Writers of Pro Football Prospectus 2008

Most Recent FO Features

SandersEmm10.jpg

» Scramble for the Ball: Quarter Pole Projections

Mike and Tom weigh the chances of this year's class of receivers, running backs and tight ends who are on pace to break the magical 1,000-yard mark for the first time.

14 Mar 2012

Adjusted Interceptions 2011

by Aaron Schatz

Last year, we introduced a new metric on FO called "adjusted interceptions." The basic idea:

  • We add in plays where the quarterback only escaped an interception because the defender couldn't hold onto the ball (dropped interceptions, which we've been tracking in game charting since 2007).
  • We subtract plays where the interception is tipped to the defender by a receiver who should have caught the pass.
  • We subtract Hail Mary interceptions as well as interceptions thrown in desperation on fourth down in the final 2:00 of a game.

Obviously, there's going to be a little bit of subjectivity here, but our game charters do their best. If the defender has to dive for a ball only to have it bounce off his fingertips, that's not a dropped interception.

Interceptions are notoriously hard to forecast from year to year, because there's so much random chance and statistical noise involved. However, we did find that this new metric, adjusted interceptions, had a higher year-to-year correlation than standard interception totals and was a better predictor of future interceptions.

However, that wasn't as true from 2010 as it was in the previous four years. Looking only at quarterbacks with at least 100 pass attempts in both seasons, the correlation coefficient for adjusted interception rate from 2007-2010 is .33, while the correlation coefficient for regular interception rate is .08. Looking just at 2010-2011, the correlation for adjusted interception rate is .15, while the correlation for regular interception rate is -.08. Yes, that's right; that's the more often you threw picks in 2010, the (slightly) less often you were likely to throw picks in 2011.

No quarterback came close to the 15 dropped interceptions we counted for Mark Sanchez during the 2010 season. Four quarterbacks tied for the league lead with nine dropped interceptions: Joe Flacco, Cam Newton, Carson Palmer, and Matthew Stafford. Of course, three of those guys played a full season, so I think it is safe to say that Carson Palmer is our dropped interception champion for 2011. His adjusted interception rate of 7.1 percent leads all quarterbacks with at least 200 pass attempts, although Caleb Hanie had 8.1 percent with 99 pass attempts. (Hanie had 102 pass attempts by official NFL totals, but we're not including clock-stopping spikes.)

And what about Sanchez? On the surface, it looked like Sanchez was making worse decisions than the year before. His interception total rose from 13 to 18. However, look at adjusted interceptions and Sanchez actually improved in 2011. Our game charters only registered three dropped interceptions thrown by Sanchez in 2011, despite more pass attempts. His adjusted interception rate is much lower than it was in 2010.

This year's most remarkable dropped interception total was a low total: zero. Aaron Rodgers had zero dropped interceptions by our count. In the five years we've counted dropped interceptions, he's the only quarterback with more than 400 attempts and no dropped picks. Now, obviously I'm sure our game charters missed a dropped pick here or there and it is possible one of those we missed was thrown by Rodgers. Readers are welcome to leave comments with plays they think we should check for dropped picks. (Be specific about game, quarter, and time.)

The other remarkable number in this year's total belongs to Tarvaris Jackson. We count four of Jackson's 13 picks as Hail Mary attempts. No other quarterback this year had more than one. For those Seattle fans who are curious, the four plays are: the last play of the first half in Week 1 vs. SF, the last play of the first half in Week 3 vs. ARI, the last play of the game in Week 9 vs. DAL, and the interception thrown with 0:49 left in the fourth quarter and the Seahawks down by 15 to Cincinnati in Week 8, the pass returned by Reggie Nelson for a touchdown. That last one isn't what we conventionally would consider a "Hail Mary," but I coded it as such because it was nigh impossible for Seattle to win the game at that point.

