Which of the NFL's 'Dumbest' Strategies Are Smarter Than You Think?

NFL coaches are stupid.

OK, no, that's not even remotely true, no matter what midday sports-talk radio and TV want to tell you.

But NFL coaches are old-fashioned, ultra-conservative and a little stubborn. They make many of their decisions as if one false move will get them ridiculed for weeks, then fired.

Can you blame them? Each NFL Sunday, we see tactics and play calls that make the fans on the upper concourse howl and cause the message boards to hyperventilate. Coaches call plays that "everybody" knows have little chance of success.

But does "everybody" really know which plays work more often than others?

With the help of the Football Outsiders in-house database of tens of thousands of sortable NFL plays, I worked out the success rates for several of the most criticized and reviled NFL tactics.

Some strategies probably should be sent to the tactical retirement home with the flying wedge and the dropkick. Others are much more effective than you might think.

So before you criticize the head coach for blowing that big decision, make sure you know the numbers.

Prevent Defense

"Prevent defense" is a broad, vague, overused term and an almost undefinable concept.

Some fans blame the "prevent defense" anytime the home team surrenders a fourth-quarter lead, even if eight defenders blitzed with battle axes on half the plays of the game-winning drive.

Coaches tend to use phrases like "two-minute defense" or "four-minute defense" to describe late-game, hold-the-lead strategies. If a coach says "prevent," he's probably talking about an anti-Hail Mary strategy or one of those plays where six defenders line up 19 yards downfield on 3rd-and-20.

So it's hard to statistically analyze how well these "prevent" strategies work. But we can determine how many defenders rushed the passer in protect-the-lead situations and calculate whether a three- or four-man rush is more effective at "preventing" anything than a five- or six-man rush. Not every four-man rush is a conservative tactic, as anyone who has ever watched the Cardinals or Chiefs can attest, but the numbers give us something objective to work with.

The table below tabulates all 2015 fourth-quarter passing plays when the defense was nursing a one- to 16-point lead; in other words, the game was still close enough to worry about an offensive touchdown. The table shows the first-down rate (touchdowns included), sack rate, interception rate and big-play (20-plus yards) rate allowed. The last thing a defense wants to do when playing with a lead is give up 20-yard chunks of real estate.

The optimal late-game strategy appears to be to rush four defenders. Predictably, blitzing increases the sack and interception rates but also results in more first downs, big plays and yards per play; late-game blitzes when leading are only common when the offense is in scoring position for that very reason. Rushing only three defenders appears to be so conservative that it's counterproductive, though three-man rushes are much more common late in two-touchdown games than when the defense is clinging to a three-point lead.

Maybe the defensive coordinator really knows what he is doing when he rushes four defenders and drops everyone else into zone coverage in the fourth quarter, perhaps sneaking in a well-timed blitz now and then. Teams blow leads in the fourth quarter because passing offenses are better, faster and more efficient than they were even 10 years ago and rules are designed to favor the offense.

So don't blame the "prevent defense." It's just NFL football.

3rd-and-Long Surrender Plays

The first-down conversion rate for third down and 10 or more yards to go last year was 20.5 percent. That's tough, but hardly hopeless, which makes you wonder why so many offenses seem so eager to run draw plays and short passes that have little chance of success on 3rd-and-long.

The 3rd-and-long handoff is an easy way to get the home crowd booing. There were 160 handoffs on 3rd-and-long (10 or more yards to go) last season, and only 12 of them (7.5 percent) resulted in first downs or touchdowns.

There are extenuating circumstances to the third-down handoff, though. Of those plays, 35 percent occurred with the offense inside its own 20. Those plays netted a respectable 6.7 yards per rush. When it's 3rd-and-24 from your own 6-yard line, seven extra yards for your punter may be the best you can hope for. Third-and-long handoffs are also much more common late in lopsided games, when the winning team is just killing clock and the losing team is trying to maintain dignity.

But what about the Blaine Gabbert special: a micro-short pass on 3rd-and-long? There were 840 pass attempts with an in-the-air distance of eight yards or less on 3rd-and-long last season, passes short enough to be beyond the reach of a receiver turning and diving for a first down. Those passes represented 41.2 percent of all 3rd-and-long plays; it's remarkable just how often teams that need 10 to 15 yards execute plays that are specifically designed not to gain 10 to 15 yards!

The conversion rate on 3rd-and-long short passes is a meager 11.7 percent—not much better than the handoffs. There are many types of plays among those short passes, from screens to running backs to shallow crossing routes to Jarvis Landry to dump-offs after the quarterback at least tried to chuck one deep downfield. While the dump-offs and checkdowns are often understandable, offensive coordinators should perform some quality control on their game plans to determine why, for example, Teddy Bridgewater threw 10 passes on 3rd-and-long to Matt Asiata last year.

Remove the handoffs and short passes, and the 3rd-and-long conversion rate rises to 29.6 percent. Keep in mind that all sacks are included in this final category, because we cannot be sure how long a pass might have been if the sack did not happen. So the real rate may be a little higher. Risk-averse strategies make 3rd-and-long look harder than it really is.

Oh, about the players I just singled out: Gabbert finished second to Nick Foles in failed completion rate last year, but Gabbert was more likely to throw a five-yarder on 3rd-and-24 than Foles (who threw more zero-yarders overall). Landry finished first and second in the NFL in failed completions in 2015 and 2014, respectively, mostly because of problems that have nothing to do with him. Asiata caught eight of 10 3rd-and-long passes last year, including an eight-yarder on 3rd-and-27. If only the Vikings had a more dangerous threat in their backfield.

