Methods
Survival Analysis
A family of statistical methods for how long something lasts before an event happens. We apply it to drives: how many yards a drive survives before it ends.
How we use it
Our drive model splits the field into zones and estimates, for each zone, how likely a drive is to end for every yard it covers. A drive that reaches the end zone, or runs out of time, has "survived" rather than failed.
Teams change those odds: a good offense lowers them, a good defense raises them. The same idea, with two competing kinds of ending (stalling and turning the ball over), drives both the ratings and the game simulation.
Where to see it
Related terms
- Drive Efficiency
- A team rating for how far drives get before they stall in a punt or field-goal attempt, adjusted for the opponents faced and the game situation. Zero is league average and higher is better, for offense and for defense.
- Turnover Propensity
- A team rating for how often drives end in a turnover (interception, fumble, turnover on downs, safety or blocked kick) for each yard covered, adjusted for opponents and situation. Zero is league average and higher is better.
- Bayesian Analysis
- A statistical approach that describes what we know about a quantity as a probability distribution and updates it as data arrives.