The following table lists adjusted interceptions for all quarterbacks with at least 200 pass attempts:

Name Team INT HM/
End Q4
Drop INT Tip INT Adj INT Pass Att INT Rate Adj Rate 2010
INT Rate
2010
Adj Rate
C.Palmer OAK 16 0 9 2 23 326 4.9% 7.1% 3.2% 4.5%
J.Skelton ARI 14 0 5 1 18 274 5.1% 6.6% 1.5% 4.4%
C.Ponder MIN 13 1 5 0 17 291 4.5% 5.8%  
R.Grossman WAS 20 1 7 1 25 456 4.4% 5.5% 2.8% 2.8%
K.Orton DEN/KC 9 0 4 0 13 252 3.6% 5.2% 1.7% 3.1%
C.Painter IND 9 0 3 0 12 243 3.7% 4.9%  
C.Newton CAR 17 0 9 1 25 514 3.3% 4.9%  
K.Kolb ARI 8 0 3 0 11 252 3.2% 4.4% 3.5% 3.5%
M.Vick PHI 14 1 6 1 18 418 3.3% 4.3% 1.5% 3.0%
P.Rivers SD 20 0 4 0 24 579 3.5% 4.1% 2.2% 2.7%
B.Gabbert JAC 11 0 6 0 17 412 2.7% 4.1%  
J.Freeman TB 22 0 2 2 22 546 4.0% 4.0% 1.2% 2.5%
B.Roethlisberger PIT 14 0 7 1 20 514 2.7% 3.9% 1.2% 1.9%
R.Fitzpatrick BUF 23 0 3 4 22 566 4.1% 3.9% 3.2% 4.5%
M.Hasselbeck TEN 14 0 7 1 20 517 2.7% 3.9% 3.6% 4.4%
M.Stafford DET 16 0 9 0 25 663 2.4% 3.8% 1.0% 3.0%
J.Flacco BAL 12 0 9 1 20 543 2.2% 3.7% 1.9% 2.4%
Name Team INT HM/
End Q4
Drop INT Tip INT Adj INT Pass Att INT Rate Adj Rate 2010
INT Rate
2010
Adj Rate
T.Jackson SEA 13 4 7 0 16 451 2.9% 3.5%  
M.Sanchez NYJ 18 1 3 1 19 539 3.3% 3.5% 2.4% 5.2%
M.Moore MIA 9 1 4 0 12 348 2.6% 3.4% 6.3% 6.3%
S.Bradford STL 6 0 6 0 12 357 1.7% 3.4% 2.4% 2.7%
A.Dalton CIN 13 0 5 1 17 516 2.5% 3.3%  
C.McCoy CLE 11 0 4 0 15 463 2.4% 3.2% 3.7% 4.5%
M.Cassel KC 9 1 2 2 8 268 3.4% 3.0% 1.5% 2.7%
E.Manning NYG 16 0 2 1 17 589 2.7% 2.9% 4.4% 4.3%
T.Tebow DEN 6 1 2 0 7 268 2.2% 2.6%  
D.Brees NO 14 0 5 2 17 653 2.1% 2.6% 3.2% 3.8%
J.Cutler CHI 7 0 2 1 8 314 2.2% 2.5% 3.3% 4.9%
T.Brady NE 12 0 4 1 15 611 2.0% 2.5% 0.8% 1.5%
T.Romo DAL 10 0 2 0 12 519 1.9% 2.3% 3.2% 3.6%
M.Schaub HOU 6 0 1 1 6 290 2.1% 2.1% 2.0% 2.0%
M.Ryan ATL 12 1 2 2 11 561 2.1% 2.0% 1.5% 2.6%
A.Smith SF 5 0 2 1 6 447 1.1% 1.3% 2.7% 2.5%
A.Rodgers GB 6 0 0 2 4 501 1.2% 0.8% 2.2% 3.0%

Here's a secondary table of quarterbacks with 90-200 pass attempts.

Name Team INT HM/
End Q4
Drop INT Tip INT Adj INT Pass Att INT Rate Adj Rate 2010
INT Rate
2010
Adj Rate
J.Beck WAS 4 0 2 0 6 131 3.1% 4.6%  
J.Campbell OAK 4 1 0 0 3 164 2.4% 1.8% 2.2% 3.6%
K.Clemens STL 1 0 0 0 1 91 1.1% 1.1%  
K.Collins IND 1 0 0 0 1 96 1.0% 1.0%  
A.J.Feeley STL 2 0 1 0 3 97 2.1% 3.1%  
C.Hanie CHI 9 1 1 1 8 99 9.1% 8.1%  
C.Henne MIA 4 1 1 0 4 112 3.6% 3.6% 3.6% 3.4%
D.McNabb MIN 2 0 0 0 2 156 1.3% 1.3% 2.9% 3.7%
D.Orlovsky IND 4 0 0 0 4 190 2.1% 2.1%  
T.Palko KC 7 0 0 1 6 133 5.3% 4.5%  
S.Wallace CLE 2 0 1 0 3 105 1.9% 2.9% 1.9% 2.8%
T.J.Yates HOU 3 0 0 0 3 132 2.3% 2.3%  
V.Young PHI 9 0 0 1 8 114 7.9% 7.0% 1.7% 2.8%

