If you've been around for a while you've likely discovered that I've got a weird fascination with the battle in the trenches. Specifically, I like to spend what free time I have watching how individual linemen do. I got into this mainly because I didn't want to just say so-and-so sucks or is terrific. I wanted data in the deep, nerdy way that Tom Clancy wants to know the ins and outs of various pieces of military hardware.

I had already done Jason Peters' season review before the end of the season. (He didn't play in the last two games.) Here's the link in case you have insomnia.

There has been some discussion about Duke Preston over the past couple of weeks.  The release of Derrick Dockery and subsequent realization the team has only two viable interior linemen (Brad Butler, Geoff Hangartner) helped these discussions along.  Specifically, people have been questioning whether he should be re-signed as depth now that Buffalo has picked up Hangartner to start at center. Short answer? "No." Long answer? "Hell no!" Actually, the long answer follows after the jump.

Ron's Grading Method
If you haven't trudged through any of my interminably long offensive line posts, I watch how each lineman does on every play. I then rate each lineman as having had a good, decent or bad play. I also note if the lineman's screw-up killed a play or a drive. (Every drive that didn't result in a TD was killed.) For the stats below, it may be useful to know that I score each good play as 95, decent play as 75 and bad play as 55. A good play is one in which a lineman does something remarkable, such as pancaking a defender or preventing a defender from elevating into the throwing lane. A decent play is one in which a lineman does his job - his defender doesn't disrupt the play, but the lineman doesn't really stand out either. Most plays fall in the decent category. A bad play is one in which the lineman screwed the pooch in some way. He might have allowed a defender to smack the QB or failed to block a guy who then either tackled or harried the RB. Killed plays are those in which the lineman's mistake caused the play to fail. To continue the above examples, the defender might sack the QB or tackle the RB in the backfield. Drive killing errors are those which literally cause the drive to stall, either with a kick (punt, FG) or turnover.

Preston while passing
Preston was on the field for 372 pass plays in 2008. Of those, he graded out as good on 2 (0.5%), decent on 344 (92.5%), bad on 26 (7%), and killed 9 (2.4%). That grades out to a poor 69.4%. Some have suggested that Preston showed progress towards the end of the year, which indicated he might grow into a decent center. This wasn't the case in the passing game. In the first 8 games (Preston had significant playing time in three - one of which was the Chargers game which wasn't broadcast in its entirety) he had no good plays, 86 decent ones, 7 bad ones and 4 killed plays for a grade of 73.1%. In games 9-12, Preston had 1 good play, 128 decent ones, 4 bad ones and 3 killed plays for a grade of 74.3%. In games 13-16 he had 1 good play, 130 decent ones, 7 bad ones and 2 killed plays for a grade of 73.6%.

That isn't growth, but rather consistency. Then I started to think about when people with the Bills were talking about Preston's improvement. I recall, perhaps wrongly, that it was during the last couple of weeks of the season. Looking at the last quarter of the season in greater depth, Preston had a terrible stretch (passes 8-13 and 15 were bad) in the passing game against the Patriots. Dropping out that week and looking at the Dolphins-Jets-Broncos stretch, Preston had 1 good play, 104 decent ones, 4 bad ones and 2 killed plays for a grade of 74.5%. That would put Preston's trajectory in the pass game as 73.1% to 74.3% to 74.5%, something that would indeed show some promise. The egg he laid in the second Patriots game, however, was smelly enough to make that upward trajectory wilt and dip back down: 73.1% to 74.3% (back down) to 73.6%. So, yes, Preston was improving in his pass blocking and then Vince Wilfork came to town.  

Preston while running
The run game wasn't any prettier for Preston. He was on the field for 269 televised run plays. He was rated as good on 24 (8.9%), decent on 148 (55.0%), bad on an astounding 97 (36.1%) and killed 19 (7.1%) for a dismal grade of 69.6%. Ugh. That's actually worse than I thought it would be. Think about that for a moment. Preston was beaten on more than one third of all running plays. He personally killed about one out of every 14 run plays.

Again, let's take a look at how it breaks down by quarter. The first 8 games (Preston saw significant action in 3 of the first 8 games) saw Preston have 5 good run plays, 39 decent ones, 7 bad ones and 7 killed plays for a grade of 70.3%. Weeks 9-12, Preston had 8 good plays, 58 decent ones, 10 bad ones and 8 killed plays for a grade of 68.7%. The last quarter Preston had 11 good run plays, 51 decent ones, 35 bad ones and 4 killed plays for a grade of 70.1%.

So, yes, Preston had a slight rebound (70.3% to 68.7% to 70.1%) in the last quarter. Once again, we're not talking growth so much as a mild swelling. It's interesting, to me anyway, to drop the Patriots game from the mix. Doing so, Preston had 7 good plays, 37 decent ones, 22 bad ones and 3 killed plays for a grade of 70.5%. It's not exactly a great leap to get from 68.7% to 70.5%, but it's almost a half percent better than when the Patriots game is included. Throw in Preston's idiotic decision to get into an altercation at the end of the first half - instead of, you know, snapping the ball to kill the clock or for a field goal - and he quite simply played himself out of a job against the Patriots.

Preston against the 4-3 vs Preston against the 3-4
There is one more aspect to Preston's run game issues to beat into the ground. Some have said that he's a good center when faced with a 4-3 defense and only struggles with 3-4 nose tackles. There's a grain of truth to this theory. Against 4-3 defenses (Preston played only 3) he had 3 good plays, 42 decent ones, 11 bad ones and 1 killed play for a grade of 72.1%. Against 3-4 defenses he had 21 good plays, 106 decent ones, 86 (!) bad ones and 18 killed plays for a grade of 68.9%.

Preston was better at run blocking against 4-3 defenses than 3-4 (72.1% to 68.9%), but that's not nearly good enough against either to play in the NFL. Besides, the AFC East is home to three 3-4 defenses, so there isn't any getting away from them. There's also no point whatsoever in bringing Preston back as depth. After all, if a guy has proven that he's incapable of starting, what's the point of having him on the roster? It would be one thing if Preston was an unknown quantity like Demetrius Bell or some other rookie. Preston has demonstrated that his ceiling is the roof of the basement. To move out of the division basement, the Bills need to pass on players who dwell there in perpetuity.