Why the box score doesn’t tell the full Blake Bortles story

Box scores don't always accurately reflect a player's performance—Blake Bortles' numbers against the Colts is a perfect example.

| 2 years ago
(AP Photo/Phelan M. Ebenhack)

(AP Photo/Phelan M. Ebenhack)

Why the box score doesn’t tell the full Blake Bortles story

Pro Football Focus’ grading often looks looks completely different to what you would expect from watching highlights or reading the box score—but there is good reason for that.

For instance, Blake Bortles’ performance on Sunday against the Colts looked good on the stat sheet—250 yards, three touchdowns, and no interceptions—but not so good in the PFF grading system. Why?

PFF grades every player on every play, and does so independently of the results of the play. QBs are graded on the plays they make, not on the plays of others after they have released the ball. If a passer dumps the ball off to his checkdown, we see that as an expected, routine play—whether that checkdown gains a couple of yards, loses a few, or breaks five tackles and takes it to the house. The quarterback’s grade won’t change based on the actions of the checkdown player and the defenders, but his passer rating certainly would.

It works in reverse too; if a passer throws the ball right to a defender, he will receive the same downgrade if the defender drops the ball as he would if the defender had caught it and recorded an interception. The throw was the same. The quarterback shouldn’t be rewarded for getting lucky and seeing the defender fail to punish him with the result of the play.

This is all by way of saying that, just because the Jaguars whooped the Colts in Week 14, it does not mean that Bortles actually played very well, overall, in the game.


PFF grades go through an extensive check and review process. They go live first thing on a Monday after the first run, but then each game passes through at least two more sets of eyes using the All-22 tape before being reviewed by a team of senior analysts. Plays then get referred to the Pro Coaches Network (a team of current and former NFL coaches with a combined 400+ years of NFL coaching experience) to finalize the whole process for the week. And that’s just the grading. We have other processes for various additional data we track.

By the time every game is finished for the week, it has had something in the region of 50 man-hours spent on it collecting and checking all of the data, so that when the grades are finalized, they aren’t simply gut feelings or the bizarre output of a series of calculations, but instead the result of an extensive and honed process that works differently to most casual analysis.

In the course of the checks and reviews using the A-22 tape, Bortles’ grade did in fact move up, and settled at -2.9 for the game. That, of course, still doesn’t match the box score or score of the game, but that’s the figure we’re working from here. It’s a below average grade, getting into “poor” territory, but that doesn’t mean that grade encapsulates what we think of Blake Bortles overall, or even this season.


Bortles, even including that game, is the 13th highest graded quarterback of 2015 on a play-by-play basis. Using our new PFF rating system that weights passing, rather than running, a little higher, and adds in some additional data, he slips a little—but is still 21st in the NFL, in the top two-thirds of passers in the league. He has undoubtedly taken a huge step forward from a year ago, when he was the lowest-graded QB in the NFL.

There is a lot to like about him—and not just his twitter game—and he is capable of making huge plays. He usually makes a couple of great ones each game. There is also a lot of negative to his play as well, and not all of it shows up in the box score.


Here’s a great example early in the game. With the score tied at 3-3, Bortles threw a pass that could—and probably should have been—picked off, setting the Colts up inside the Jaguars’ red zone, and likely leading directly to points.

The Indy defense was lined up with off-coverage against Rashad Greene in the slot, and Bortles tries to pick up some free yardage with a quick slant. The problem is, he doesn’t ever see LB D’Qwell Jackson reading his eyes and heading right to the ball.


This is a bad decision and a bad read, and only a pretty well thrown pass (from a trajectory standpoint) saves this from being a walk-in touchdown going the other way. In the box score, it goes down simply as an incomplete pass: 0 for 1.

Didn’t happen.

Only, it did happen, whether Jackson came up with the ball or not. We grade this decision and throw as an extremely bad one, and don’t act like it didn’t happen just because Jackson couldn’t complete the play. This isn’t an “if” or “but,” it’s a bad throw, whether the defender punished it or not.

That was only 8 minutes into the game, and it wasn’t even the first pass Bortles threw that could (and maybe should have been) intercepted. The first pass he attempted ended up as a 17-yard completion, but required a Colts defender conspiring to get it from his hands to the Jaguars receiver’s for it to do so.


Scrambling out of the pocket to his right, Bortles is trying to find somewhere to go with the ball, and after thinking better of trying to force the ball into a tight window to Allen Robinson, and then trying to direct Robinson to open space, he instead turns to the other side of that space in front of him to see Clay Harbor coming into it. Bortles fires without thinking, but puts the ball closer to the Colts’ defenders closing from the other side with the flow of the play than he does to Harbor. D’Qwell Jackson again gets both hands on the ball, but in going to the ground, it pops up and lands right in Harbor’s hands.


