I have often stated that our own turnovers and penalties, and unfavorable
decisions by referees are all "part of the game," like boundary lines,
turf, goal posts and weather. They must be overcome as well as the
opponent in order to win consistently. All the greatest teams do. No team
is perfect, but great teams overcome these adversities.
After observing the difference between Golson's outstanding season in
2012 and his performance this season, I wanted to study the consequence
of turnovers on the game outcomes.
Of course, it is self-evident that our fumbles and interceptions are
undesirable, but I have wanted to see more quantitatively how they
directly affected the outcomes of our games.
Inasmuch as turnovers and penalties are unplanned, and since I have not
yet discerned objective metrics for their causes, I will regard them as
stochastic for our purposes here. In other words, assuming that not all
teams that would like to force opponents into turnovers can necessarily
I limited my population to Notre Dame football only, and only to our last
186 regular and post season games (2000-2014 seasons).
I calculate turnover margin as the difference between the number of ND's
turnovers and our opponent's.
Since the outcomes of games depend on so much more than just turnovers, I
am keenly interested to have seen a strong historical correlation between
the turnover margin and the margin of victory in ND's games since 2000.
These results reaffirm that possession of the football is of paramount
importance to deciding the outcome of the entire game, not just to an
The data shows that each count in turnover margin has correlated to 12.9
points in the margin of victory or loss.
The data shows that fumbles and interceptions are equally painful. NO
fumbles and NO interceptions during a game has been a relatively sure
path to success.
In 13 of the 186 games we won the turnover margin, but still lost the
game. In 27 of the 186 games we lost the turnover margin but still won
the game. In the 24 games in which there were no turnovers by either
team, or the turnover margin was equal, we lost 12 games and won 12 games.
Some interesting results (2000-2014):
• Overall record: 115 - 71 (.618)
• Where turnover margin = -0-, ND record = 12 - 12 (.500)
• Where ND lost NO turnovers, ND record = 40 - 5 (.889)
• Where ND won NO turnovers, ND record = 11 - 20 (.354)
• Where turnover margin > -0-, ND record = 74 - 14 (.841)
• Where ND fumbles > -0-, ND record = 41 - 46 (.471)
• Where ND int's > -0-, ND record = 50 - 55 (.476)
I will expand this survey to as much of ND's history as I can collect
data, and will perform more supervised segmentation, such as relates to:
1.) Time of possession
2.) Rushing and passing
3.) Series for individual coaches and players
4.) Series for opposing teams
I would like to expand this survey to other schools as well, but only
after I discern causal correlations that will make more in-depth research
beneficial. If I detect any errors I will revise and release.
1.) Not all turnovers result in immediate points scored by the recovering team.
2.) Sometimes a team recovers its own fumbles.
3.) Risk of more fumbles increases as the number of rushes increases.
4.) Risk of more interceptions increases as the number of passes increases.
5.) Turnovers occur with special teams.
6.) Turnovers can be unforced by the opponent.
7.) Occasionally more than one turnover occurs during a single play.