Filed under:

# Pro-Football-Reference Measures The Support QBs Receive In Playoff Games

I can't recommend more the great work the guys at Pro-Football-Reference do.  They have tons and tons of data throughout the NFL's history, and they put it to good use.  It's one of those "must read" sites on the Internet, especially if you love numbers and football.  These are the same guys who host our favorite stat here at Stampede Blue: Adjusted Net Passing Yards / Attempt.  Anything that can measure QB play, without using the outdated Passer Rating, is a win in our book.

Yesterday, Chase Stuart wrote a very interesting article, trying to quantify the amount of support Quarterbacks have received in playoff games, both from their running game, and from their defense.  First, his disclaimer:

Game-ending stats are deceiving -- just one of the many caveats in my head as I wrote paragraph 1 -- but I figured there was little harm in doing some back of the envelope calculations. If nothing else, this post can just add some layers to the typical discussion of post-season records.

He's absolutely right.  I love when stat-heads throw ideas out though, as it gives an opportunity for other people to take the idea, refine it, try new things, and hopefully we'll get a stronger sense of this extremely complex issue.  I think this is interesting enough to pass along, and I'm thinking I may take a stab at refining the idea, although I don't quite have the database of stats that PFR does.

You can read his entire explanation of how he compared each game and gave it a rating, but I'll give you the Reader's Digest version:

• Defined "Support" as Rushing yard Margin (RBs) and Net Points Allowed (adjusted for defensive/Special Teams TDs).
• Found Averge and Standard Deviation for all games going back to 1950. Rush Yd Margin was Average of 0, StdDv of 93.1 Yards; Adj. Points Allowed Average was 18.9, StdDv was 11.4.
• Calculate the Standard Deviations of Support each gave the QB, and add them together.
• Put them in 5 buckets: Impossible, Difficult, Neutral, Easy, Gimmes, based on the Support given

If that's confusing, hopefully the tables of data I pulled of Peyton Manning will help clarify.  If you're still confused, please ask questions in the comments.

First, a table of the five categories, complete with an expected winning percentage, and how QB's fared:

Category Wins Losses Expected Wins StdDv Range
Impossible 0 111 0.00 <= -2.0
Difficult 13 129 0.09 -1.99 < x < -1.0
Neutral 176 188 0.48 -0.99 < x < 0.99
Easy 133 12 0.92 1.0 < x < 1.99
Gimme 117 0 1.00 x >= 2.0

The fact the Difficult and Easy categories are pretty strong in their records makes me think this is a great start to figuring this out.  We get the Expected Wins there, which Stuart used to rank each and every QB since 1950.  I'll show you a few, and remember this doesn't include 2010:

Quarterback Record Exp. W% Impossible Difficult Neutral Easy Gimme
Wins Losses Wins Losses Wins Losses Wins Losses Wins Losses
Brett Favre  13-11 51.8% 0 2 0 6 4 3 6 0 3 0
Joe Montana  16-7 57.1% 0 2 1 2 6 3 6 0 3 0
John Elway  14-7 49.0% 0 4 0 2 9 0 2 1 3 0
Terry Bradshaw  13-5 64.6% 0 1 0 2 3 2 4 0 6 0
Tom Brady  14-4 58.9% 0 1 1 0 7 3 4 0 2 0
Peyton Manning  9-9 39.8% 0 2 1 4 5 3 2 0 1 0
Dan Marino  8-10 30.0% 0 8 1 0 5 2 1 0 1 0
Ben Roethlisberger  8-2 57.4% 0 0 0 1 5 1 3 0 0 0

Does this chart tell you anything new that you didn't already know?

• Terry Bradshaw had great support, tying for the most "Gimme" games with 6, and only had 3 games below a "Neutral".
• John Elway had the only "Easy" game loss of these 8 QBs, but was 9-0 in "Neutral" games.  He definitely outperformed his support.
• Dan Marino played in 8 "Impossible" Games.  Holy cow.  Think that might be the reason for his poor record, and lack of Super Bowl ring?
• Ben Roethlisberger has only played in one "Difficult" game (which he lost), and no "Impossible" games.  He also hasn't had a "Gimme" game yet (Jets game from this year might be close).
• Tom Brady has had great support as well, and has won the only "Difficult" game he's played in ('07 Chargers).  Makes a big difference when 16 of your 18 playoff games you get a minimum of "Neutral" support.

That brings us to Peyton Manning, who by these numbers, should have only been expected to win 7 games out of 18, based on the support he was given by this teammates.  Want to see it game-by-game?

Opponent Year Rush Margin Rush StDv Adj. PA Adj PA StDv Support W/L Score
NYJ 2002 -128 -1.4 41 -1.9 -3.3 L  41-0
MIA  2000 -159 -1.7 23 -0.4 -2.1 L  23-17
NWE  2004 -164 -1.8 20 -0.1 -1.9 L  20-3
KAN  2003 -54 -0.6 31 -1.1 -1.6 W  38-31
SDG  2008 -103 -1.1 23 -0.4 -1.5 L  23-17
SDG  2007 -55 -0.6 28 -0.8 -1.4 L  28-24
TEN  1999 -119 -1.3 19 0 -1.3 L  19-16
PIT  2005 -54 -0.6 21 -0.2 -0.8 L  21-18
NWE  2003 -14 -0.2 24 -0.5 -0.6 L  24-14
DEN  2004 -2 0 24 -0.5 -0.5 W  49-24
NOR  2009 48 0.5 24 -0.5 0.1 L  31-17
DEN  2003 -61 -0.7 10 0.8 0.1 W  41-10
NWE  2006 32 0.3 20 -0.1 0.2 W  38-34
NYJ  2009 15 0.2 17 0.2 0.3 W  30-17
BAL  2009 -45 -0.5 3 1.4 0.9 W  20-3
BAL  2006 17 0.2 6 1.1 1.3 W  15-6
CHI  2006 80 0.9 10 0.8 1.6 W  29-17
KAN  2006 144 1.5 8 1 2.5 W  23-8

Well, look at that.  His six most supported games were the four when the Colts won the Super Bowl, and the two wins in 2009.  Even last year's Super Bowl was close, barely getting edged out by one of those Bronco blowout games.  Once again, we see that football is a team game, and placing blame for losses on a single player, especially one that gets terrible support from his teammates at times, is lazy and flat-out dumb.

By the way, his one "Difficult" win was the game in Kansas City back in '03, which I thought was his best playoff game ever last year.  I feel somewhat vindicated, even though only 8% of you agreed with me.  He was unreal that day, and I have another data point to prove it.

Stuart also had a paragraph dedicated to Manning vs. Brady, which you should go check out as well.  Despite what the picture the media has drawn of the two, they are very similar... He also summed the article up perfectly, so I'll just quote him:

Once again, I don't know what any of this means. I don't have an overarching conclusion, but I wanted to add to an interesting discussion in the comments by providing some data.

How can I/we improve this?  Fire suggestions away at me.