You’re on your very first visit to Las Vegas and have never in your life previously played blackjack. You’re sitting at the table having been dealt a seven and a six. Though you’re inexperienced you know enough to know this is not a very good hand. However, the dealer is showing a three, which also isn’t a good hand. What, you wonder, is the best decision, stand or hit?
Luckily for you, others have done the hard work here and figured out the best approach to this hand. In fact, smart people have researched every type of hand under every scenario and know the best decision, statistically, in every situation. They’ve gone so far as to develop a simple grid that makes the decisions easy, allowing even a novice player to make the “proper” play every time. This doesn’t guarantee the player wins such hands, but gives the novice player the highest probability of winning each individual hand.
These blackjack guides are widely available and have been used by players for decades. A good example of one can be found at wizardofodds and looks like this:
If the table isn’t clear it instructs the player to either:
- Hit (red)
- Stand (yellow)
- Double-down (if allowed) or hit (blue)
- Surrender (if allowed) or hit (white)
Even someone who had never played the game before could use this simple guide to play a “proper” game of blackjack. (Sidenote: none of this means the player will win. The rules dictate the house will win about 0.5% more than the players on average even if the player employs “proper play”.)
You’re an NFL head coach. You’ve invested your entire life into playing and coaching football. You’ve reached the pinnacle of your profession, holding one of only 32 head coaching positions. You’re team is on the opponent’s 40-yard line, midway through the second quarter, facing 4th-and-eight. You have three choices:
- attempt a field goal
- go for it
A field goal attempt would be from 57 yards and a miss would give the opponent’s the ball the their own 47. Going for it seems out of the question due to eight yards needed to reach a first down. So you decide to make the sound decision and punt the ball. The ball bounces into the end zone for a touchback and your opponent takes over at their 20-yard line. You’ve gained 20 yards in field position while surrendering the ball.
Congratulations! You’ve joined virtually every NFL head coach in the history of the league in making a statistically irrational decision that reduced your likelihood of winning the game.
And yet no one will really question your decision. The media and talk shows won’t beat you up endlessly for your irrational decision because that’s what NFL coaches have always done; it’s conventional NFL wisdom.
And yet, just like blackjack, there’s a simple table that shows the statistically optimum decision when it comes to fourth-down decision making. There are years of research into this topic and it’s been covered repeatedly by many. The best summary of the topic can be found at Advanced Football Analytics by Brian Burke.
The concept is pretty simple. Every down-distance-yardage situation has an “expected point” number. For instance, a first-and-ten situation from a team’s own 27-yard line has 0.7 “expected points”. Plus, every play has a subsequent result, which affects the “expected points” of the prior play. For example, touchdowns aren’t really worth 7.0 points, they’re worth 6.3 points because the opponent will receive the ball on the following play and, on average, will have 0.7 “expected points”.
The math can get a bit mind-numbing at times but the concepts are actually very simple. Every play has a statistical likelihood that yields the “expected points” number. From the article referenced above, the details on a 27-yard FG attempt:
A successful FG is worth 3 points minus the value of the ensuing kickoff for a total of 2.3 points. A missed FG is worth the EP value of a first down for the opponent at the spot of the kick (or the 20 yd line, whichever is larger).
For example, with the ball on the 20 yard line, the NFL average FG percentage is 82%. The spot of the kick would be the 27, which corresponds to 0.7 EP (that’s -0.7 EP for the FG kicking team). Therefore the EP value of a field goal attempt from the 20 would be:
(0.82 * 2.3) + ((1-0.82) * -0.7) = 2.0 EP
Thus, by understanding the expected points of every field situation possible, a table can be developed that shows the optimum, statistically-supported decision for every situation. In fact, these tables have been available for a number of years and have been widely published. Again, from Brian Burke’s work:
The chart illustrates that between the opponent’s 30 and the 50-yard line, going for it on fourth down is the optimum decision unless there’s more than six yards to go (and even then it’s frequently the right decision).
Here’s a version from the New York Times that compares a statistically-driven decision compared to NLF coaching decisions:
And yet we know NFL coaches almost never make the rational, logical decision, instead opting for less-optimum field goals (or even lower-optimum punts). We also see that teams should be going for it much more often on 4th-and-short (four yards or less) but teams generally limit their fourth-down attempts to two yards or less, and only if they’re at midfield or beyond.
Why are NFL coaches making illogical decisions?
