Mid-Season Report Part II: The Rest

I want to preface this post with a little discussion about QB Support and this new value added metric. The value added metric is designed to measure how much credit or blame a QB deserves for his team’s record. Before you take this as some bizarre claim that Alex Smith is better than Tom Brady, here are two points to consider:

1. This is ONLY for the first half of 2012 (Weeks 1 – 9). To the extent that the data tells us anything, it only tells us about those 8 or 9 games for a QB. It does not take into account twelve years of a hall of fame career or 3 or 4 seasons of mediocrity.

2. A QB’s value added metric is most affected when a QB (1) leads his team to victory despite low support or (2) fails to deliver a win in spite of high support. A lot of QB play gets overlooked. For example, when a team is getting crushed, a QB may have a strong or weak game and lose anyway. In those circumstances, the value added metric is very forgiving to QBs who play poorly in games that were all but lost regardless. Similarly, it gives little credit to the QB who plays very well as his team cruises to a 24-3 (Alex Smith vs. the Cardinals) or 45-7 (Tom Brady vs. the Rams) win. The idea is to measure how many wins a QB adds or "how good a QB is at winning." It's great to play well in a blowout, but you're not actually adding any wins if your backup could have likely won the game too. 

With all that said, let’s move on to “the rest.”

III. Good Team + Good QB = Good Record

This next group includes QBs who are receiving high support numbers and doing an adequate to above average job in turning that support into wins. For QBs like Jay Cutler and Matt Schaub, it’s hard to say that their teams record is because of who’s standing behind center. Nevertheless, the QBs in this group have at least not squandered the support they received, and, in most cases, added their own solid play to help their teams to a winning record.

TEAM	QB	AVG. SUPP.	Expected Wins	Actual Wins	VALUE ADDED
STEELERS	B.Roethlisberger	3.26	4.38	5	0.62
TEXANS	M. Schaub	4.58	6.39	7	0.61
BRONCOS	P. Manning	3.19	4.44	5	0.56
49ERS	A. Smith	4.81	5.53	6	0.47
GIANTS	E. Manning	3.90	5.62	6	0.38
BEARS	J. Cutler	5.50	7.00	7	0
PATRIOTS	T. Brady	3.82	5.08	5	-0.08

HIGHLIGHT ON JAY CUTLER AND TOM BRADY

Jay Cutler is the most supported QB in the NFL this year. Cutler was 1-1 in his first two games after receiving somewhat average support. In both of those games, Cutler had QB Support levels where NFL QBs, on average, win 57 percent of the time. Since then, Cutler has played in six straight games with a QB Support level above five. QBs who received that level of support in 2012 have won 44 out of 45 games (Sorry Matt Cassell).  In those six games, Jay Cutler has only had to overcome an average of 3.6 points per game, and never more than 16 (accounting for defensive and special teams TDs). His solid play in some of those games helped turn a “just a win” into a blowout, but I don’t think he deserves the bulk of the credit for the Bears 7-1 start.

Tom Brady is the only QB in this group whose value added is below 0 (though just barely). Regardless of how excellent Brady has played in his career, he’s lost some winnable games this season and, given the small sample size of only 8 games, it’s enough to drag his value added number down below 0. Despite having a QB Support level of 3.9 against the Cardinals, Brady could not capitalize. Yes, his kicker missed a makeable field goal at the end of the game, but (1) it should not have come down to that and (2) his kicker also made field goals from 46, 51, and 53 yards in that game (as well as 34 yard chip shot). No QB can expect his kicker to be perfect. Brady’s QBR of 30 was a big reason for that loss. Brady’s numbers are also lowered by the loss against the Seahawks (QB Support = 2.305), but not as much. Additionally, Brady has played some of his best games when his team is already rolling over the opponent (for example, the blowouts against the Titans and the Rams). Brady hasn’t quite been what people have come to expect of him. Still, it’s been enough to keep the Patriots at a winning record, and just about where one would expect them to be.

IV. AVERAGE ALL AROUND

This group includes QBs who have performed right at average (or just barely above) while their teams have given them just about average to below average support. Their records may be average or slightly worse, but that’s because their QBs have played about average or slightly worse… oh yea, and the entire rest of their team too.

TEAM	QB	AVG. SUPP.	Expected Wins	Actual Wins	VALUE ADDED
BILLS	R. Fitzpatrick	2.24	2.9	3	0.1
CARDINALS	K. Kolb	3.80	3.93	4	0.07
EAGLES	M. Vick	2.29	3.00	3	0
RAIDERS	C. Palmer	1.99	3	3	0
RAMS	S. Bradford	2.52	3.16	3	-0.16

HIGHLIGHT ON CARSON PALMER

Carson Palmer is interesting in that his average support is only 1.99 (30^th out of 32 rank\ed QBs) and his expected wins are 3.00 (tied for 24^th of 32 ranked QBs). The reason that the two do not have a perfect correlation is Palmer has played in a number of games with extremely low QB Support. In his first eight games this year, he has had three games where his QB support was below 0. When QB Support is below 0.5, all QBs  have a winning percentage of 0 % this year. Therefore, a game where a QB’s QB Support level is 0.300 is treated the same (as an automatic loss) as one with -0.800, even if the two games might average differently with a second game where there is high support. Palmer’s “expected wins” then represent a better  representation of where his team should be than a simple average of his QB Support.

V. PLEASE SEND HELP!

These two QBs are on teams with struggling records, and they also have value added metric’s that are slightly below average. Sure, maybe these QBs could be playing better, but they aren’t getting much help from their teams. If these teams want a winning record, they’re going to have to play better from top to bottom. There’s a lot more than QB play that’s sinking these two teams.

TEAM	QB	AVG. SUPP.	Expected Wins	Actual Wins	VALUE ADDED
TITANS	M. Hasselback	1.92	2.15	2	-0.15
JAGUARS	B. Gabbert	2.09	2.32	2	-0.32

HIGHLIGHT ON MATT HASSELBACK

Hasselback is the second least supported QB (only Drew Brees has received less), and his expected wins are actually even slightly lower than Brees’. Compare Hasselback with the league’s most supported QB, Jay Cutler. Cutler’s played two games with QB Support between 2.500 and 3.999, and he’s gone 1-1. Hasselback has played in three such games, and he went 2-1. One big difference is that where Cutler’s enjoyed QB Support levels above 5 in each of his six additional games (averaging at 6.23), Hasselback never saw QB Support above 2 in his three addtional games (averaging at 0.775). That's at least part of the reason why, at the end of week 9, Hasselback was 2-4 as a starter and Cutler was 7-1.