In which a whole bunch of meaningless statistics (thanks, PaVa!) are used to come to some rather obvious conclusions...
Although the quarterback for the Baltimore Ravens, Joe Flacco, is still part of the post-season, for my purposes this doesn't matter, because I'm going to confine my analysis to the regular season and compare my pre-season prediction to the actual outcome of the season.
So let's begin with said prediction. Before I made it, I put up a great many charts comparing Joe Flacco, Ben Roelisberger, and Andy Dalton in their rookie season, and comparing Ben Roethlisberger's first four years with those of Joe Flacco. I threw in some college numbers for Brandon Weeden, just to be thorough. This was the conclusion I reached:
I think Ben Roethlisberger will stay reasonably healthy behind his new improved offensive line and put up excellent numbers, possibly even as good as his 2007 or 2009 seasons. I think Joe Flacco will play well, especially with the new confidence he gained by helping get his team to the playoffs and almost to the Super Bowl. But I also don't see his upside to be as high as the other QBs in the AFC North. I believe Andy Dalton will continue to show the Bengals got a steal in Round Two of the 2011 draft, and he will improve significantly, just as he did between his first and second year of college. This is partly because of his history and partly because he has more than just A. J. Green to throw to this season. Brandon Weeden is a big question mark, and will probably struggle. Not necessarily because he isn't up to the job, but because the Browns haven't addressed the WR issue. Since he only managed a single rushing TD in his entire college career, and since he's an old guy, relatively speaking, he's probably not going to be able to pull a Tebow and just take the ball in himself on a regular basis.
How close did I get? Well, let's look at the numbers. First off, here is what you might call the bottom line. I'm giving both the average NFL rating and the Pro Football Focus quarterback rating:
Ben Roethlisberger did stay relatively healthy, at least at first, until his new, improved offensive line became subject to the same frequent injuries as the old, unimproved model. (Ironically, the only offensive lineman to play every single snap was afterthought Max Starks.) Roethlisberger had an average of almost 15 attempts per sack this season, which maybe doesn't sound great, but it is notably better than any other season in his career. And despite the unfortunate back-alley encounter with a couple of gentlemen from the Kansas City Chiefs which kept him out for three games and reduced his effectiveness when he returned, he put up quite good numbers. As you can see from the charts above, whether you use the NFL quarterback rating or Pro Football Focus's quarterback rating (which eliminates spiked balls, dropped passes, and suchlike,) Roethlisberger still put up the best numbers in the division.
Next in seniority is Joe Flacco. I thought he would play well this season, and he did—at home. But more of this later. Despite a few stinkers of games, he ended up with a respectable 87.7 average rating for the season. While this is considerably improved from his 2011 figure of 80.9 (only marginally better than the 80.3 of his rookie season) he still hasn't again matched his high of 93.6 in 2010. (Of course, Ben hasn't matched his high of 104.1 in 2007 again, either.) I am watching the playoffs with interest, and think perhaps my comment about Flacco's upside was premature.
Andy Dalton did indeed improve considerably, just judging by the 87.4 quarterback rating, compared to the 80.4 he averaged in his rookie season. Furthermore, he got his team to the playoffs for the second year in a row, although, like last season, they were one and done in the playoffs. However, although the Bengals did pick up a few more receivers, they dropped like flies throughout the season, other than Andrew Hawkins. In the end the offense is still too dependent upon the awesomeness that is A.J. Green, in my opinion.
Finally, I think it is fair to say Brandon Weeden struggled. Towards the end of the season his receiver corps began to look a bit less like an episode of "Keystone Cops" and a bit more like an NFL unit, and this definitely helped. Furthermore, he was not only the least-sacked quarterback in the AFC North, by a good bit, but he was only sacked five times during the latter half of the season, giving him a total of 25 sacks. (The most-sacked quarterback was, surprisingly, Andy Dalton, with 46—almost twice as many as Weeden.) He did steadily improve during the season, though. This might be more meaningful in terms of next season if anyone had any idea what is likely to happen in Cleveland. He might be released by the new coach, or groomed as the centerpiece of yet another offensive scheme, or some other currently unimagined scenario. Watch this space...
Here's a chart comparing some of the salient numbers for the quarterbacks. I have calculated everything so that higher is better (and thus I used touchdowns divided by attempts, rather than the more natural attempts divided by touchdowns, because in the latter case a lower number would be better. When necessary I have multiplied or divided the resulting number to put it into a meaningful range.)
I'm sorry it is so small, but if you click on the chart it will open in a new window and be actual readable by human eyes. Here is the legend, rather larger:
Now to return to a comment I made above, stating Joe Flacco played well this season, at home, with the obvious implication that things were otherwise away from home. I noted this myself partway through the season, and recalled a comment, read I know not where, that Joe Flacco was a different quarterback at home than away. I ran some numbers at the time and the difference was indeed striking at that point in the season. I figured I might just as well run the numbers for all of the AFC North QBs while I was at it. Here's what I came up with for the season as a whole:
The error bars show the standard deviation. In other words, if they overlap the difference probably isn't meaningful. Note that two of the quarterbacks don't show a meaningful difference in their home and away average, Roethlisberger and Dalton. However, both Flacco and Weeden show a distinct difference in their rating when playing at home or away. Curiously, though, Weeden is distinctly better away than at home, or at least he was this past season. Joe Flacco, however, does definitely seem to play better at home than away. I decided it was worth looking at the Pro Football Focus rating to see whether the same pattern showed up, and here's what I found:
For whatever reason, although the differences are still evident, according to the PFF ratings they fall within the margin of error.
But this didn't satisfy me, and so I decided to go back to 2008, year by year, and see whether we can spot a tendency in one or more of our quarterbacks. (I went back to 2008 because that is as far as the PFF rating goes, and I wanted to compare both.) So here are the charts:
Curiously, we note Ben Roethlisberger was the one who was way better at home in 2011. Also curiously, Andy Dalton was better away than at home in his rookie year, although not by a great deal. Joe Flacco was better at home, but also not by a great deal. And we can see that the differences are greatly minimized by the PFF ratings, just as for this season:
Once again Joe Flacco is better at home than away, but within the margin of error. And in the PFF ratings, his home average is actually slightly worse than his away average:
This season the difference is definitely statistically significant between Joe Flacco at home and, as it were, abroad. The difference is still marked, if just within the margin of error, with the PFF ratings.
And here comes the kicker:
In his rookie season, just like Brandon Weeden, and to a lesser extent Andy Dalton, Joe Flacco was worse, by a long way, at home than he was away. And as you can see, in the PFF ratings it's back within the margin of error.
Finally, I decided to do a career average, home and away, dating back to 2008. Here's what it looks like:
Interpret it how you like, but it looks to me like the difference between Joe Flacco at home and away are not large enough over the course of his career to be particularly significant. As so often happens when looking at football stats, the combination of the small set of data and the team nature of the sport makes it difficult to determine such things with any degree of assurance. It makes cool-looking graphs, though, so I'm happy.
And so, after collecting a ton of data over the course of the season, I find my gut-level assessment in July was pretty good. At least so far. We've got a lot of positions to get through. It's a long off-season, though, although longer for some of us than others...