There’s two simple reasons why coaches are blind to rational decisions in these situations:
- group think (conventional wisdom)
I’m just guessing that today’s average NFL coach has spent 40+ years playing and coaching football (a pure guess but if anything it’s probably too conservative). No doubt throughout those decades of football they’ve been taught to make the conservative decision in these situations:
- punt the ball and make the other team go farther to score
- kick the field goal and take the three points
Thus, decades of experience have created an entire population of coaches, players, media and fans that largely accept such decisions without a lot of criticism. Thus, the punt that flies through the end zone and results in a touchback generates little blowback from anyone.
If, however, the coach were to follow the logic of the statistics, go for it in that 4th-and-8 situation and then see his play fail the knives are sharpened, the criticisms pointed and widespread.
Which brings us to the second reason why coaches don’t follow the numbers: fear. It’s simply much easier for NFL coaches to follow the “conventional wisdom” and do what’s always been done than do something “radical” and get crucified should it fail.
When “radical” becomes standard practice
Professional baseball has already been through this type of situation. Advanced analytics have revolutionized the way the game has been played and what was once “radical” is now standard practice. This is particularly evident the use of the stolen base and sacrifice bunting. Both tactics have seen a dramatic decline over the last 15 years as a result of analytics showing they more often decreased the likelihood of scoring than increased the likelihood of scoring.
The following comes from Anthony Castrovince of Sports on Earth:
A simple Google search on the decline of the sacrifice bunt shows the following results:
The bottom line: statistical analytics have revealed that “conventional” baseball wisdom for 100+ years has been wrong and now teams are adapting and avoiding “bad dugout decisions”.
All this, of course, is well known. Every sport is being transformed by advanced analytics which can reinforce or challenge any long-held strategy. So it’s somewhat surprising that it took the Philadelphia Eagles’ Doug Pederson’s 2018 approach to fourth downs to really highlight how rigid NFL coaches have been.
Pederson’s “aggressive” fourth-down decision-making elicited widespread praise when he famously went for the touchdown when facing a 4th-and-short situation late in the first half of Super Bowl LII. The fact it was successful and employed a daring reverse-tight-end-to-quarterback pass just made it all the more impressive.
But this was Pederson simply doing what he’d been taught by the numbers. All season long he’d been aggressive on fourth down. Earlier in the year when facing a situation very similar to our hypothetical outlined above he went for it on fourth-and-eight. On the play quarterback Carson Wentz was sacked, the Eagles turned the ball over.
What’s interesting is that decision didn’t generate a tidal wave of “what was he thinking” articles. Instead, most articles noted that while unconventional, the decision was supported by analytics:
But note that within the sympathetic BGN article there’s a poll that showed 69% of Eagles fans disagreed with Pederson’s decision. I’m just guessing, but I’d bet that if Pederson makes the exact same decision next year he’ll have near unanimous support from Eagles supporters.
Which brings me to my final point. Like in every other sport, it’s inevitable that Pederson’s “radical” approach to fourth-down decision-making will become widespread and eventually become the “conventional wisdom” within the NFL. My guess is it will take about five years before pretty much every team is going for it in that situation.
Yes, there are those who will claim NFL football isn’t blackjack and there are other considerations. And they’re 100% correct. There are a number of variables that come into play in NFL football that are absent from blackjack:
- time remaining
- quality of field goal kicker
- quality of punter
- quality of offense and defense (both your own and opponent’s)
These variables can move those numbers a bit here and there. But they don’t change the fact that NFL teams have, for decades, been irrationally conservative when it comes to fourth-down decision-making.
Here’s the really interesting thing to me. Were you to take a group of gamblers, mathematicians and statisticians who knew nothing about football, showed them the fourth down table above and asked them to make fourth-down decisions for an NFL football team they’d make make more sound decisions than today’s NFL coaches.
In fact, many blogs, media members and fans are farther down the road to acceptance of such ideas than NFL coaches.
This befuddles me because NFL coaches are smart people. I guarantee the vast majority (if not every) NFL coaching staff has had some stats guy give a presentation on exactly the topic we’re covering right here today. And the collective response of those NFL coaches has been “yeah but...” and they came up with some excuse why they should continue making statistically irrational decisions.
All but one. Doug Pederson said “tell me more”, decided to incorporate that strategy into his game-day decisions and he’s a Super Bowl winner.
You tell me who’s the “logical” coach.
Will Doug Pederson’s "radical" fourth-down decision-making become "conventional wisdom" within the NFL?
This poll is closed
Duh! Who would argue with the numbers AND the results?
Yeah - but it will take a while. A lot of stubborn coaches out there.
Maybe - but these stats-driven tables don’t take enough variables into consideration; there’s more to football than numbers.
Nah - old guys gonna old guy.