Todd Grantham is on to something.

“In pro ball you don’t have to recruit, so you’ve got to do something,” Grantham said. “So you do stats.”

It turns out in his spare time Todd Grantham is a stat-head.  And the stat he’s focused on at Georgia is turnover margin.

That seems to have paid off.

… On defense, Georgia couldn’t go anywhere but up after 2009. The Bulldogs recovered two fumbles, last in the nation, but have grabbed 10 each of the past two seasons.

It forced 18 in 2009, 22 in 2010 and has 29 already this year.

“We need to recover some more,” Richt said. “When you’re knocking the ball out, you’re going to get a chance for turnovers.”

The Bulldogs have gone from 76th in the nation in interceptions gained (10) to 27th last year (16) to 15th this year (13). Georgia starts each practice with three minutes of ball disruption drills.

“We’re putting pressure on quarterbacks so DBs can get interceptions,” said defensive end Abry Jones. “Just group tackling and really getting the ball out of there.”

Georgia has gone from minus-16 in turnover margin in 2009 to +19 in the past two seasons combined.  Grantham believes the current eight-game win streak is no coincidence.

… The fact that Georgia’s current winning streak coincides with its turnover success — it came out ahead of its opponent in six games with Mississippi State and Tennessee even — is consistent with what Grantham found out in his own study when he was an NFL assistant.

Grantham said a minus-one turnover ratio meant a team’s chances of winning were about 20 percent and if a team was plus-one, a team’s chances of winning are at least 80 percent.

He said he saw Georgia’s defensive turnover statistics when he arrived, but didn’t look at reasons for them.

“I just knew that it was critical that we got turnovers and we were in the plus-margin,” he said.

And before you yell about the anomaly of the 2010 season – plus-10 in turnover margin and a 6-7 record – look at the game log.  The Dawgs were 6-1 in games with a positive turnover margin and lost all four games with a negative one.


Filed under Georgia Football, Stats Geek!

44 responses to “Todd Grantham is on to something.

  1. It’s amazing how defensive pressure with an athletic front seven will help there. The turnover is the result of aggressive defensive play. Overall scheme (43 vs. 34) doesn’t matter. Athletes combined with a GATA attitude make it happen.

  2. Zdawg

    I still think the run d is the difference between this year and last.

  3. This is the kind of stuff that gets my blood pumping. This is being proactive by monitoring your defensive statistics and recognizing where you can improve then making it happen.

    It seems obvious you need to have a better turnover margin than your opponents. The thing that makes this interesting is how Grantham has been able to make it happen. The article mentions quarterback pressure, but I also have to thing that the technique they are teaching the DBs has made a difference as well. Heck the techniques they have implemented throughout the defense.

    There must be a way to create a statistic based on running yards allowed, passing yards allowed, sacks, tackles for loss, interceptions, and fumbles. I’m too lazy though.

  4. You know just looking at Grantham, I never would have taken him for a stat nerd.

  5. Bad M

    Seriously, what the fark was Willie doing?!

  6. Mayor of Dawgtown

    Gee Senator, I thought you said repeatedly on this blog that turnovers were totally random. How can an improved UGA defense “force” totally random events?

    • Find where I said that, Mayor.

      The only thing I’ve pointed to as being totally random are fumble recoveries.

      • BenG

        Yes, and if Team A forces more fumbles than Team B then that means that Team A has more chances to recover a fumble than Team B does. So even though recovering a fumble once it happens is random luck, forcing more fumbles is not, and it can contribute to a better turnover margin.

      • Mayor of Dawgtown

        May I assume from your statement above that you agree that hard-hitting, aggressive defense causes fumbles? If so, would you also agree that, all other things being equal, hustle aids in fumble recoveries and a team that hustles more has a higher likelihood of recovering fumbles than a team that does not hustle as much?

        • Yes to your first point.

          As to your second one, “all other things being equal” as a qualifier leaves a hole big enough to drive a Mack semi through. Fumble recoveries are random events. Hell, hasn’t watching the last five years of Georgia football shown you that?

          • Mayor of Dawgtown

            No. Watching the last 5 years of UGA football proves that the teams with a good attitude who hustle get more fumble recoveries than the teams that don’t. Also, trying to pick up a ball and run with it in a crowd and batting it around until it goes out of bounds or a player from the other team recovers it instead of just falling on the ball results in fewer recoveries. The reason why we have a better turnover ratio now than when CWM was the DC is, in a word–Grantham. Better coaching=more fumble recoveries. Period.