Posted by: Aaron Schatz on 14 Mar 2012

39 comments, Last at 19 Mar 2012, 8:48pm by Justsomeguy

Comments

1
by max mulitz (not verified) :: Wed, 03/14/2012 - 5:13pm

In fairness, Palmers number is deceptive based on his week 7 3 interception performance where he didn't know the playbook. Not that he looks that much better after that anyway...

2
by are-tee :: Wed, 03/14/2012 - 5:14pm

This info comes too late for Sanchez. The "he regressed in 2011" mantra has been repeated so many times that it must now be accepted as fact. And this despite that his touchdown production (26 passing and 6 rushing)increased dramatically from 2010. He also had the same number of fumbles as in each of his first two seasons, but we all know about fumble recovery luck (that took a huge turn for the worse). Well, at least he got his contract extension in the end.

4
by Eddo :: Wed, 03/14/2012 - 5:43pm

His DVOA and DYAR did decrease from 2010 to 2011 though.

6
by Drakos (not verified) :: Wed, 03/14/2012 - 6:08pm

But aren't his 2010 DVOA and DYAR based on his low actual interception rate and not his higher adjusted rate?

12
by Theo :: Wed, 03/14/2012 - 8:07pm

I won't accept mantras as fact because they are repeated enough. The arguments though... that's another story.

19
by CoachDave :: Thu, 03/15/2012 - 9:34am

The only tired "mantra" I see on this site is your repeated cherry picking of stats to minimize his crappy 2011 performance.

He is a below average QB who got worse after another season of starts, training camp and preparation...you debating against some self-invented mantra isn't going to change that.

And here's the thing about the "luck" of recovering fumbles...if your boy didn't put the ball on the turf so often, it would be a moot point.

23
by are-tee :: Thu, 03/15/2012 - 11:50am

I certainly did not "invent" the idea that Sanchez regressed between 2010 and 2011 - I've heard and read enough commentary from the local and national media to that effect. And I'm not saying that Sanchez had a good year. He did not improve significantly the way most people expected (or hoped for),especially in light of his performance in the 2010 post-season (which was very good). The premise of the article's commentary re. Sanchez was that he actually improved in the area of interception rate when you take into account drops.

3
by John Courage :: Wed, 03/14/2012 - 5:41pm

"although Caleb Hanie had 8.1 percent with 99 pass attempts. (Hanie had 102 pass attempts by official NFL totals, but we're not including clock-stopping spikes.)"

But are you counting pump fake, intentional grounding non-spikes?

Ahh, Caleb Hanie. He's almost bad enough to make one nostalgic about the Grossman/Orton days. Almost.

5
by tuluse :: Wed, 03/14/2012 - 5:54pm

Don't look now but Jay Cutler is firmly in the bottom quartile of adj interception rate.

9
by TomC :: Wed, 03/14/2012 - 7:38pm

Yeah, and check out the 2009-2011 trend:

year #INT INT % (non-adj)

2009 26 4.7
2010 16 3.7
2011 7 2.2

I *think* that's real, not just statistical noise.

20
by Steve in WI :: Thu, 03/15/2012 - 10:55am

That really impressed me. I knew that he threw relatively few interceptions last year (and of course, playing only 10 games helped with that) but my perception was that he got lucky with a lot of dropped interceptions, and that wasn't really the case I guess.

7
by Led :: Wed, 03/14/2012 - 6:37pm

Curious: Did the same game charter(s) chart Sanchez in 2010 and 2011? I wouldn't be surprised if some charters are more forgiving to QBs than others in terms calling something a dropped interception.

8
by zlionsfan :: Wed, 03/14/2012 - 6:53pm

Not completely, I would guess: typically games are split between charters (1 half per charter), so anyone on the Jets' schedule in 2011 and not in 2010 would likely have a new charter. (For example, I did the first half of Lions-Jets in 2010.) Plus there's charter turnover from year to year, and I don't recall if the Jets' "regular" charter was the same this year to last.

