QB Support: Why Manning isn't really a Choke-Artist

Proposing a New Statistic in A QB Centric World

When discussing the NFL, fans and analysts love to talk about a Quarterback's winning percentage, especially in the playoffs. And so we count how many of their fingers are decorated with Superbowl Rings and talk about which QBs "know how to win." What about the rest of the team? Do they make a difference?

Well, today I am proposing a new “aggregate statistic” somewhat akin to QB rating. Instead of adding up different measures of the QB, why not look at the measures of the rest of the team. The better defense and the ground game, the more "support" for the QB. Think about it this way If your team is running the ball for 200 yards and holding the other team to 10 points while creating 3 turnovers, it doesn’t take an all-pro to lead you to victory.

"QB Support" takes into how account who wins the rushing battle, points allowed by the defense, turnovers generated by the defense, and also defensive and special teams TDs. I include exactly how I calculate "QB Support" at the end of this post (for those interested in the numbers). For now though, let's look at three pro-bowl caliber QBs, how they've fared in the playoffs, and what light, if any, "QB support" can shed.

BEN ROETHLISBERGER, TOM BRADY, AND PEYTON MANNING: A CASE STUDY

If you look at playoff wins and losses, Roethlisberger is 10-3 (77% wins) with two rings, Brady is 14-5 (74 % wins) with three rings, and Manning is 9-10 (47 % wins) and 1 ring. Just for the sake of this argument, let’s momentarily exclude 2006 for Manning (The year he won the SB). That would put him at a dismal 5-10 (33 % wins) and no rings.

The win-loss records suggest that Manning just plays horrible during the playoffs, which has been possibly backed up in people’s mind by two ineffective playoff games against the then stingier Patriots defense in 2003 and 2004. Still, Manning hasn’t played comparatively poor in the playoffs. He has, however, continually received low levels of "QB support" in his non-Superbowl years. Let's look at QB rating in the playoffs, win-loss records, and the average support levels received by all three.

	Playoff QB RATING	Playoff WIN-LOSS Record	Average Support Level
Roethlisberger	85.7	10-3	2.90
Brady	85.4	14-5	2.96
Manning (excluding SB year)	93.0	5-10	0.58

Averages are sometimes not as informative though, so let’s look at how the support level affects individual games. Specifically, let’s look at games where support was above 1.5, where it was between 0 and 1.5, and where it was below 0.

WIN-LOSS RECORDS ACROSS DIFFERENT SUPPORT LEVELS
	Above 1.5	Between 0 and 1.5	Below 0
Roethlisberger	10-0	0-1	0-2
Brady	13-0	1-2	0-3
Manning (excluding SB year)	3-0	1-2	1-8

Now, let’s take a look at some averages of the individual statistics that lead to support levels.

A LOOK AT THE INDIVIDUAL STATISTICS
	Rush Yards for	Rush Yards Against	Turnovers generated/gm	Def or SP TDs/gm	Pts allowed
Roethlisberger	116.17	78.69	1.77	0.31	21.17
Brady	108.74	100.42	1.89	0.32	19.26
Manning (excluding SB Year)	77.87	142.867	1.53	0.00	22.20

LOOKING DEEPER: MANNING vs. MANNING and BRADY vs. BRADY

Throughout this exercise, we’ve been excluding Manning’s superbowl year to make a point that all those playoff losses might have more to do with factors Manning does not directly control than with his poor play. And yet, he won a superbowl anyway. It should be even more impressive considering his typically low levels of support. Right? Meanwhile, Brady took full advantage of high support levels to start his career by going 10-0 in the playoffs. Since then, he’s gone 4-5. He must have lost his magic, right?

Wrong on both counts. I said earlier that “averages” aren’t that useful. They can be, but the more you can break it down, the better. This is why I took out Manning’s superbowl year. Let’s take a quick look at that year for a comparison.

