Bills "Game Film," Part I: The Defense

The 2010 NFL Draft is over, and a good portion of Buffalo Bills nation is drinking Buddy Nix’s Southern-style, cranberry-flavored Kool-Aid, embracing - or at least accepting - the fruits of the off-season. However, some fans remain confused and upset that the Bills have failed to address widely perceived needs at left tackle, quarterback, and so on.

Part of the answer as to why the Bills made the off-season decisions they have could be how the new coaching staff has evaluated the players currently on the roster. For instance, we’ve heard more than a few times that our front office felt our current crop of quarterbacks are just as good as any QB in the 2010 NFL Draft. I’d imagine the incoming staff has reviewed last year’s game film in-depth and given grades for every player’s performance.

Now, wouldn’t it be fun to get a hold of the grades the coaching staff assigned to a player? Sadly, that’s like my chances of flipping through television channels and stumbling onto "Charmed" - it ain’t happening. Instead, we’re going to have to ride with the next best thing - ProFootballFocus.com (hereinafter "PFF").

PFF "analyzes and grades every player on every play in every game to provide you with the most in-depth statistics you can find anywhere outside the team's film room." After the jump, we’ll dive into PFF’s grades for last season’s Bills defense. (Note:  For the faint of heart, there may be some ugliness ahead.)

For anyone not familiar with PFF, I’ve included some excerpts from an interview with a member of PFF through SB Nation's Blogging the Boys, who did a similar study. Basically, PFF charts games on television and grade it like it's game film. Follow the links provided for more information, because for the purpose of this post, I’m assuming that the PFF data is true - even if it makes certain Bills players look like they need to be given the Old Yeller treatment.

PFF has a unique way of grading players. They look at game tape, assign a grade for every play and then ‘normalize’ the data so that the average player for a given position is graded at zero. The higher the positive grading the better the performance and vice versa. In their own words:

"The grading takes into account many things and effectively brings "intelligence" to raw statistics. For example a raw stat might tell you a Tackle conceded a sack. However, how long did he protect the QB for before he gave it up? Additionally when did he give it up? If it was within the last two minutes on a potentially game tying drive it may be rather more important than when his team is running out the clock in a 30 point blow out." From the PFF Q&A.

 

"In our opinion [our approach] is at least a step change above what anyone could get by simply "tracking results". We can give you two examples:

1. A cornerback is beaten badly on a post route and the wide receiver drops the ball in the end zone. Tracking the result gives this as an incompletion against the cornerback (a positive) whilst we will mark this down as a significant negative.

2. A QB throws a perfect strike over a linebacker to hit his open slot receiver on an out. Once more, the wide receiver drops the pass and it cannons off his chest to a Safety who catches the ball even though he'd initially made a bad job of the coverage. The QB gets a INT listed against his name and the free safety gets an undeserved INT against his. How is tracking this result more accurate?"

Clearly there are limitations as to what is shown on TV. The biggest issue is that of not being able to see downfield coverages on untargeted defenders and we accept this as an inherent error in what we do. That said, nobody outside the teams has access to this either so should we stop our "more accurate" analysis on the basis that it's not 100 percent perfect?" From an interview with Sam Monson at PFF.

PFF allocates each player an NFL rank and a "score." Therefore, we can do a comparison of how each Bills player performed last season and compare them to their contemporaries at the same position. I’ve included the players' penalty scores, but personally, I believe that penalties are weighted too greatly in the overall score. Additionally, there’s a "percentile rank," to demonstrate where that player stood relative to other NFL players at his position (higher is better). 

Green is a positive number, yellow is roughly an average grade, and red is negative (also one of the Bills’ primary colors and something we’ll see a lot of). The higher the number, the better that player performed; alternately, the lower the number, the worse they were last year according to PFF’s grading system.

Defensive Backs:

Position last year	Player	NFL Rank	# of total players	% Rank	Overall	Pass Rush	Cover	Run Defense	Penalties	NFL QB Rating Against
CB	Terrence McGee	75	107	30%	-6.4	0.0	-3.2	-2.2	-1.0	88.0
CB	Drayton Florence	79	107	26%	-6.8	0.0	0.7	-2.5	-5.0	74.2
CB	Reggie Corner	86	107	20%	-8.9	0.6	-3.7	-2.3	-3.5	83.7
S	George Wilson	4	85	95%	11.4	3.1	4.6	3.7	0.0	20.5
S	Jairus Byrd	37	85	56%	-1.3	0.0	4.9	-4.2	-2.0	37.5
S	Donte Whitner	65	85	24%	-6.0	-0.7	1.8	-4.1	-3.0	58.1
S/OLB	Bryan Scott	19	53	64%	3.6	1.3	6.9	-4.6	0.0	93.9

I expected to see something akin to a Christmas tree from the Bills’ defensive backs - lots of green with a mix of yellow and sprinkles of red. However, that’s a lot of red, especially for the area of the team most people think is our strength. I was pretty shocked to see PFF think our best overall cornerback was Terrence McGee. Drayton Florence was actually better in coverage (the only positive Bills cornerback), but his grade gets crushed by the five penalties he had (see what I mean about the grading system being overly harsh on guys who were penalized?). Leodis McKelvin didn’t have enough snaps to qualify. According to PFF, Bills cornerbacks as a whole didn’t perform too well last season.

