Turnover Ratio is Neither Consistent Nor a Defensive Stat

Yesterday, Danny Kelly considered an important aspect of Pete Carroll's coaching approach both in college and with us: getting the ball. An insightful article, worth a read, and one that certainly has seen enough evidence in both what the Seahawks coach says and what he does, with weekly days dedicated to turnover drills.

And before I dive into anything else, let me make one thing crystal clear: I don't disagree with the focus on turnovers, because there's little the correlates quite as well with wins and losses as turnovers, or rather getting more of them than the other guys. It's worth focusing on as a coach, and I'd imagine all of the best teams in the NFL do. There's a reason why someone like Antonio Cromartie has a place in this league other than pure athletic upside, it's because his tendency to give up big plays is offset well by his tendency to make big plays. It's not ideal, but it is acceptable.

But there's a dose of realism needed here. Every team in the NFL values the turnover ratio. Some maybe more than others, but no one ignores it. But no team ever quite mastered the turnover ratio to a consistent degree. There is no such thing as consistency in turnovers from year to year as an NFL organization. And what consistency there is, is not a result of defensive play. After the break, let me chart two of the best organizations with franchise QBs (Colts, Patriots) and two of the most dominant defenses of the 00s (Ravens, Bears), and a "model franchise" with a dominant defense that would later add a franchise QB (Steelers).

Three things immediately stand out: 1) It is very difficult to maintain a high positive or negative turnover ratio in the NFL, the outlier performances in this chart are usually followed by regression to the mean or outliers in the other direction, Manning at his peak being the exception 2) even for the best, multi-superbowl appearance franchises like the Colts, Patriots and Steelers, there is little year-to-year consistency in turnover ratio. 3) any consistency in turnover ratio does not correlate well with great defenses, but rather with great quarterbacks. The Colts and Patriots are not consistent but they are nearly always positive, the Ravens  appear to be on the right path with the addition of Joe Flacco.

Pete Carroll seems to know this. He has stressed the need for the offense to keep the turnover ratio down more than once. He has put pieces into place to make it easier to keep turnovers down (throwing a ball inside Sidney Rice's reach but outside the CB's reach is easier than throwing a ball outside Deon Butler's reach but outside the CB's reach).

But we're missing that most important piece, and the QB we're plugging in has never been a consistent ball security guy. Our machine overlord, Football Outsider's KUBIAK, projects Tarvaris to have about 19 INTs (and Whitehurst about 15, but with a lack of actual reference material on him that's largely meaningless). That projection sounds pretty realistic, and with a low amount of attempts that is unacceptable - KUBIAK has him at 495, again, realistic, maybe even on the high side.

While I do agree turnovers aren't just random, and certainly fumble-forcing-and-recovering techniques can be trained, it seems very unlikely for the Seahawks to turn the corner to consistent positive TO ratios based on defensive training alone. With Hasselbeck at the helm from 2008-2010, the Seahawks posted TO ratios of -7, -8 and -9. That's an important corner for Pete to turn to build a winning team. But even if we improve to, say, a TO ratio of 0 (around the highest realistic number with our QB situation), historic evidence suggests one can not build a consistent turnover-winning team through the defense. Historic evidence also suggests that once you have your franchise quarterback, even then your turnover ratio will be all over the place from year to year, the best you can hope for is it's always positive.