	QB Rating	Average Support level	WIN-LOSS Record
Manning (excluding SB year)	93.0	0.58	5-10
Manning’s SB Year	70.55	4.82	4-0

The truth is, Manning was not playing his best playoff football that year. It wasn’t that he finally got over the hump. It was that his team did. A two-headed ground game featuring Rhodes and Addai and an opportunistic defense that came alive in the playoffs was what the Colts were missing years past. It wasn’t poor QB play that was holding them back. I want to break down some of the individual statistics because I think they’re instructive.

	Rush Yards for	Rush Yards Against	Turnovers generated/gm	Def or SP TDs/gm	Pts allowed
Manning (excluding SB Year)	77.87	142.87	1.53	0.00	22.20
Manning’s SB year	151.00	83.25	2.75	0.00	17.00

I think this also makes the case that these statistics are not merely “products” of good QB play. Manning had been playing well in the playoffs for years, but with no ground game and a weaker defense than other strong quarterbacks had, his team struggled to generate those January wins. Without any significant improvement in his personal play (in fact, it seems he played worse), his team gelled as an entire unit. With high support levels in 2006, he finally got “his” ring.

Now, let’s take a look at Tom Brady. What’s happened to his once spotless playoff record? Well, the numbers suggest, that at least on some level, it wasn’t “his” playoff record, but the record of a team that is no longer together.

	QB Rating	Average Support level	WIN-LOSS Record
Brady’s First 10	91.0	4.10	10-0
Brady’s Next 9	80.6	1.66	4-5

Brady did have a higher QB rating that corresponded with the winning (I’ll address that shortly), but the drop-off in support levels seem more dramatic. Same as we did with Manning, let’s look at the individual statistics:

	Rush Yards for	Rush Yards Against	Turnovers generated/gm	Def or SP TDs/gm	Pts allowed
Brady’s First 10	116.9	84.0	2.2	0.4	15.8
Brady’s Next 9	99.67	118.67	1.56	0.22	23.11

You can see that they were not rushing as effectively, defending the run as effectively, making as many big plays on defense OR keeping the point total down. It’s significantly easier for a QB to win when opponents put up 15 points than when they put up 23. Now, there was also a drop-off in Brady’s QB rating since his first ten wins. Let’s look deeper into that.

	Yds/Att	Comp. %	% TDs	% Ints	Atts/game
Brady’s First 10	6.50	62	4.22 %	0.90 %	33.1
Brady’s Next 9	6.42	62	4.56 %	3.70 %	39.0

So Brady completed the same percentage of his passes, threw for TDs more often and also kept his yards/attempt pretty steady. The big difference is that he had to throw more often, and he threw picks at a much higher rate. I would argue that in years past, with the defense holding teams to lower point totals while the Patriots were outrushing opponents, Brady was less often forced to make something happen. I would argue it’s less about him becoming ineffective and more about him finding himself in difficult situations.

Final Thoughts

I think the “QB Support Level” can be a useful metric, particularly when examining individual games. Brady, Roethlisberger, and Manning are all undefeated when they have a support level above 1.5. Likewise, all of them have abysmal records when the support level is below 0. Do some Quarterbacks “just win” when it comes to crunch time, or do some QBs get more support from the rest of the 53 man roster?

APPENDIX I: CALCULATING "QB SUPPORT."

RUSHING YARDS

Rushing Yards for :.015 points for each rushing yard a QB’s team generates (.75 for 50 yards)

Rushing Yards against :-.015 points for each rushing yard a QB’s team allows (-.75 for 50 yards)Essentially, a QB gets +/- .75 points for every 50 yard disadvantage/advantage in the ground game

POINTS ALLOWED

9 points or less = 3 points

10 to 16 points = 2 points

17 to 21 points = 1 point

22 to 30 points = 0 points

Over 30 points = - 1 point

*In calculating points, you subtract 7 points for any INT returned for a TD. Those TDs fall on the shoulders of the QB and shouldn’t be calculated in how much “support” they are getting from the overall team.

DEFENSE OR SPECIAL TEAMS TDs

1.5 points for each TD

TURNOVERS GENERATED

.75 for each turnover your defense creates.