PFF rated George Wilson as the fourth best safety in the entire league last year. Wilson’s QB Rating Against is 20.5, which was the lowest in the league for a safety who played a large number of snaps. PFF dings Jairus Byrd for his clear weakness - run support - and he also ends up positive if you take away his penalties. Mr. Polarizing, Donte Whitner, was graded below average with a poor run defense figure. However, his coverage grade and his QB Rating Against were excellent, so there’s room for improvement in his grade if his run defense improves. PFF counts Bryan Scott as an OLB for their rankings; he would have been the No. 12-ranked safety had they included him there. Obviously playing outside linebacker last year for the Bills flattened his run defense grade.

One of the quibbles I have with PFF’s system is how it takes interceptions into account. In watching the games, Byrd basically won two games by himself, and I don’t believe PFF properly grades the impact of interceptions. For those wondering which CBs got the top PFF grades, it was far and away Charles Woodson with a 29.2 and Darrelle Revis with a 27.4; Revis had an absurd 32.3 QB Rating Against. Woodson had a Run Defense grade of 15.0, which vaulted him slightly ahead of Revis, and his 31.0 Coverage grade.

Linebackers:

Position last year	Player	NFL Rank	# of total players	% Rank	Overall	Pass Rush	Cover	Run Defense	Penalties	NFL QB Rating Against
ILB	Paul Posluszny	6	54	89%	15.0	2.9	6.5	5.6	0.0	47.4
ILB	Andra Davis (Den)	5	54	91%	16.3	4.8	-0.5	13.0	-1.0	103.2
OLB	Kawika Mitchell	45	53	15%	-8.8	-1.3	1.6	-7.6	-1.5	108.5
OLB	Keith Ellison	48	53	9.5%	-10.1	-2.4	-4.3	-3.4	0.0	98.1
4-3 DE	Aaron Schobel	30	73	59%	2.3	2.5	1.5	1.3	-3.0	N/A
4-3 DE	Chris Kelsay	71	73	3%	-20.3	-10.4	2.0	-11.9	0.0	N/A

For the linebackers, I chose to include the players who will be playing linebacker in our 3-4, while listing their positions last year. New addition Andra Davis had the fifth-best grade according to PFF, and Paul Posluszny had the sixth-best grade among inside linebackers. You can also see that PFF graded Davis as struggling in coverage, which is generally considered his weakness. If those grades continue into this year, we could have a very formidable inside linebacker duo. Posluszny also had the lowest QB Rating Against among all inside linebackers, with a 47.4. Provided the Bills graded Poz even remotely near the figure PFF gave him, I’d expect to see long-term extension talks soon. Kawika Mitchell struggled mightily in the Tampa Bay and New Orleans games; he was average in the other three games he played before being placed on Injured Reserve. 

Keith Ellison graded really well in exactly one game - the Cleveland game (the entire defense had good grades for that game, probably because Derek Anderson played as well as I could have). If Ellison’s grade is accurate, he’ll have a tough time making the team. Aaron Schobel was inconsistent; he’d have a top-notch game (e.g. a three-sack performance in Atlanta) followed by a miserable game. Kurupt will be pleased to note that Chris Kelsay graded out as the third-worst 4-3 defensive end in the league. However, Kelsay did get a positive grade in coverage, so perhaps he has better skills in space than previously thought, and can transition to OLB successfully. Recently acquired LB Reggie Torbor finished with a -5.6 overall rating, slightly below average among players with a qualifying number of snaps. 

Defensive Line:

Position	Player	NFL Rank	# of total players	% Rank	Overall	Pass Rush	Cover	Run Defense	Penalties
DT	Kyle Williams	13	87	85%	7.3	2.6	0.0	4.7	0.0
DT	Spencer Johnson	55	87	37%	-6.5	-4.1	1.0	-1.9	-1.5
4-3 DE	Ryan Denney	66	73	9.5%	-11.6	-3.8	-1.0	-5.8	-1.0
DT	Marcus Stroud	84	87	3.5%	-20.8	0.9	1.5	-21.2	-2.0
3-4 DE	Dwan Edwards (Bal)	18	73	75%	6.2	-7.9	1.5	13.6	-1.0

We all knew Kyle Williams had a good year; it’s comforting to see that confirmed. He registered a higher grade in run defense than I expected to see. You can see why the Bills let Ryan Denney walk - negative numbers all around there. The big shock here, at least to me, was what our defensive end situation would be without signing Dwan Edwards. I know Marcus Stroud had a bunch of nagging injuries last season and didn’t play well, but he was Kelsay-level bad. In fact, Stroud actually earned a lower PFF grade than Kelsay did. Stroud’s run defense in particular flabbergasted me. Stroud posted a +14 overall grade in 2008, so it’s entirely possible last year was a fluke because of various injuries. There’s also the very real chance that he’s reached the decline phase of his career.

If the Bills graded their game film even remotely similar to this, it’s easy to see why the Edwards and Alex Carrington additions occurred. Perhaps defensive end was just as big of a hole as QB and LT? Edwards undoubtedly excelled at run defense, but was a giant black hole when rushing the passer. He’s here primarily to stop the run, which clearly plays to his strengths.

Once again, if the Bills game film grades are similar to PFF’s grades, the defensive line got a horrible grade last year except for Williams; the off-season spent bolstering the defensive line makes a whole lot more sense in this case, especially with the off-season switch to the 3-4 defense.

Are these evaluations a correct indication of how the Bills may have graded their players? Their off-season acquisitions and decisions so far could indicate that they are, especially the focus on the defensive line. In Part II, we'll go through the offense (expect more red).