            • austintwo

              I sure was glad Sean Jones didn’t just “fall on the ball” in Knoxville in 2003 ….

              “Scoop and scores” in a game correlate greater than turnover margin with victories, which is why our team practices them.

              Overall, I agree the recovery ratio is a function in some way – over time – of good coaching, hustle, running to the ball to tackle, etc., though like Senator I don’t know how to prove it.

          • Cojones

            Yes and some years we were unluckier than other years. We apparerntly are not only making more balls hit the ground, but we’ve gotten luckier. Perhaps the Mayor is implying that it is “made” luck thru more hustle after the ball is on the ground. Unlucky to me means the ball bounced away from Dawgs and they never had the chance to recover it, not that they didn’t hustle faster than the other team.

            • Look, if you guys can find a statistical study that shows fumble recoveries aren’t random, lay it on me. I’d love to see it. Hell, I’d love to post it.

              All kinds of folks I respect have looked at this and come to the same conclusion, that they are random from season to season.

              • Newt

                I think the point they’re trying to make (and I tried making below) is that while the bounce of the ball may be random, your chances of recovering go up the more bodies you have around the ball. I’m sure there’s no study to prove it other than the eye test. I know from watching our defense this year that we’re getting more players to the ball on defense and that gives us a better chance of falling on the ball when it comes loose.

                • Then that should mean as a rule, good defenses recover fumbles at a higher rate than bad defenses. Not the case.

                  • Dboy

                    Senator, I am of the opinion that “good defenses” that are properly coached should create more fumbles than a bad, poorly coached defense. Then, if 50% of the balls that hit the ground are recovered by the defense, there is a net increase in fumble recoveries. The situation that you talk about being up to the bounce of the ball only applies to compeltely random fumbles or fumbles that are 100% screw-up of the offense or not reliant on the Defense disloging the ball. Take Crowell’s 2 fumbles Vs Auburn. One was a screw – up by the offensive player Crowell – a random occurance. The second fumble was forced by the defense being in Crowell’s grill as he got the ball. What if AU had done that 2x more times and forced 2x more fumbles by crowell… the odds say the AU defense recovers one of the two and the final turnover margin changes in the game to even instead of -1 for AU.

                    Similarly, good defenses force the opposing QB into situations where INTs are more likely to happen (i.e. pressuring the QB, confusing the QB, Reading the QB / play) based on preparation.

                    I think turnovers are partially random… but many are created.

                    • What odds?

                      Again, I’d love to see a statistical study that backs up assumptions like this. To my knowledge, there isn’t one.

                    • Mayor of Dawgtown

                      Senator, if you saw a study that said the earth was flat would you then believe the earth was flat? There are flawed studies out there about every topic you can imagine. The concept that recovering fumbles is totally random goes against everything I was told by every coach I ever had and it also defies logic. If that is the case why do some teams consistently, year after year, recover more fumbles than they lose? Under the “random” premise that would be impossible.

                  • Mayor of Dawgtown

                    It is impossible to statistically measure motivation, i.e. “hustle.” That said, if there is a ball on the ground and 2 players, 1 from each team, go after it the player who goes after it the quickest and the hardest is going to recover it more often. Is there an element of luck involved? Sure. Is luck the primary reason the player who got the ball ended up with it? Hell no.

        • James Stephenson

          Plus you can not account for stupidity either. Like the Tenn Fumble recovery that wasn’t. No way the Tenn player had that ball between his legs.

  7. Mike Cooley

    I’m with the guy that asked what Willie was doing was doing all that time? Rambo spoke up and mentioned lack of intensity. I would love to be able to talk to a player off the record and find out what the real differences are between this defensive staff and the old one. Obviously this one is much, much better. But what did Willie and Jancek have them doing? What was their time spent on? I remember people moaning because we lost Nikel Robey to USC because of the defensive house cleaning. He’s looking good at USC but what do you think he would have looked like under Martinez? I have two words for you; Bryan Evans.

