Turnover Statistics Only Partially Repeatable

Note: Long time reader Nik Greene sent the piece below to me and asked me to post his cursory thoughts on turnover statistics. – Dave

By Nik Greene

I\’m really fascinated by how much a defense can control turnovers, and Dave\’s article about Dick LeBeau and turnovers, and it inspired me to look into the matter more closely.

Let\’s look at fumbles. For the statistic of opponent fumbles, the San Francisco 49ers led the league as their opponents fumbled 1.9 times per game. The 49ers then recovered about 50% of those fumbles (middle of the pack statistically). On the other hand, the Steelers opponents fumbled only 0.9 times per game (third to last), and Pittsburgh recovered only 33% of the time. (Interesting side note, this statistic does not take into account turnovers that occurred because of pure error on the part of the opponent offense or special teams: a botched snap for instance.) If the 49ers had only recovered 33% of opponent turnovers, like Pittsburgh, they would have had 5 less takeaways last year (a significant drop over an entire season). I\’m sure it can be debated how much control a team has to recover opponent fumbles, however, these statistics tell me that a team’s takeaway total is affected by chance.

Interceptions paint a slightly different picture. The Green Bay Packers led the league with opponents throwing 1.9 interceptions per game. In addition, the Packers intercepted 4.77% of all opponent passes, highest in the league, and 2% higher than the league average. This might make someone think that interceptions are partially chance, however, Green Bay has been in the top four of opponent interception percentage for each of the last four seasons. This causes me to think that interceptions are less reliant on luck, and more a indication of quarterback pressure and defensive back ball skills.

Ultimately, I am led to believe that turnover statistics, good or bad, are only partially repeatable, and Pittsburgh fans should not be too concerned. In addition, consider that in the last five years Pittsburgh has finished in the top ten in total team takeaways two times, and in the bottom ten in total team takeaways three times. The Steelers record during that time was 55-25, with two Super Bowl appearances. Maybe turnovers aren\’t that great of a measure of team success and we should pay more attention to what really matters – team wins.

Of course, this is only a cursory investigation into the matter. I encourage others to do more research and share their results.

  • 8forty

    I’ll take a 12-4 regular season over good take-away stats any time.

  • Kenneth Wilt

    When you look at fumble and fumble recoveries, there are sooo many factors it is hard to really give a good indication of what really happens. How? When? Where? Who was close by?

    Interceptions are mistakes typically. Without Harrison and Woodley for large stretches last year, it doesn’t surprise me that these numbers were down. As we get pressure back on the QB more this year combined with more athletic corners in Lewis, Allen and Brown playing larger roles, I expect this stat to improve.

  • Dcpfjr

    I’d rather see a 12-4 team, like last year’s Steelers, rank high in defensive stats (low yards/gm, low points per game) with low turnovers, than a 12-4 team that creates a lot of turnovers, but has higher yds/gm and pts/gm. Why? Because if you rely on turnovers to win games, then eventually, especially in the playoffs, you’re not going to get them, and your defense will be exposed for what it really is.