So sure, there will be some differences from year to year, although the FO folks give us consistent instructions (of course) and do what they can to point out where our tendencies differ from those of the group as a whole. There are also occasions where our instructions change from year to year, but I don't believe that involved dropped INTs either year.

10
by jonnyblazin :: Wed, 03/14/2012 - 7:56pm

Do QB's with stronger arms have more dropped INTs? Maybe their balls are harder to catch. Here are the top QB's with the most dropped INTs:

Flacco
Stafford
Palmer
Cam Newton
Big Ben
T-Jax
Hasselbeck
Grossman
Vick
Bradford
Gabbart

I'd say of that list, Flacco, Stafford, Ben, and Vick have some of the best arms in the league. The others are either inexperienced (Cam, Bradford, Gabbo) or INT-prone (Grossman, T-Jax). Not sure where Hasselbeck fits. But I've seen Vick throw lasers that LB's drop, I just don't think they have the hands to deal with his velocity.

11
by tuluse :: Wed, 03/14/2012 - 8:05pm

Newton has a pretty strong arm I would say (doesn't change the fact that he's also inexperienced).

13
by Jerry :: Wed, 03/14/2012 - 8:09pm

The question that leaps to mind is "Do dropped interceptions correlate with dropped passes?" Even if receivers presumably catch a higher percentage than defenders, I'd still guess that a guy whose passes are hard to catch is hard(er) for anyone to catch.

16
by justanothersteve :: Wed, 03/14/2012 - 9:53pm

That's an interesting question. During Favre's prime in the 90's, I don't know how many INTs I saw dropped because the ball was thrown so darn hard the DB couldn't adjust. I even think that may be why sometimes Favre didn't mind flinging it in coverage; he probably thought that even if the defense got their fingers on it, they couldn't handle the ball anyway. He used to break his own WRs fingers. I think some of the other gunslingers may feel the same way.

18
by the cat in the box is dead (not verified) :: Thu, 03/15/2012 - 9:02am

That's also true for Donovan McNabb, another passer with a really low INT rate. I lost count of the number of times he would throw a 5-yard in like he was trying to take the receiver's head off.

35
by JonFrum :: Fri, 03/16/2012 - 2:13pm

On the other hand, you might note that Favre holds the career record for total INTs. Obviously, that bazooka of an arm didn't always protect him from sticky DB hands.

36
by Shattenjager :: Fri, 03/16/2012 - 2:22pm

However, he also has a 98 career INT%+, essentially average. Especially when you consider that his interception rate was much higher 2003-2010 than previously, it actually lends credence to the idea that he had a lot of dropped picks.

Much as I detest him, he has the career INT record because he threw 21.7% more passes than anyone else in history.

14
by db :: Wed, 03/14/2012 - 9:12pm

Without knowing the play call so as to ascertain whether the receiver made the correct read and was in the proper spot this is meaningless. Add to that the subjective nature of the bad throw or bad hands argument and it is totally without merit as a statistic.

17
by Will Allen :: Wed, 03/14/2012 - 9:59pm

We don't know whether interceptions are due to a receiver or qb error, either, but that doesn't stop us from incorporating interceptions in qb stats, nor should it.

Yes, there is subjectivity as to what is a dropped interception, but it is too important of an event to ignore.

25
by jfsh :: Thu, 03/15/2012 - 12:07pm

I think the idea is not that this is the ultimate statistic, but rather that it is more helpful than raw interception totals.

Which it is, judging by the correlation coefficients.

15
by Dolphinsfan44444 (not verified) :: Wed, 03/14/2012 - 9:52pm

Not to be a jerk, but I think M. Moore played for Miami last year, not Carolina.

33
by Aaron Schatz :: Thu, 03/15/2012 - 4:23pm

Whoops. Will fix.

21
by Steve in WI :: Thu, 03/15/2012 - 11:00am

One thing that jumps out at me is that the vast majority of QBs have a higher adjusted interception rate than actual interception rate. My gut based on watching games is that it would be the opposite, but it's obviously more common for a defensive player to drop what should be a pick than for a QB to throw one that's not his fault because a catchable ball got tipped off his receiver's hands or it was a Hail Mary/end-of-game desperation play.

