When the Raiders scored a touchdown with 47 seconds left in the fourth quarter of their Week 1 matchup in New Orleans, everyone just assumed that Jack Del Rio would send Sebastian Janikowski out for a game-tying PAT. Everyone was wrong. Del Rio kept the offense on the field, deciding that a two-point conversion was their best shot at pulling out the victory on the road. Derek Carr lobbed a fade to Michael Crabtree in the corner of the end zone, and the rest is history, as the Raiders pulled out a 35-34 victory.
But did Jack Del Rio make the right call? It's easy to say yes after the fact. After all, the Raiders won, so what's the issue? The beauty of Del Rio's decision is that he indeed made the right call, and he was rewarded for it (there's nothing that makes an analytics guy happier than when a coach makes a decision with positive expected value and it works out in his favor!).
In order to understand Del Rio's decision, one must understand expected value. Since the NFL moved the PAT out of “gimme” range and back to the 15-yard line last season, extra points are no longer a virtual guarantee. In 2015, NFL kickers converted on 1146/1217 (94.2%) of their PAT tries. It should be noted that Sebastian Janikowski is an above average kicker, although his main strength is kicking power, which is not vital in a PAT scenario. With that being said, Janikowski missed just one PAT in 2015, connecting on 38 of 39 tries, for a solid 97.4% converstion rate, and although it's an extremely small sample size, we'll use that number for our evaluation.
In order for a two-point conversion attempt to be worth the same amount as a PAT, one must expect to convert at a rate of the PAT divided by two, since the two-point conversion is worth double the points of a PAT. Therefore, if the two-point conversion is achieved at a rate of more than 48.7% (Janikowski's conversion rate divided by two), there is positive expected value on the two-point conversion.
Here are the league-wide two-point conversion rates over the past three years:
2015: 45/94 (47.9%)
2014: 28/59 (47.5%)
2013: 33/69 (47.8%)
What you're likely to immediately notice is that that the two-point conversion rates for the last three years are all lower than the target rate of 48.7%. Case closed, right? Wrong.
In this instance, we had a league-average offense against the league's worst defense
. Oakland ranked 18th
in passing efficiency and 14th
in rushing efficiency in our final league-wide rankings prior to the season
, while the Saints ranked 31st
in defensive pass efficiency and 30th
in defensive run efficiency, making them the worst defense in the league. Those rankings also don't take into account the injury to New Orleans' top cornerback, Delvin Breaux, that was suffered during the game yesterday. Without Breaux on the field, the Saints defense is even worse, which at this point, is almost unimaginable.
According to our simulations, the Raiders were 49.7% likely to convert a two-point conversion against a league-average defense, while the Saints were 58.2% likely to allow a two-point conversion against a league-average offense. When we run the Raiders offense against the Saints defense, we get a 55.3% conversion rate for the Raiders. Del Rio did make a mistake. His mistake was not going for two points after every touchdown.
If the Raiders had kicked the PAT to tie the game, they would have had a 43.3% win expectancy at that point in the game. So while some people may have considered kicking the extra point "the safe thing to do", it actually would have still put the Raiders in a situation where they were expected to lose more often than not. This wasn't a decision that was made with no time left on the clock; 47 seconds is an eternity for a quarterback of Drew Brees' caliber, especially with a kicker that has the leg of Wil Lutz (yes, he missed the 61-yard FG to tie the game but it had plenty of leg). All in all, with all factors considered, and regardless of the outcome of the game, we strongly believe that Jack Del Rio put his team in the best position to win on Sunday afternoon.