Signature Stat – Accuracy Percentage

| 6 years ago

Signature Stat – Accuracy Percentage

This past spring we introduced Accuracy Percentage, and two months ago we went over it again looking at just one week of this season’s data. However this Signature Statistic is so interesting, that we’re looking at it again.

This time we’ve got so much more data it boggles the mind. Whereas after Week 1 we have only 16 games worth of numbers, now we’ve got a frankly quite awesome 129 (after adding San Diego at Oakland to the mix). So you’re not just looking at a one-off with this article, you’re looking at patterns and trends of a very decent sample size.

So take a load off and get ready for the return of the PFF Signature Stats where we’ll look at two of the most accurate passing seasons we’ve seen, one of the best games we’ve seen, and give Tim Tebow haters more ammunition. Now, where to start …

Accuracy, thy name is Rodgers

Before we drop the bombshell of all bombshells let me quickly remind you what the Accuracy Percentage is. It is, as simply as possible, the Pro Football Focus way of ‘fixing’ the completion percentage stat which can so badly misrepresent a player. Our ‘Accuracy Percentage’ excludes all dropped passes, throwaways and spikes, leaving you with a stat which is all about looking at the accuracy of a quarterback, looking only at passes which are actually aimed at a receiver. Now let’s start with the exceptional shall we?

It may have escaped your attention but Aaron Rodgers is having a pretty good season. As close to flawless as we have seen anyway during our time as he has completed an incredible 81.3% of all his aimed passes. We all know Tom Brady was pretty good last year, but he only completed 76.1% of aimed passes in 2010 (the highest percentage along with Drew Brees). So Rodgers (who was third in 2010 with a 74.4% rating) is redefining expectation right now with the kind of quarterback play that is as good as you’re likely to see from anyone. That’s enough praise for him right now as there’s more to come later. Instead, let’s turn our attention to other members of the Top 5.

Overall - Top 5

RankNameTeamAtt.CompDropsTASPAcc. %
1Aaron RodgersGB265192168181.3
2Drew BreesNO379269186377.6
3Alex D. SmithSF206132186075.0
4Michael VickPHI266165228273.0
5Andy DaltonCIN2571582011072.4


A Rookie Revelation

Perhaps the most interesting name there is Andy Dalton. The Cincinnati Bengals’ QB (who jumped up to third in our Race for Rookie of the Year) is a great example of how the Accuracy Percentage stat works – the rookie is nowhere near the top of any list for the common measures of a quarterback with a completion percentage of 61.5% (13th highest mark in the league). But, break it down like the Accuracy Percentage does and you start to appreciate his season all the more. When the 20 passes Dalton’s receivers have dropped, and the 11 balls he has thrown away rather than force, are taken away, he’s a healthy fifth place with 72.4% of his passes ending up complete.


Searching for that barn door

So you’ve got a glimpse of some of the good, now let’s move onto some of the bad, and it shouldn’t surprise anyone who is at the “top” of that list. Have a look for yourself and see. Tim Tebow has been criticized  for his accuracy, and the numbers completely back that up. While he may have gone viral, and he led his team to an impressive fourth quarter comeback against the Dolphins, he’s also managed some of the most impressive (or depressive) overthrows of the year. He may be a special talent, but one of those talents doesn’t center around being accurate.

He’s followed by a rookie quarterback in Blaine Gabbert who is, to put it kindly, taking his lumps with a spell of inaccurate quarterbacking that is as bad as it is consistent. Still you don’t get to a really interesting name until you see that of Joe Flacco. Yes indeed, Flacco is struggling so much in accuracy terms that you’re comparing him to names like Tebow and Gabbert. Quite the fail as the kids say.

Overall - Bottom 5

RankNameTeamAtt.CompDropsTASPAcc. %
36Tim TebowDEN974546154.4
35Blaine GabbertJAX17379147156.4
34Kerry CollinsIND984854257.6
33Curtis PainterIND195104117061.2
32Joe FlaccoBLT3091691713062.8


Practically Perfect

If you want some truly mind-boggling numbers regarding quarterback accuracy you need to look at individual games. More precisely you need to turn your attention to the Packers and the Vikings, where that man again, Aaron Rodgers, was as close to perfect as it gets. He had just six incomplete passes where two were drops, two were thrown away, and one was spiked. Yep, that leaves just one incomplete aimed pass all game. Credit also to Drew Breess who has two of the top five individual performances, while the ‘where-did-that-come-from’ award goes to Donovan McNabb. The one-time Viking starter helps show that accuracy isn’t everything by doing his utmost not to throw incompletions (as opposed to challenging a defense). Here’s the Top 5:

