"Good game, Andy, but you might want to throw to your running backs more often... - Frank Victores-US PRESSWIRE
In which we go a bit deeper than the raw stats and take a final look before bidding them adieu.
Disclaimer—statistics can help you look at something which has already happened. It can't tell you why, necessarily, or if it will happen again. There is utility in looking at these things, however, including possibly seeing a direction the team ought to take in terms of drafting and/or coaching.
My original plan for this post was to move on to the wide receivers, and so I began comparing their drop rates. Earlier in the season I had noted that some quarterbacks seemed to throw a great many less "catchable" balls than others. I do this based on the Pro Football Focus Signature Stat for receivers of "drops".
So what is this stat? Rather than just noting how many balls a given receiver did or did not catch, the fine folks at PFF analyze each play in which the receiver was targeted and try to determine who was at fault, in essence. If they decide the receiver didn't have a chance to catch the ball they remove it from the list of his drops.
Obviously this is to some extent a subjective determination on their part. The ball may not have been in position for the receiver to catch it, but was that because the quarterback made a poor throw because his arm was hit, or the ball was out of position because it was very slightly tipped at the line of scrimmage, or because the receiver was held just long enough to make him unable to get to the right spot (called or not?) Conversely, did the receiver run a very convincing-looking route, only it was the wrong one? Without knowing every circumstance of every play, I would presume they get it wrong on occasion. I'm also guessing they get it right often enough so useful information can be derived from it.
In other words, I think we can look at a couple of things about the quarterbacks using this stat. First of all, who are they throwing to most often? What percentage of the balls they throw are deemed to be "catchable"? Does this correlate well with their completion percentage? And are they generally more accurate when throwing to tight ends and running backs, or less? The latter would seem to be counterintuitive, as presumably most of those throws are short, high-percentage ones. Finally, we can look at who (or what position group) they mainly threw to, and perhaps determine if deficiencies in any of the position groups need to be addressed in the draft.
Let's begin with throws to wide receivers. The average number of throws to wideouts during the season in the AFC North was 312. Pittsburgh was closest to the average, with 318. (This includes all quarterbacks.) Andy Dalton was way over the average, with 352, and both Joe Flacco and Brandon Weeden (read all Cleveland quarterbacks) were well under, with 288 targets apiece. Of these targets, what percentage was deemed by Pro Football Focus to be "catchable?" Let's have a look. [Note all of the Catchable Balls charts begin at 50% and end at the appropriate level.]
The blue bar is the league average for balls considered to be catchable. Interestingly, the entire AFC North, with the exception of Ben Roethlisberger, was at least slightly below the average. Joe Flacco is distinctly subpar in this stat. Although in theory it should have nothing to do with this, I threw in the average drop rate for all wide receivers from each team, and every team except, naturally, the Browns have a lower-than-league-average drop rate.
But quarterbacks throw to more than just wide receivers. Let's take a look at tight ends and running backs. The presumption is, they are thrown to less often, but the accuracy should be higher, as the greater percentage of the throws should be short, high-percentage throws. Indeed, this is true league-wide. The average percentage of throws to wide receivers deemed to be "catchable" is 67.18%; to tight ends it is 74.85%, and to RBs it is a whopping 86.28%.
So do our QBs follow this pattern? Here are the TEs:
Once again we see Joe Flacco's accuracy is well below the league average when throwing to his TEs, with Brandon Weeden better but not great. Andy Dalton is well above the league average, and Ben Roethlisberger and the PIT quarterback stable in general are far above the crowd. And once again the Baltimore TEs are far better at holding on to passes close enough to catch than their AFC North peers, although Pittsburgh is reasonably close.
What about the running backs?
This is the category in which Joe Flacco is closest to the league average, being just below it. But he is far outdone in the Hall of Shame, if you will, by Ben Roethlisberger, who barely managed to get three out of four throws to RBs in the general vicinity of the back. As you can see, the other Pittsburgh quarterbacks did enough better at this that the overall average is slightly better than the league average. Odd, that. The PIT running backs were also far more likely to drop passes from Ben than from the other quarterbacks.
The AFC North quarterbacks threw to the running backs an average of 96 times. (The league average was 93.) Once again Baltimore was well over the average, with 117 targets to backs. Andy Dalton, on the other hand, was at 58% of the average, with only 54 attempts to connect with one of his running backs. Both Pittsburgh and Cleveland were well above the average (103 and 110 respectively) which is the only reason the AFC North was anywhere close to the league average.
Here are comparison charts that make the balance easier to see:
Now let's compare the "Catchable Ball" rate with the completion percentage for each quarterback.
The league's most "average" quarterbacks for 2012 are Carson Palmer (61%) and Ryan Fitzpatrick (60.6%.) Ben Roethlisberger is in the top ten quarterbacks in completion percentage. The top quarterback, Alex Smith, isn't in the playoffs, but his team is, and he was instrumental in getting his team there. Tony Romo, Philip Rivers, and Ben are the three quarterbacks in the top ten who weren't in the playoffs. Joe Flacco and Andrew Luck are the only two quarterbacks who were below average in Completion Percentage and yet ended up in the playoffs. Clearly the accuracy of your quarterback isn't everything..
I'm not done here, although some of you may have nodded off a long time ago : ) I have broken down the information even further to see exactly who each quarterback was throwing to, and whether we see a problem with this. Here's how often each quarterback targeted his top three receivers, compared to the league average:
This is how many times the receivers were thrown to, compared to the average. This chart is by actual target, so Andrew Luck chucked the ball downfield almost 400 times.