Punting from Beyond Midfield

Teams punted when the line of scrimmage was at or inside the 50-yard line 448 times last year. That's nearly once per game per team! Most of those punts—365 of them—came when the game was within 10 points, so there aren't a lot of meaningless, late-game punts in the data. There are lots of 4th-and-long situations from around the 49-yard line in the data, but 113 punts occurred with five yards or less needed for a first down and the game still close.

The fourth-down conversion rate in the NFL last year was just below 50 percent. The "discretionary" rate is even higher if you take out 4th-and-15 situations at the end of games and such and concentrate on situations where the coach has a real choice. The rate jumps to 58 percent if you remove fourth-quarter attempts and attempts with more than five yards to go. A 4th-and-short attempt from the opponent's 45-yard line will probably succeed two times out of three. Coaches still choose to punt far too often.

The average punt from inside midfield nets about 34 yards, with the opponent getting the ball at the 13-yard line. Pins at or inside the 10-yard line occur roughly 40 percent of the time. The percentage chance of retaining possession with excellent field position is roughly equal to the chance of pinning the opponent at the goal line. Unless I have a 21-point lead, I'll keep the ball, thank you very much.

The Grand Poobah of dumb punts last year was Ken Whisenhunt. His Titans punted on 4th-and-2 twice in the first half against the Bills in Week 5: from the 39- and 36-yard lines. The Titans also settled for a 21-yard field goal. The Bills won, 14-13, and Whisenhunt lost his job a few weeks later. 'Nuff said.

The Wildcat

The Wildcat fad may be as out of date as your MySpace page, but every once in a while, a non-quarterback takes a handoff and plunges off-tackle for a yard or two. There were 54 Wildcat plays last season, as defined by an offensive snap to a non-quarterback. The Wildcat netted 5.2 yards per play, three plays of 20-plus yards and 12 stuffs for no gain or a loss (plus two incomplete passes).

The Wildcat results are even worse than they look. Two runs by one player resulted in 80 yards, meaning that everyone else averaged just 3.9 yards per play. The table below shows who the Wildcat superstar was and lists the results for other direct-snap recipients.

The Rams often cut out the quarterback middleman late in the 2015 season and let Todd Gurley take direct snaps and either keep them or hand off to Tavon Austin. As the table suggests, other teams used the Wildcat the same way the Rams did: as a desperation tactic when their quarterback situation was in shambles. Le'Veon Bell took his direct snaps when Michael Vick and Landry Jones were the Steelers quarterbacks. Cecil Shorts III and others took direct snaps for the Texans when T.J. Yates was the quarterback or Brian Hoyer was playing through injuries.

So if you are down to your third or fourth quarterback and just need something to diversify your scaled-down game plan, the Wildcat won't kill you. Otherwise, there are better ways to average 5.2 yards per play.

Super-Short Field Goals

You probably know what the analytics say about 4th-and-goal from the 1- or 2-yard line: GO FOR IT! Even if you fail to score a touchdown, your opponent is extremely likely to give you the ball back on a short field, making it easy to kick that field goal you are thinking of settling for anyway.

But it's hard to put faith in that spreadsheet when three easy points are staring you in the face. Teams kicked field goals at or inside the 5-yard line 86 times last season. Meanwhile, there were only 38 fourth-down conversion attempts inside the 5-yard line. The conversion rate on 4th-and-goal was 39.4 percent (15 conversions), whereas the field-goal rate inside the 5-yard line was 100 percent. It's hard to blame coaches for taking the sure thing.

The problem comes when coaches get too risk-averse. It's one thing to settle for a short field goal with a 14-point lead and another to do so when trailing. Fourteen times last year, coaches settled for field goals at or inside the 5-yard line when trailing by more than three points; in other words, the field goal would not give them the lead. Those coaches ended up winning three of those games and losing 11. Settling for the safe points is too often a recipe for settling for the close loss.

Drawing the Defense Offside on Fourth Down

What a dumb idea! Those are professional defenders across the line of scrimmage! They know what you are trying to do! You must think they are stupid to fall for this old trick! What a waste of…

Oh wow, it actually works.

Defenses jumped offside (or encroached, committed a neutral-zone infraction, etc.) five times on "draw 'em off" plays last year. Defenses also jumped offside on three actual attempts to go for it on fourth down—runs or passes that were executed, as opposed to the quarterback standing there saying "Hut…HUT…huthuthut." It's impossible to tell just how many draw-'em-off attempts there were without combing through 35,000 plays from last year, but five free first downs are worth the delay-of-game penalties and timeouts that offenses had the luxury of burning anyway.

The draw-'em-off play has gotten more sophisticated in the last few years. The Chargers added a hurry-up wrinkle to it to draw two defensive flags: Philip Rivers raced to the line as if to run a sneak, the defense scurried into position and one "HUT" got some jumpy lineman to move. The Bills sent quarterbacks Tyrod Taylor and EJ Manuel on the field at the same time against the Jets and looked ready to run some nutty Wildcat play. Manuel barked a signal and a befuddled defender jumped across the line of scrimmage.

The more common fourth-down conversion attempts become, the more successful draw-'em-off attempts should become, because they are less obviously draw-'em-off attempts. In fact, some of last year's five successes may have become real fourth-down conversion attempts if the flags had not started flying. If coaches are going to put a lot of engineering in the form of motion, strange snap counts, no-huddle tactics and weird personnel into any play, it might as well be a play that means the difference between retaining or losing possession.

Mike Tanier covers the NFL for Bleacher Report.