By Thomas Negron
This past Sunday, Bill Belichick did the unthinkable, he chose to kickoff in overtime. Controversy erupted as there seemed to be communication errors between the referees and Patriots Special Teams Captain Matthew Slater. However, after the game, Coach Belichick did stand by his decision to kickoff. He was hoping his defense would be able to stop the Jets quickly and give the offense better field position than a normal kickoff. Was this unorthodox decision really the correct one to make?
Above are the breakdowns of drives that started with a kickoff for the Jets Offense and the Patriots defense this season, before yesterday’s game. Both sides are better than each other’s average opponents. This data relates to the following probabilities:
Finding an exact probability for certain events is difficult here, as the Jets offense and Patriots defense interact. It does seem reasonable to say that both sides would perform worse than these probabilities imply due to the skill of their opponents. The probability that the team gets less than 20 yards would imply that the Patriots get the ball in a position better than the 20. This is based on a net punt average from Ryan Quigley of 36.3 yards.
However, with the new rules in overtime, unless a touchdown is scored, both teams get their chance on offense. Due to this, the only truly bad outcome for the defense would be the offense scoring a touchdown. If anything else occurs, the offense will get a chance to at least match the other team. In fact, it is probably better to have the ball second so that you know if you need to score a field goal or not, allowing you to go for it on fourth down.
Therefore, the decision on whether to receive or kickoff should be based mainly on the probabilities of touchdowns for the two sides. Similar drive statistics with the Patriots offense and Jets defense give us the following probabilities:
The probability breakdown between offense and defense is pretty similar for both decisions. While the Patriots are the one of the top scoring offenses in the NFL (third), they are still not very likely of driving 80 yards to score a touchdown to start overtime against the eighth best scoring defense in the NFL. This also holds true for the Jets offense (12th best scoring offense) and the Patriots defense (6th best scoring defense).
Ben Zauzmer has already looked into the effect of new overtime rules for HSAC. We can use this idea of Monte Carlo simulations to see if Belichick made the right decision to receive. The Monte Carlo simulation gives us a matrix with the probabilities that the receiving team will have the following point differentials at the end of overtime: -6, -3, -2, 0, 2, 3, 6. I conducted simulations for both decisions, finding the probability of each of these outcomes after 6 total drives (3 drives each). As Zauzmer explains, this is the number of expected drives, based on the length of an average drive, if the overtime went the full 15 minutes. I used probabilities associated with each offense and each defense separately. The offense-probability model will underestimate the likelihood of a tie, and overestimate the importance of having the ball first, while the defense-probability model will do the opposite. The matrices below are the stationary distributions for all the possible outcomes (again, corresponding to the point differentials of -6, -3, -2, 0, 2, 3, 6):
Jets Receive: [0.3259 0.1671 0 0.0262 0 0.1258 0.3549]
Patriots Receive: [0.1891 0.1768 0 0.0247 0 0.2300 0.3795]
Jets Receive: [0.2299 0.2279 0 0.0699 0 0.1789 0.2935]
Patriots Receive: [0.3014 0.1506 0 0.0790 0 0.1828 0.2863]
We can add these numbers up to get the probability that the Patriots win in each scenario. The offense-based model gives the Patriots a 49.3% chance to win if they defer and a 60.95% chance to win if they receive the kickoff. The defense-based model gives the Patriots a 45.78% chance to win if they defer and a 46.91% chance to win if they receive.
Based on these probabilities, it seems that Belichick did in fact make the wrong decision here, regardless of which probabilities are used. While Belichick will always be known as a coach who likes to think outside of the box, he has over thought this one. After two overtime losses this year, it will be interesting to see Belichick’s strategy if the Patriots go into overtime in the playoffs. It is impossible to predict what Belichick will do, but it is clear what he should do in most cases: elect to receive.