In the final minute of the game, the Seahawks had an 88% of winning Super Bowl XLIX. To make grandiose statements about the Patriots passing attack and football analytics based on New England winning the Super Bowl would be silly given the way the game ended.

Okay, whew.  But I do want to talk about the Patriots offense, and more specifically, ANY/A.  As regular readers know, Adjusted Net Yards per Attempt is calculated as follows:

(Gross Passing Yards + 20*PassTDs – 45*INTs -SkYdsLost) / (Pass Attempts + Sacks)

ANY/A correlates very well with winning, and it’s my favorite basic metric of passing play.  But ANY/A, based around yards per attempt, is not perfect.  And I think SB XLIX provides a good example of that.  Tom Brady finished the day with 320 net passing yards, 4 TDs, and 2 INTs on 51 dropbacks, which translates to an ANY/A of 6.08.  Russell Wilson had 234 net passing yards, 2 TDs, and 1 very fateful INT on his 24 dropbacks; that translates to an ANY/A of 9.54.

That means the Seahawks averaged 3.46 more ANY/A than the Patriots. ^[1]For both teams, only the starting quarterbacks attempted a pass. And they lost! Only 6 times in the previous 48 Super Bowls had the team that averaged fewer ANY/A won the game: the ’01, ’03, and ’14 Patriots are on that list, in addition to the ’82 Redskins (John Riggins!), ’97 Broncos (Terrell Davis!), ’05 Steelers, ^[2]Insert obligatory officiating comment. Also of note: Yes, this was Seattle’s other Super Bowl loss. And this game was even worse for Ben Roethlisberger than you might think, since … Continue reading and ’08 Steelers (James Harrison!). Take a look at the graph below, which I have also included in table form at the end of this article. It shows the ANY/A differential (i.e., winning team ANY/A minus losing team ANY/A) on the Y-Axis for each Super Bowl (the X-Axis).

As you can see, no Super Bowl champion ever lost the ANY/A battle so badly. So, how did the Patriots manage to win despite also being significantly outrushed? Well, that’s a very good question. But perhaps this game highlights a couple of the drawbacks of ANY/A.

But there’s something missing from the above analysis: passing first downs. Russell Wilson executed a vertical offense extremely well, but was it any more efficient at picking up first downs than Tom Brady’s horizontal passing attack?

Seattle completed 12 of 21 passes for 10 first downs. Here’s the double-edged sword: Seattle was just the third team all season ^[3]Behind not one, but two, Brian Hoyer led efforts. to pick up a first down on 83.3% or more of its completed passes. The less-impressive part? Gaining a first down on just 47.6% of your attempts is not as good, and even worse to do it on just 41.6% of your dropbacks.

The Patriots, meanwhile, completed 37 of 50 passes for 21 first downs. And while Brady may have only gained a first down on 56.8% of his completed passes, he also converted 42% of his passes into first downs, and most importantly, 41.1% of his dropbacks turned into passing first downs.

So that works to level the playing field quite a bit. On the other hand, it’s not as clear cut as you might think that passing first downs trumps all. Consider that Seattle had 11 drives in the game, while New England had 10. ^[4]I am excluding the end-of-game “drive” by the Patriots that consisted solely of kneel downs. You might think New England had the methodical offense, while Seattle was more boom or bust.

Well, the Seahawks had five drives that went nowhere, picking up 20 or fewer yards, five long drives of 50+ yards, and one 33-yard drive. But New England also had five drives that picked up 20 or fewer yards, and five long drives of 50+ yards. Looking just at drive length, one could never tell which team had the boom/bust passing game.

So what does this mean? I’m not quite sure, so I open things up to you. I have so far ignored the pretty relevant point that the Patriots ran more than twice as many passing plays as Seattle, and presumably that’s part of the analysis of which passing game was more effective. But what do you think? Was the Patriots passing game more efficient/productive than Seattle’s? Does this game highlight some of the flaws in ANY/A? If so, how would you adjust the formula? One thing we can do, of course, is to give 9 yards for each passing first down. But I’ll open this one up to the crowd.

Data Dump
Below is the full list of ANY/A from each Super Bowl participant. Change the number of rows shown using the dropdown box to the left.

With less than a minute remaining in the Super Bowl on Sunday, the Seattle Seahawks lined up in a three-wide-receiver set at the 1-yard line. To the right of quarterback Russell Wilson, Ricardo Lockette stood nearly directly behind another wide receiver, Jermaine Kearse, a yard back and on Kearse’s outside shoulder — an alignment commonly used on pick plays.

New England Patriots cornerback Brandon Browner stood across the line of scrimmage from Kearse, a couple of yards in front of Malcolm Butler, who was tasked with guarding Lockette.

For Seattle, the concept was simple. Kearse would run what is known as a clear-out route: With Browner and Butler aligned so close together (in response to the way the Seahawks’ receivers were set), Kearse’s job was to make Browner backpedal. That would block Butler from cutting across the field to cover Lockette, who was to run a slant to the inside.

You can read the full article here.