The box score says 1 for 1 for 17 yards, but this is another bad decision and ball location where Bortles got lucky—even luckier than the other pass, because this one wasn’t just broken up, but actually bounced its way to the intended receiver.

Bortles gets a lot of credit from fans for “giving his receivers a chance to make a play” or playing to the strengths of his receivers. This is where things get murky, because Ryan Fitzpatrick gives his receivers a chance to make a play like no other quarterback in football, and it often works, but it means there are a lot of dangerous passes in the air that other quarterbacks never attempt. This season, he has the fifth-highest number of “turnover worthy plays,” with 25, but only 11 of them have been caught. Sometimes he goes on a lucky run and none of them are picked off, and sometimes he looks like he can’t stop throwing picks.

The other extreme of that scale is Alex Smith, with a near pathological aversion to taking risks. This isn’t an either/or situation. The best quarterbacks in the league reside between those two extremes.

Every pass is different, but in general terms, ill-advised passes don’t become good decisions just because you’re aiming the ball at Allen Robinson.


Take this pass from a week ago. When Bortles puts this ball in the air, it’s an objectively bad decision, and any coach in the NFL would tell you that. He’s throwing it deep right at a safety who has position over the top of the play with his eyes on the ball. This is what the defense wants, and is an offense’s nightmare. As it turned out, the receiver was able to go up and get the ball, taking it away from the safety who had the position to do it. From a results standpoint, this was a fantastic outcome, but the decision was a poor one, because over time, the balance of that play would not be a good one for the offense. Giving your receiver a chance to make a play still only makes sense with the right defensive look. A waiting safety over the top of the play is not a good look. Having good receivers mitigates those throws and increases the chances they will work out, but doesn’t make them all of a sudden good decisions.

Here’s another example this week.


This was still a close game by this point, and in fact, the Colts had a 6-3 lead. Jacksonville are facing 3rd-and-4, and Bortles decided to go for it all. That’s not necessarily a problem, but the decision to throw deep against the coverage he was seeing was an issue. The Colts showed and played two high safeties all through this play. At no point was a post from the receiver isolated to the top of the screen ever on. Giving your receiver a chance to make a play still needs to come on advantageous looks. If the Colts were showing single-high on this play with the FS to the middle of the field, and the wide out isolated like this, then by all means, put it up and trust your guy wins against their guy. Here, however, it’s not one-on-one. It’s a receiver on a corner running right into a waiting safety with eyes only for that receiver. Oh, and the receiver was Marquise Lee, not Hurns or Robinson. The fact they only needed 4 yards on the play on third down only compounds the poor decision.

As it happens, the Colts defended the underneath stuff pretty well, but there was a route to the bottom of the screen open early before the curl-flat defender could get out to it, had Bortles been looking that way instead of for the home run.

Obviously, this is not an exhaustive breakdown of his game this week, or over the season. Our play-by-play grading is the culmination of that, and over the season, it doesn’t hate Bortles. In this game, it certainly didn’t love him. This is rather an attempt to show that, just because the team scored 51 points (two touchdowns of which were returns), and his passer rating (which is really just a metric rating the production of the passing game as a whole) was 114.6, doesn’t mean Bortles had a great game, and the above are some of the reasons why that is true.

If D’Qwell Jackson had a little fresher tack on his gloves, or just slightly better hands, Bortles’ stat line could have read 15-of-30 for 233 yards three touchdowns, and two interceptions, which is a passer rating of 81.6 for the game. One of those picks would almost certainly have cost the team points. That hypothetical scenario doesn’t require anything different from Bortles to happen. Same throws, different outcome, based on something entirely outside of his control.

We graded the throws, not the outcome, because Bortles had no influence on the latter once the pass left his hand.

Bortles also fumbled the ball, turning it over late in the first half when he tried to escape the blitz. The Colts had a 13-3 lead at this point and began a drive at the Jacksonville 34-yard line with 1:11 left in the half, and a chance to deliver a hammer blow off the back of the turnover. Instead, Hasselbeck returned the favor, and the Jags scooped and scored, starting the comeback. Again, though, the result meant that Bortles got lucky, and wasn’t punished for a bad play that doesn’t show up in his passing stats (but is downgraded in our system). That play might as well be an interception—it had the same outcome—but because it doesn’t get recorded in his passing numbers, and the Jags’ defense actually turned it into a score of their own on the next play, we pretend it never happened at all.

That’s why he ended this game with a negative grade, despite some strong work and fine plays in the second half.

PFF doesn’t hate Blake Bortles, nor do we think he is a poor player, but we do think that, in this game, his stat line and the point-total the Jags amassed significantly flatter his performance.

| Senior Analyst

Sam is a Senior Analyst at Pro Football Focus, as well as a contributor to ESPN and NBCSports.

