What Turnovers Really Mean

Written by OKC Dave

Photo Attribution: EK Nielsen

When I wrote about turnovers before the season started, I promised I would keep a close eye on them as the season progressed. I think it’s time to take another look.

In 2011, OSU gained a turnover on 4.0% of defensive plays (3rd in the country). The NCAA average is around 2.5%, so while it took an average team 40 plays to generate a turnover, our defense produced one every 25 plays.

As it turned out, last season’s offense was also pretty decent at holding on to the ball. We turned it over on 2.3% of offensive plays (49th in the country). Not great, but above average.

This gave us a per-play margin of 1.7% (5th in the country). That alone would give most teams a great chance at winning a lot of games. Look at the chart below. It shows the relationship between a team’s per-play turnover margin and winning percentage. The two have a fairly strong relationship (67% correlation). The line running through is a simple closest-fit linear trendline. It shows that a team with a +1.7% margin should have a winning percentage in the 75% to 80% range based on turnovers alone. Obviously, our team last year overachieved due mostly to our explosive offense.


Okay, so how are we doing this year? (Hint: not as good)

This year, we gain a turnover on 1.6% of defensive plays, or 1-in-62 plays. (104th).

We lose a turnover on 2.7% of offensive plays or 1-in-37 plays. (83rd).

Our per-play margin is -1.1% (104th).

Turnovers have cost us dearly in our losses, but our team is still overachieving the trendline. A -1.1% margin tells us that a team should have a winning percentage in the 35% to 40% area. We are sitting at 67% right now (see the orange dot in the chart).


K-State is having a very similar season to our 2011 adventure. They are leading the nation in per-play turnover margin at 2.8% (the dot closest to the top right — where you want to be — is K-State).

One more thing. I was curious to see if our change in per-play margin from 2011 to 2012 was the largest. It is. Last year we were +1.7%. This year we are -1.1%, so the change is -2.8%. UCONN is tied with us for the largest fall.

  • Jeff

    Great stuff. Thanks

  • Awesome work as always!

  • Chicago Poke

    So what does this really mean…? We were extremely lucky last year as every ball bounced our way? And we are very unfortunate this year but overachieving? Or is this simply the way things go in college football? I tend to think it is a little of both. Young qbs throwing a TON of picks (think if all the pick 6s that have been dropped as Kyle has pointed out) and a defense that hasn’t had the same fortunate bounces as last year’s squad.

  • Jr

    Man those grey jerseys look slick!