Note the charts are percentages now. These two begin at zero and end at 60%. Andy Dalton is definitely a one-receiver man. Joe Flacco not only threw fewer targets than average, but he spread them almost equally between Anquan Boldin and Torrey Smith. The same can be said for Brandon Weeden, who split the targets between Josh Gordon (who ended up as the No.1, just, and Greg Little.) I think the tally would have been more even had Antonio Brown not been out for several weeks. In the event, Mike Wallace ended up as the clear No. 1.
I wondered whether it's a good thing or a bad one to be dependent upon a single receiver. Instinctively, it seems bad, as you may be in trouble if said receiver is hurt. So I checked out which teams, league-wide, threw a lot more to a single receiver. The Bears and the Lions both heavily favored a single receiver. Calvin Johnson and Brandon Marshall were the most heavily targeted receivers in the league, both numerically and compared to the other receivers on their team. Johnson was thrown to more than three times more often than the No. 2 receiver on the team, and Marshall almost four times more. Neither team made it to the playoffs.
Cincinnati falls in a group of teams who targeted their No. 1 receiver about twice as often as their No. 2. Three of those made the playoffs: the Texans, San Francisco, and of course the Bengals. The other two are picking early in this year's draft: Buffalo and the Jets. So I suppose I should stop getting on Cincinnati about being so reliant on A.J. Green. But there may be mitigating circumstances, which we will look at later.
There were six teams who targeted their No. 1 and No. 2 receiver within a narrow margin. (I put a team in this category if the No. 2 receiver was within 10% of the No. 1 receiver's score. In other words, Josh Morgan was targeted 24.8% of the time, Pierre Garcon 22.7%. Adding 2.5 to 22.7 gives you 25.2, or more than 24.8, so I considered them to be equal. And if Pierre Garcon hadn't been injured for a chunk of the season he would almost certainly have been the No. 1, but that's another story.) The six teams to fall into this category were Atlanta, the Ravens, Cleveland, the Titans, Green Bay, and the Redskins. Four play-off teams, two not so much...
Here is a comparison of the No. 3 receiver targets and of all the rest of the receivers (which was anything from one more receiver to seven more. The Jets really burned through the receivers this season.)
Note both of these charts end at 40%.
Now let's look at the tight ends and running backs:
Look at Detroit! Not only was Calvin Johnson the receiver with the highest number of targets in the league (199) but Stafford threw 200 more footballs to his tight ends. The Bionic Arm? As you can see, the AFC North is right around average, although as already noted Joe Flacco throws a lot to his tight ends.
Here are the running backs:
Every AFC North team except the Bengals was in the upper half of the league in terms of throwing to their running backs. (Cincinnati was barely above the league low.) But how they divided those targets varied immensely:
(Once again this is a percentage of targets, stopping at 80%.) The Steelers had the league low in percentage of targets to their No. 1 RB, and the league high in percentage of targets to the other RBs. Those of you who were paying attention will know very well why that is. For one thing, who was the No. 1 RB?
Well, in terms of total targets, that would be Jonathan Dwyer, with 25. But it's even more complicated than that. Dwyer was the No. 1 back in terms of targets during all 16 games, but if you take out the games Ben wasn't playing (and the portion of the KC game he quarterbacked) the No. 1 back was actually Isaac Redman. Although Pittsburgh was above average, league-wide, in number of targets to backs, those throws were split up between six different backs, including a significant number to Will Johnson, the fullback.
And finally (before the draft bit, that is) here is a ranking of the top four receivers for each team, regardless of position. Non-wideouts are noted:
As usual you can click on the chart for a bigger version if it is too hard to read. The upshot is, Ben Roethlisberger and Joe Flacco rely on their top four guys for close to three fourths of their passing offense. Andy Dalton appears to do the same, but in fact Marvin Jones and Jordan Cameron (TE) both had 31 targets, considerably lower than any of the other teams' No. 4 guy. So really they should be a top three, not a top four, in which case Andy Dalton would be dependent upon three players for 64% of his offense. I still say he needs to spread the ball around : ) Brandon Weeden appears to have the most balanced passing offense, with over a third of his offense coming from players outside of the top four, and almost half of his top four being non-wideouts. But perhaps this only speaks to his not having a really dangerous receiver. Greg Little would be it, except that he drops the ball too much. He has improved a great deal this season in that regard, however.
Now for my draft advice. First, Cincinnati.
Andy Dalton really does need a back-up plan for "throw it to A.J. Green." Houston took that option away in much of the wildcard game, stopped the run, and it was lights out. The team may not need to draft another receiver, though, as the Bengals lost a couple of promising receivers to injury, including last year's second-round pick Mohamed Sanu. They could also stand to send Jermaine Gresham a secret bottle of stick-um...
Next, Cleveland. Cleveland's receivers are an emerging group. The question is whether they are emerging quickly enough. It didn't help that Mohamed Massaquoi went down after playing well at the beginning of the year. Some of the young 'uns were stepping it up towards the end of the season, with Josh Gordon surprisingly overtaking Greg Little in number of targets. So they might be okay for receivers next season, depending on how well Massaquoi comes back. Really, they need to get more out of Little.
Joe Flacco appears to require a fairly high level of comfort with his receivers. He's got a nice group, although I would say the wideouts are one injury to Anquan Boldin away from being in trouble. Hey, they're in the Super Bowl, though, so they don't need any advice from me.
And finally, the Steelers. If the team chooses not to draft a tall strong receiver and/or tight end and/or stud running back with great hands, Ben is going to have problems if Heath's recovery takes as long as predicted. I always have great advice for the Steelers, which they always ignore, and they seem to do okay anyhow (although perhaps not so much this year. I don't know that my advice would have helped, however.) But it seems as plain as the nose on Ben's face. We have the most accurate quarterback in the division. Protect him adequately and give him some stud receivers to throw to, and next year things should be different.