Individual Games - Top 5

RankNameTeamOppWeekAtt.CompDropsTASPAcc. %
1Aaron RodgersGBMIN7302422196.30
2Andy DaltonCININD6322540090.63
3Drew BreesNOIND7353100088.57
4Drew BreesNOTB9362731188.24
5Donovan McNabbMINCHI6241920087.50


Not so fearsome Jaguars

Now for the really bad. Jacksonville fans may not like this, but it seems so long ago that David Garrard finished in the Top 10 of our Accuracy Percentage (both 2008 and 2010). They have three of the worst five individual performances and it’s only that truly atrocious performance from Mark Sanchez against Baltimore that prevents them having the worst. Given how well the Jags’ defense is playing, it’s scary to think where some competent play could have led them this year.

Individual - Bottom 5

RankNameTeamOppWeekAtt.CompDropsTASPAcc. %
1Mark SanchezNYJBLT4351112036.36
2Luke McCownJAXNYJ219610036.84
3Blaine GabbertJAXHST8301023044.44
4Rex GrossmanWASPHI622902045.00
5Blaine GabbertJAXNO4421630045.24

The Bottom Line 

Since the NFL started keeping track of passes attempted and completed in 1932, the incomplete pass has been a negative for a quarterback. It’s always important to realize why a pass is incomplete, and sometimes a quarterback shouldn’t be downgraded for it. We like to think of it as adding a bit of intelligence to a stat that can be otherwise easily manipulated. It might seem like taking these into account doesn’t make a huge change, but it does. Just look at how accurate (pun intended) a picture it paints of Andy Dalton.


Follow us on Twitter: @ProFootbalFocus


| Director of Analytics

Nathan has been with Pro Football Focus since 2010. He is the Director of Analytics, an NFL analyst, and a fantasy writer.

  • hnrqoliv

    Wow, it’s just one stat, but Andy Dalton being in the same discussion with Brees and Rodgers is pretty encouraging signs considering he’s just a rookie. Not to mention a rookie without any minicamps or OTAs…

  • temporaryusername

    I don’t know that there’s a good way to quantify this, but is there any way to incorporate sacks that result from the QB simply holding on to the ball too long? As a Packer fan, the one thing that drives me crazy about Aaron Rodgers accuracy stats throughout his career is that it seems like he often holds on to the ball too long. It may be smarter in the long run to just take a sack rather than risk a turnover in what are often difficult throw-aways under pressure, but if his offensive line statistics suffer as a result why don’t his own? Maybe there’s a way to categorize sacks — coverage, missed blocking assignments, line breakdowns, overwhelming pass rush, QB extension (sacks > 3 seconds), etc.

    Thanks for your consideration.

    • Nathan Jahnke

      While sacks are incorporated into the PFF pass rating, I don’t think this statistic is the place for it. Accuracy Percentage is solely trying to quantify how well a quarterback get the ball to the receiver he chooses to throw to. This isn’t supposed to say that every quarterback ranked high in this statistic is better than every quarterback who ranks low, it is just showing part of their game. Something like decision making isn’t a factor in this statistic, because it doesn’t factor things into play like did he throw it to the best possible receiver on that play, or how long he held onto the ball before throwing it. If a quarterback gets sacked, he doesn’t have the chance to get the ball to his intended receiver, so I don’t believe sacks should be included here. Of course it says something about a quarterback, just not his accuracy.

  • urrymonster

    These are pretty interesting statistics and it’s phenomenal what Rodgers is achieving this year. One thing that does scew these stats though are the ‘west coast’ QBs and the guys with the offensive systems that take advantage of matchups to pick up easy gains. Would you be able to considering accuracy per distance increment? I only ask because someone who throws double the amount than average under 10 yards is going to have a statistical advantage. How about including and over under the average for starters etc… on each distance as well?

    • Nathan Jahnke

      That is all very true. Clearly this doesn’t account for difficulty of throw(which part of that is distance), as well as other things such as if a quarterback was a little off on the pass and the receiver just made a great play. Therefore this isn’t a set it stone, player ranked 13th is more accurate than player ranked 16th… but still an improvement to completion percentage.

      With looking at things by distance, that is something that I am looking to do more once it’s the off-season. When you break things down by distance, the sample sizes for each player in each category gets smaller. While plenty of players might have a big enough sample, I just think at the end of the season we’ll have an even better idea, and see if there is a need to and an ability to improve on accuracy percentage.

      • urrymonster

        Awesome, look forward to that. Like you say this article is a great step up from looking at base completion percentages. Keep up the good work!