4:25 left, 2nd quarter: You’re down 7-6 at home, trailing a better team. But you are playing well. Your fans are into the game. Your NFL team gets stopped after completing an 8-yard pass on 3rd and 9— just short of a first down. You have the ball at the 50-yard line. They bring out the sticks, and sure enough– the football is 4 chain links short What do you want your team to do?
Sure enough, most NFL coaches opt to punt here. The announcer makes some ignorant comment, usually something along the lines of: “Well Jim, the fans here at _____ (insert stadium) don’t seem to like the playcall, but Coach isn’t going to let fan opinion influence his decision making.” Which is ironic—because just the opposite is true (more on that in a bit). What’s fascinating is that week after week, coaches make sub-optimal decisions for their teams, and no one speaks out. Using basic outcome EV decision-logic, we can see just how wise punting is on 4th down and less than a yard. The inability of coaches to see how –EV punting on 4th&short has created an arbitrage opportunity that every coach is irrational not to seize.
In football, we can measure our field position using a system called “Expected Points”. For instance, suppose we have the ball at midfield and it was 1st&10. This is worth 2.0 expected points. If we threw a five yard pass to our tight end, creating a 2nd&5, this is now worth 2.13 expected points. Had we lost 5 yards on a sack, bringing up 2nd&15 from our 45, this is only worth an expected value of 1.2 points.
Now, let’s apply expected value to our situation above. If you have the ball at the 50 yard line, you have 2.0 expected points. However, as you travel down the field, your EPs go up. Using stats from the past five years, teams that have attempted 4th and less than one yard tries have been successful 72.8% of the time. But let’s be fair and say your team has a lousy offense. Say you are playing the Ravens, whose formidable defensive line has given up 3.5 yards a rush this year. Let’s say you have a 66% chance of converting the first down (and that is being generous). That means 2/3 times you will be successful, for every one failed attempt, on average. For the two successes, your team will be granted a first down at midfield. For the one failed attempt, the other team will get the ball at midfield (approximately).
Or you can offer your opponent the ball ALL THREE times at their own 15 yard line. Is 35 yards worth giving up possession 100% of the time? No.
[(2.0 EPs)*2/3 + (-2.0 EPs)*1/3 ] vs. -0.8 EPs each time.
There is no comparison. (Note: When I say 0.8 EPs, it is because statistically having the ball on your 20-yard line is worth 0.8 points to that team) So why do coaches continuously make dumb decisions, when study after study by decision-experts show that going for it on 4th down is a +EV decision? Some claim that our aversion to loss causes coaches to punt more. Failing to convert a fourth down is painful. The brain is constructed to avoid losses more than an equivalent gain. Instead, coaches view a punt as a compromise option, and see it as a neutral outcome. Better that than gambling over what amounts to be a net of less than 2/5 of the field most of the time.
Just like firms try maximize profits, NFL teams should attempt to make decisions that maximize their win percentage. And yet coaches have not making said decisions. In fact, as shown in a study by economist David Romer, punting on 4th and less than 3 anywhere past your own 40-yard line costs your team a 3% chance to win the game. So why do coaches do it? It makes no sense. Except for our perceptions. When a coach makes a gutsy call and it backfires, the coach reaps the majority of the blame. If he opts to punt and the other team drives down, it is the players fault, and they just have to try harder. What people miss is that far too often coaches worry about media reaction and their own jobs. They don’t maximize their EPs because they play scared. A lot of people give Rex Ryan slack, but I saw him make two good decisions early in the Houston-New York MNF matchup. Going for it twice early in the game on 4th and short was smart—and I give Ryan a lot of credit (not something you’ll hear me utter too often) for actually making the BEST DECISION for his team. Kudos to coaches who utilize optimal decision making throughout an NFL game.