There have been quite a few stories on where the various quarterbacks rank in the NFL, based on pundit opinions, coaches' opinions, etc. However, what we haven't really seen is a data-driven look at the NFL's signal-callers. I'm a big believer in the expected points model, so I decided to take a look at QB performance based on various ESPN Expected Points Added (EPA) sub-components -- passing EPA, sacks EPA, total EPA, etc. I also had to do a little bit of work to combine pre and post season data, which ESPN doesn't do for the public (boo!).

The first one I'll throw out there is raw QB production (includes running and penalties) per game since 2012, including playoffs. For Kap, I used 7.5 games for the 2012 regular season:

This list generally matches your perceptions of the game's QBs...almost. Kaepernick is 7th, which remarkably high given that the 49ers have been a run-first team since 2011 -- this is aggregate production, and the opportunities simply aren't there as much as they are for other QBs in pass-heavy offenses. In addition, the 49ers have a very good/great defense, which reduces his production even more.

So what about on a per-play basis?

On a per play basis, Rodgers and Manning are significantly better than every other quarterback in the NFL. But after that, we've got Brady as a pretty clear 3rd, then a muddle of 5 to 6 quarterbacks in tier 2 -- including Kaepernick.

Now, what about just passing -- after all, pundits claim he's a run-first quarterback, right?

We see some other young QBs here that don't get credit from the media either, like Tannehill. Kaepernick is t-9th with Stafford. Pretty incredible that Kaepernick produces as much with this arm -- in aggregate as Matt Stafford does, as much as he throws. Here are the per-attempt figures:

Kap is 5th in passing efficiency, a shade behind Wilson (surprisingly) - we see guys like Stafford fall off the list here and down to 19th when measured this way.

Now, how about sacks? Kap gets sacked a lot right? Well...yes. So let's take a look at that metric.

In the aggregate, Kap falls all the way down to 12th or 13th, Wilson falls to 15th, Luck drops to 17th. Lots of quarterbacks are having sack problems. On a per ATT+SACK basis:

Luck falls to 20th. Cam Newton is a putrid 27th (!). Kap and Wilson have nearly identical figures in this metric and are t-10th. And once again Manning and Rodgers are heads and shoulders above the rest of the NFL.

So what does all this mean? Well, unless you think Kaepernick dramatically regressed in 2014, it means that we have a Top 5 to Top 12 quarterback who is still figuring things out. I think the 'regression' argument is bunk. While his INT% went up, his completion percentage also went up. And while his yards per attempt went down, the very poor play of the offensive line prevented him from having enough time to get down the field. Finally, weak receiver play -- evident by the investment in Torrey Smith and departure of Crabtree -- prevented a lot of great opportunities -- 6th in team drop rate, lack of separation, Vernon Davis injuries, etc. There are great reasons to remain optimistic! Over a 45 game sample, Kaepernick looks like a wonderful quarterback.