  • Marcus Johnson

    The final grade is on Wednesday that why I never trip out about the initial grade and I don’t see why pff just doesn’t wait until Wednesday to release them it would calm peoples quips with the website I believe.

  • Scott Kohler

    Why don’t you guys grade Brady accordingly when he runs a dink and dunk offense then?

    • PFFSamMonson

      We have done in the past, and taken equal heat for that. The fact he is grading so well this season should tell people he’s doing more than just that. His deep ball has been unusually accurate (for him) this season and he is genuinely playing some exceptional football.

      • crosseyedlemon

        Some players are also going to find themselves in systems that inhibit their potential grades somewhat but coaches have to consider what’s best overall for the team to succeed rather than worrying about soothing individual egos. Every QB would love to have carte blanc to bomb opposing defenses at will but that may not always be the best approach for achieving success.

    • Kyle Ferguson

      I would imagine a lot of small green grades and 0.0s, then a few reds in there. Palmer probably has big green grades but some big red grades. The fact that he’s so efficient in that deep passing offense is crazy. Why can’t Stafford do that?

  • crosseyedlemon

    The explanation of the grading procedures was quite interesting but still left me wondering about a few things like:
    a) is the grading process a blind one where each grader does his evaluations unaware of what other grades have been assigned.
    b) if there is a defined set of metrics used unilaterally how would the weight placed on each of those metrics be determined
    c) is the deviation from the consensus grade noted for each grader to determine if any personal biases are in play.

    • Forrest

      The How We Grade link at the top of the page answers some of those questions.

      • crosseyedlemon

        Thanks for the info. While most readers are concerned with the actual number awarded players I find the evaluation process itself more interesting as it certainly must provide some unique challenges.

        • Kyle Ferguson

          Very true. The biggest thing I wanna know is what grade a certain play gets. That part of the grading is not as concrete and can be up to interpretation. Though overall it shouldn’t be too off. Miss tackle = bad, sack = good

  • Jaguars28

    So making agreesive throws is bad? What?

    Like, it’s hard to take this seriously when you improve his grade from -5.4 to -2.9 in two days. Seems like they know the original grade was dumb.

    • KWS13

      Getting lucky isn’t always being aggressive. “Taking what the defense gives you” is a common moniker for QBs, and it doesn’t mean dink and dunk. There’s a reason offensive coordinators exist, and it’s to find ways to beat defenses. Robinson’s route exists, for example, to a) give an option to take a shot against single high or cover 0 (allowing you guy to make a play), b) clear out space, or c) exploit blown coverage against zone. None of those things happens for Bortles, but he says “a screw the offense I’ll throw it anyway.” Further, with Robinson double covered and an open guy heading to the flat the throw should never have been made in that situation. When “good” QBs make “aggressive” throws its because they know they can reduce the risk by knowing the situation and using the right reads. It isn’t “easy” to make the right read, it’s smart- it’s easy to just ignore the reads and take ill advised shots downfield when they aren’t open or necessary.

    • crosseyedlemon

      You can’t know whether a grade shift is significant unless you know what scale is being used. A shift from -5.4 to -2.9 is significant if the overall scale ranges from -10 to +10 but not nearly a significant if the scale is much wider.

  • crosseyedlemon

    Maybe my math skills are even worse than I thought but I only counted 10 Colts on defense in that first screen shot….where is the 11th guy?

    • HenryT

      6 in the box and 5 dbs =11

    • PFFSamMonson

      Off screen as a deeper safety.

      • crosseyedlemon

        That must be it Sam. If the Colts were playing the game with only 10 on defense giving up 51 points doesn’t look nearly as bad.

  • David Appelson

    Great stuff.

    • Anyone sane


  • Keith

    30 TDs so far. Must be the luckiest QB to ever live

    • The Mysteries of Bob

      That’s 2013 Nick Foles.

      He threw 27 TDs, has thrown 20 since.

    • Jaguars28

      But Bortles is good

  • Turing

    I’m not saying it’s a total waste of time, but it’s impossible to have a grading system like this without expectation bias creeping in.

  • Joe

    I enjoy PFF’s articles but there are a couple of points I disagree with here. I don’t agree that you can call the deep pass to Robinson as bad decision just because other QB’s sometimes get picked off in the same situation. The fact remains he gave his receiver (One of the NFL’s best at that) a chance to make a deep play. If you are going to criticize the decision, surely you have to reward the execution?

    Secondly, this doesn’t take into account the value of deep passing in terms of opening up the offense and ensuring a defense respects the threat of the deep pass. I would be very confident that Greg Olson instructs Bortles to throw deep and often because he knows it creates not only big plays, but space underneath and help for the running game.

    I’m no Jag’s fan and perhaps PFF recognize these limitations above, but it seems a fairly big thing to miss.