|
|103||San Diego St.||-0.58||8||4|
|109||Fla. Atlantic Univ.||-0.75||4||8|
Well, let's take a look at osu. Do you really think they won 11 games because of their TOM? They only had 3 close games (Miami, Illinois, Iowa). In the Miami game they were +3, Illinois was -1, and Iowa was -1. In the loss to Wisc, they were -0-. Most of osu's large TOM came in blowout games and the TOs had nothing to do with the wins. Marshall +3, Ohio +3, Minn +2, U-M +2
For Wisconsin, they were +6 in the NW game that was a 70-23 blowout, +3 in the Indiana blowout, +3 in the Purdue blowout.
For these two teams, TOM had NOTHING at all to do with them winning games.
If the turnover margin would have been 0 in those "blowouts" there is a good chance the game isn't so lopsided. In the close games, perhaps that's the difference between an undefeated OSU and a 1 loss OSU.
Your simply looking at victory margin and saying this game was close despite turnover margin rather than this game was close because of turnover margin. The data you showed is pretty good evidence that turnover margin strongly plays into victory (or deficit) margin.
I'm not saying turnover margin means everything, but as sure as hell means something, a whole lot of something to be exact. Some teams are good enough to overcome it, most aren't, and all would be much better off (in terms of wins and winning margin) had they had a better turnover margin.
I absolutely agree that TOs benefit the team that gets them. It would be silly to think otherwise.
But, the real queston: "Did the TOs help you win the game?" It is obvious that TOs do NOT impact the final result of many games.
When I looked at TOs in 2009, there were wild swings in TOM for many teams with no corresponding change in win/loss record.
One of my favorites was Florida: 2002 TOM -9 & 8-5 record, 2003 TOM +7 and the same 8-5 record. This is a swing of 16 TOs with no difference in record!
Or Florida in 2005 TOM +18 & 9-3 record, 2006 TOM +5 & 13-1 record. A swing of 13 TOs less and 4 more wins!
To me this screams that Situational Analysis on a Game by Game basis is the only sane way to draw conclusions about TOs.
Thanks for tracking this all season
I've said it a lot, but because its true. With the turnovers this season, the team really paid the price by going with a first year starter for a third year in a row. I think the ends justifies the means, in this case.
I know everyone is tired of it. And sick of the excuses. And hate hearing things over and over again
But, this will get better as the quarterback becomes an upperclassman and the team as a whole around him continues to gain experience. Thats pretty much how it works.
If our TO margin ends at -10, the silver lining is historically speaking, this foreshadows a better record next year. Teams with a double digit TO margin have improved the overall won/loss record 69 percent of the time since 1996 (80 percent same record or better). Michigan fell into this category after 2008. And improved its record the next year. Michigan fell into this category after 2009. And its improved its record the next year.
It will happen again in 2011, the record boost, and if it comes with (finally!) an improved TO margin, then this outfit will contend for the Big 10 Title.
(stats courtesy of Phil Steele)
Few other thoughts
*I like the situational TO breakdown. Not doing so is one of the problems I have with Luck metrics
*TOs are still important if you think one third of the outcomes and the direction, arguably, of half the games played came down to it.
* We tend to focus on the offense. MIchigan needs to generate more takeaways on defense as much as anything to get the TO margin stat on the rebound.
Yeah, it's both offense and defense. Three straight years deep in the negative range is starting to each my soul. Hopefully a better secondary and an experienced QB will get us back to even.
In the MSU game, the 1st EZ pick, on a 3rd-and-4 ball thrown behind Roundtree, was a killer. More than ANY T.O. this season, that one had an impact to a game and perhaps a better season. If I could re-wind the season and get one free do-over it would be to go back in time to that 3rd-and-4 play. Up until then, we looked like we were going for a TD on the drive. I talked with an MSU friend about things after the game...he felt sick to his stomach on the 1st drive and was shocked with the INT. He felt Sparty may not have responded well after a swift TD drive and a fired up home crowd. At home, 5-0, and opening drive TD changes the complexion of the game. The 2nd EZ INT was bad, but people forget we were going to get hit with a chop block PF (someone chopped Worthy) that would have pushed us out past the 20 with 3rd-and-forever and no kicking game. Even the long bomb INT was in a situation where we were down by 14.
Iowa was just a complete goat frick game. Iowa had the gift of turnovers all season and still ended with the same record (7-5) and a crappier bowl game (with a more lopsided match-up against them with Missouri). The kicks OOB had a huge impact, as did our True FR DB whiffing on a tackle that Troy or Donovan (if he had stayed) make and force a punt down by 7 with plenty of time to score.
Two "non" turnovers killed us in the PSU game. Down 14-10, we had Royster stopped in the backfield on a 4th and short --- that would have been a huge lift for our D. He spun out of the tackles and picked up the 1st. They scored a few plays later. Then all-do-nothing-returner Jeremy Gallon runs over to the sideline and tries to catch a kick at his 2. Ball bounces out of bounds, we go three and out with a roaring crowd, get a crappy punt, then give up one last TD before halftime. What should have been 14-10 with our ball and Big Mo on our side...turns into 28-10.
I'll concede the Wisky and OSU games. We are not even close to them in talent, size, etc.
Momentum is huge in CFB. When the our D gets a turnover it moves the mojo meter in a positive direction. However, an Offensive turnover is 3x as powerful on the mojo meter in the negative direction (this is based on a double blind scientific test based on the amount of cuss words directed at the offending player).
Conclusion: To have a positive mojo meter reading at the end of the season, it takes 3x as many takeaways to balance out the level of negative mojo created by @#$%ing turning the ball over.
the title of your diary and your user name definitely do not agree with one another in the case of this football team. Reading "Turnovers - The Year in Review" followed by Enjoy Life just makes it that much more depressing