    • Cojones

      Not only coaches, but player issues were happening. I’ve consistently posted for several years that problems started with a large number of injuries, players had to be retrained to positions they weren’t proficient in, coaches had to share players and when it came to game time the cobbled patchwork caved because they ran into situations that weren’t practiced nor forseen(i.e. Bryan Evans, etc.) . That was trhe start of coach and player dissention that we saw acted out during game play. Confusion literally rained(reigned). It was also occurring off the field , the cancer started and the operation took place last summer. We are now here, sans some coaches and players ,who, under other circumstances might still be with us. That down cycle was slow to heal and the effects lingered. We are all better now.

  8. Newt

    The biggest factor, to me, in the defense’s ability to not only force, but recover more fumbles is that we’re getting more players to the ball. More tacklers hitting a ball carrier equals more shots from different angles to knock the ball free and more bodies close by to fall on it. Just one more benefit of a good scheme and a GATA attitude.

    • yurdle

      Fumble recoveries are awfully close to random. On average, the only way to recover more is to force more, and the only way to lose less is to fumble less.

      • Newt

        Agree to an extent. Yes, the ball bounces funny, etc. But you can’t tell me that getting more bodies to the ball on defense doesn’t play some role in forcing more fumbles and going from increasing your recovery percentage from just over 10% to over 33%.

  9. Bright Idea

    Look for Baylor to give Willie’s secondary a fit Sat. night like TT did. Okie State will light them up. Amazing how there were so few interceptions during that error.

  10. Cojones

    Hope Kentuck’s Ga entourage don’t do the same to us tomorrow. Richt is keeping the troops focused, but sometimes other teams play beyond their usual perceived limits. If they give our O big trouble, make fumbles happen, intercept(isn’t one of their LBs an SEC leader?) and run back a couple, it might not matter how well the D plays.

  11. This article made me think about the Georgia Tech fumble situation. There fumbles have gone up since PJ has taken over. Some think it is due to the fact more people touch the ball in his option oriented offense. I’m not sure how true that is, but they had 37 fumbles last year and lost 20. Wow.

    They have had at least 34 fumbles each season since PJ took over. This year seems to be the exception. They are sitting at 23 with three games to play (including a bowl game). They have a good chance at reaching 30 fumbles. I hope we help them out with that.

    Agree with the Senator. Fumble recoveries are random. They only thing you can do is to encourage good habits handling the football to reduce your fumbles.

    One thing that always bothers me is when I see RBs running to the left and carrying the ball in their right arm. My peewee coach told our RBs to carry the ball to the outside. They also focused on gripping the ball properly. That is the kind of thing that disturbs me. Is it that they don’t know or they don’t care that carrying the ball on the inside arm increases the chances for a fumble?

    • X-Dawg

      Thinking about the other side of the ball for GTU: In the VA Tech game last week, the Hokies seemed to pick up a lot of extra yards after intial contact while the GTU defense was trying deperately to cause fumbles by batting at the ball carrier’s arms. In one instance, a VT runner gained an extra 10 yards while the GTU defender was running along beside him using this “technique”. It was high comedy

    • Cojones

      If you saw the Va T/NC game last night you noted that Va T had a bunch of fumbles and got about 90% back. They were lucky. Many weren’t caused by NC. Those that were fell right near their feet, but Va T luckly saw it first and pounced, even though NC was playing hard. On a punt return, the receiver lost the ball, was confused where it was, even though it was closest to him, it rolled further away and Va T jumped on it while NC players were in the air. Straight dumb luck. And a NC recovery could have possibly won the game. Va T didn’t look like a top10 team last night, but won anyway.

      The point is that fumbled balls recovery appeared to be luck only and Va T was lucky to win.

      • I think people like to believe that their favorite sports teams are truly in charge and want to believe in intangible concepts such as being “clutch” or a team making its own luck. The reality in a team sport like football where there are constantly so many moving parts is that pure dumb luck plays just as big a factor in the outcomes of most games as does the preparation and ability of the teams on the field, but it’s hard to convince a fan this is the case.

      • I never think VT looks like a top ten team, but they just win. Congrats to Beamer for getting #250 last night.

  12. W Cobb Dawg

    CTG may be a ‘stat guy’, but he sure as heck can coach. Next time we have an opening for an assistant coach, we should search far and wide for a clone. If the candidate doesn’t talk, act and think like CTG, don’t hire him.

  13. Scooter

    Has anyone kept a running total of interceptions that we dropped this year? I swear it has to be another 4 or 5 turnovers that we simply let slip away.

  14. Cojones

    Maybe Coach G has hit upon a lucky theory that works. Wonder if he has taken a look at string theory.