clock menu more-arrow no yes mobile

Filed under:

Paradigm Shift - Ohio University

3-0 is 3-0, but Tennessee has a lot of work to do.

We won't get to see this again.
We won't get to see this again.
Randy Sartin-USA TODAY Sports

We at the Paradigm Shift would be remiss if we did not begin by stating that the Vols are unbeaten, and most teams cannot claim that anymore. All of Team 120's goals remain ahead of it so long as it takes care of business. You know, Idiot Optimism and all that.

But we're in a tight spot.

But we're in a tight spot.

To say nothing about the numerous and costly injuries suffered by Tennessee before and during the game, Saturday afternoon was aesthetically unkind to the Vols (and the Bobcats, for that matter; these teams traded 10 punts during the second and third quarters). We will get into how each team fared statistically in a moment, but first there is one glaring statistical abnormality that appeared in patient zero Appalachian State, achieved outbreak against Virginia Tech in Bristol, and went full on pandemic Saturday against Ohio. I am referring to "fumble luck", or Tennessee's statistical improbability to recover fumbled footballs.

Fumble Luck

To explain why appearing to be really good at recovering fumbles is so odd, it is important to understand that fumble recovery tends to be random. That is to say that while you can coach effort and technique to recover a fumbled ball, fumbles statistically tend to wind up either with Team A or with Team B about half of the time, and fumble recovery rates for individual teams do not correlate year-to-year.

That means, for the purposes of modeling expectation versus reality, we can treat fumble recoveries like a flipped coin coming up heads. An outcome distribution of flipped coins yields a reasonable expectation of performance, but Tennessee is an outlier. Like nearly-four-standard-deviations-above-the-mean outlier.

That little red dot near the axis? Some lucky Volunteers.

That dot right above '15' on the x-axis? Some lucky Volunteers.

To date, there have been 18 fumbles in the games Tennessee has played. The image above shows a binomial distribution for the results of those fumbles. The dots represent discrete outcomes from zero recoveries to 18 recoveries. The dots' positions on the y-axis are their respective probabilities. The center-most dot, nine recoveries, is the expected outcome more than 18% of the time. In fact, the range of 7-11 recoveries is the expected outcome more than 75% of the time. To get all C-3PO on you, the odds of Tennessee recovering 15 of 18 fumbles are 320.3 to 1.

What we're trying to say is that Tennessee has been lucky to this point. Absurdly lucky. But they have also been extraordinarily careless with the football. It is impossible to know for certain if Tennessee's good luck recovering fumbles will continue, but it is unlikely that we can count on it. One thing we can say for sure is that the Vols would do well to take better care of the football. Which brings us to...

Turnovers and Explosive Plays


  • Tennessee - 1
  • Ohio - 0

Explosive Plays

  • Tennessee - 9
  • Ohio - 2

The Vols, lucky as they are, turned the ball over once through the air but were not able to do the same to the Bobcats. Fortunately, a dramatic advantage in big plays helped Tennessee seal the victory.

Average Starting Field Position

  • Tennessee - 28
  • Ohio - 29


Total Drives

  • Tennessee - 13
  • Ohio - 13


Positive Drives

  • Tennessee - 6
  • Ohio - 6


Scoring Opportunities

  • Tennessee - 5
  • Ohio - 5


Points per Scoring Opportunity

  • Tennessee - 5.6
  • Ohio - 3.8
3rd/4th down Conversion Percentage

  • Tennessee - 50%
  • Ohio - 32%

Oh, that's how we won. Touchdowns, not field goals, folks. The conversion percentages look very flattering for Tennessee, but the truth is a bit more complex. Ohio and Tennessee actually converted the same number of chances, Ohio just faced a lot more of them because they were able to drive the ball on occasion. We may take a look in the future at better ways to account for 3rd/4th down performance.

Thanks for reading, y'all. See you in the Post Script.