Of course, as someone pointed out, this doesn't factor in receiver error, so presumably if it was possible for that to be included the adjusted rate would drop some. (But then, wouldn't we have to consider plays in which the receiver makes an awesome interception-saving play on a bad throw? Boy, am I glad I'm not a game charter.)

22
by tuluse :: Thu, 03/15/2012 - 11:04am

Eventually you're talking about needing the a coach to walk you through the play and tell you were each receiver was supposed to be, what the QBs keys were supposed to be, etc.

24
by DenverCheeze (not verified) :: Thu, 03/15/2012 - 12:00pm

The QB should be able to account for Receiver error at the pro level dont you think? If we are going to consider them 'elite' qb's and they have practice regularly then it is their job to know if the receiver is (or is not) going to run the right route and adjust accordingly. The QBs that dont do this have higher interceptions rates...is that the receiver's fault? Why would a QB throw a pass blindly assuming the inexperienced receiver is going to make the right play (assuming the experienced receivers do this much much less often)? If he is throwing before the receiver breaks to an inexperienced receiver I am guessing the QB is either not as good as others (therefore more interceptions) or has a high level of confidence that he and the receiver are on the same page (lower interceptions). If he continues to throw before the receiver breaks after being intercepted, then he is also not a good QB because he is not learning from mistakes. My point (restating) is that even receiver running the wrong pattern is still the QBs fault and should continue to factor into the interception rate for the QB.

26
by tuluse :: Thu, 03/15/2012 - 12:08pm

Well I don't think QBs should take as much blame for interceptions as you do.

Even Marvin Harrison has ran the wrong route on occasion.

27
by DenverCheeze (not verified) :: Thu, 03/15/2012 - 12:23pm

My point is that Peyton (when Marvin ran the wrong route) still didnt throw the int because he is a better QB than say...Caleb Hanie who likely would have in the same situation.

28
by tuluse :: Thu, 03/15/2012 - 12:40pm

2006 playoff game between the Colts and Chiefs. Peyton threw two picks because Harrison cut the opposite direction he was expecting.

30
by DenverCheeze (not verified) :: Thu, 03/15/2012 - 1:52pm

What you don't know is who was at fault...did Harrison run the wrong route or did Peyton misread the defense and positioning of the defender? I am saying that this is the responsibility of the QB and should be reflected in his interception metric.

31
by tuluse :: Thu, 03/15/2012 - 1:55pm

"What you don't know is who was at fault"

That's right, and I'm not trying to determine it really. You're the one who think it's 100% the QBs fault.

37
by LionInAZ :: Sat, 03/17/2012 - 4:22pm

Have you ever heard of timing plays? Sometimes a receiver runs the wrong route, or gets bumped off the route. It's not necessarily the QB's fault.

29
by gtliles82 :: Thu, 03/15/2012 - 12:48pm

Stafford played with a broken finger against Carolina and Chicago, and racked up 6 of his 16 INTs in those games. I'd be interested to know how many dropped INTs were also in those games.

32
by Nevic (not verified) :: Thu, 03/15/2012 - 2:19pm

Breaking news: Aarom Rodgers is GOOD!

34
by Will Allen :: Fri, 03/16/2012 - 11:17am

It might be interesting to see an interception rate study that attempts to determine if there is a correlation, or how strong a correlation, between a qb throwing bad passes, and the qb's defense giving up a lot of first half points that did not follow the qb's interceptions. That is, if a team's defense stinks the joint out in the first half, how much does the team's qb's adjusted interception rate climb, due the qb being in a situation where his offense has to go downfield?

It seems to me that our qb evaluations don't yet do a good job of reflecting the difference in difficulty of playing qb when the qb's defense is allowing a lot of early scoring, and playing qb when the qb's defense is not only preventing scoring, but getting the ball back to the qb quickly. In other words, the gap between Tom Brady and Joe Flacco last year may have been significantly larger than what DYAR and DVOA suggest, even though those stats account for the strength of the opposition.

38
by LionInAZ :: Sat, 03/17/2012 - 4:25pm

I agree that would be interesting to look at. One could even start simply, by counting 1st half INTs versus 2nd half INTs.

39
by Justsomeguy (not verified) :: Mon, 03/19/2012 - 8:48pm

Eli with only 2 dropped INTs? Seemed like he had a lot more than that. What were his 2 dropped INTs according to the charters? I'm assuming the T-New dropped INT and...the one he had against the Redskins?