Who has been the best quarterback in the NFL over the last five seasons? It’s a question that is sure to spark debate, with some going with Tom Brady, others with Drew Brees, still more with Peyton Manning, and Aaron Rodgers making a push to be in the conversation. If rings are your thing, then Eli Manning has the most in that span.

The answer, though, is pretty clear cut, and it’s none of them. Or rather, it’s all of them, and all the other guys who have thrown passes in that span, from Derek Anderson to Vince Young (unfortunately, Eric Zeier and Jim Zorn no longer play in the league). The best quarterback in the NFL has been the one that gets to throw on first down when backed up near their own end zone, by a pretty large margin.

Don’t believe me?

Over the last five season, quarterbacks have run 627 pass plays on first down and 10 inside their own 10 yard line. The results: 8.7 yards per attempt, a minuscule 1.9% sack rate, a turnover 4% of the time, and immediate first downs on 33% of throws.

For comparison, here are Drew Brees’ averages over the last five years: 642.8 passing plays, 7.7 yards per attempt, an excellent 3.0% sack rate, and a fumble or interception 3.9% of the time.

The list of individual passers who have averaged more yards per attempt than are collection of quarterbacks back against the end zone, over the last five years (min 500 passes)? Aaron Rodgers in 2011 and Philip Rivers in 2010.

So yes, they are very far away from the end zone, so the touchdown rate is low (7 touchdown passes). The likelihood of throwing a touchdown pass is more than 3 times greater than a quarterback throwing one on first and 10 at the 20 yard line, even though it is further back.

All this data comes from pro-football-reference’s new play index finder, where you can find the results of groups of plays by team, field position, down, distance, type of play. The best way to compare the tradeoff of running and passing on first down is by looking at how teams behave when they start after a touchback.

Starting at their own 20, teams call a pass on first down 51% of the time–almost exactly even. The pass plays average 7.4 yards per play, while resulting in turnovers 4.6% of the time, and sacks 6.5% of the time.

The run plays average a healthy 4.6 yards per carry, with a turnover rate one-third of pass plays (1.5%).

A fair amount has been written about game theory, and run/pass equilibrium, so going off on that tangent would take too long, but feel free to check out Advanced NFL Stats and run a search of other articles. The very simplified gist is that teams should increase their passing (or running) percentages until the find equilibrium in the payoffs.

It’s fairly complicated in football because you balance bigger gains and greater chances of picking up another first down (passing) against safer, smaller gains (rushing). Let’s assume, in arguendo, that the teams are at equilibrium when they play call at the 20 on first down. They call roughly equal passes to runs, and they get a payout of of about three yards more per play, and a new first down rate on that play of a little less than three times as high (31% vs. 12% on runs), in exchange for more negative plays, and a turnover rate about three times as high.

Well, if we assume those payouts to be equal, let’s compare that to the pass/run balance on first down inside one’s own ten yard line, along with the payouts (yards per play, first down percentage on that play call, and turnover rate). For turnover rate in this range, I include safeties as a turnover, since you relinquish possession a little further away, but also give up points.

Here’s the breakdown:

Just so you are reading that right, teams call a pass 27% of the time on first down inside the 3, and gain 10.19 yards on average when they do.

It seems pretty clear to me that we have a case where one side of the ball is willing to gamble that the other side is going to call the play a certain way, and the other side adhers to dogma and plays right in to their hands, largely. If defenses were playing straight up, like they do at the 20, there is no way you would see those numbers. Defenses are selling out to stop the run on first down, and succeeding, and the equilibrium of passes to runs called by the offenses is out of whack as a result.

How do we know, other than personal observation about teams handing it to the fullback and immediately getting stuffed?

• turnover rates are virtually equal once we account for safeties right at the goal line. Teams are selling out and hitting backs early, causing fumbles or catching them behind the line for safeties.
• sack rates are abominably low. There were actually no safeties recorded on passes inside the 3 on first down over the last five years, and only one sack (fumbled, for a touchdown). There’s no way the sack rates on first down should three to four times less frequent in a tighter space unless defenses are simply not playing the pass, and instructing their edge players to hold and play run first.
• the yards per play numbers are way out of whack compared to our “starting at the 20″ numbers. 10.2 yards per play versus 3.1? Are you kidding me? Five times more likely to pick up a first down and actually get out of the shadow of the goal posts? Passing payoff is way better, considering there is no downside at this point given the relative turnover rates.

I don’t know if those numbers at the 20 are equal when we compare running and passing. I do know that the numbers inside the 10 are way out of whack, and show that teams run the ball way too much given the defenses they are getting. Coaches think too negatively, worrying about creating space for the punter. If you think negatively, that’s what you get. Negative results.

Instead they should view it as an opportunity. Passing near the goal line turns the league wide collection of quarterbacks into Hall of Famers when going against defenses selling out to stop the run. If teams start passing it more, maybe the defenses adjust, and then coaches can actually create that space they so crave.

[photo via US Presswire]