≡ Menu

Quarterback Heat Maps

Since my running back heat maps post was so popular, I thought it made sense to perform the same analysis for quarterbacks. So here’s what I did. And as a reminder, BLUE means GOOD or above-average, while RED means BAD, or below-average.

I looked at all quarterbacks with at least 100 dropbacks (i.e., pass attempts + sacks) in 2014, and then measured on what percent of their dropbacks did each quarterback gain at least 0 yards1, at least 1 yard2, at least 2 yards, etc., up to 10 yards. I also calculated the percentage of runs that went for at least 15+, 20+, 25+, and 30+ yards.

For example, among this group, the quarterbacks gained at least 0 yards on 92% of their dropbacks, and at least 1 yard on 56% of their dropbacks.3 Last year, Drew Brees gained at least 0 yards on 94% of his dropbacks, 1 yard on 63%, and 2 yards on 62% of his dropbacks; in other words, he was 2% above average at avoiding negative yards, 7% above average at getting at least 1 yard, and 8% above average at gaining at least 3 yards. In fact, Brees was between 7-8% above average at gaining between 1+ and 5+ yards on his dropbacks, although a hair below average at gaining over 20 or 25 yards on his dropbacks.

Drew Stanton and Brian Hoyer have a pair of interesting splits. Both were above average at avoiding sacks but below-average in completion percentage because they were vertical throwers; as a result, they’re each in the blue in the first column, then in the red for the next several columns, before getting back into the blue because they made a number of big plays.

qb heat map

And for those who want the information in table form:

Quarterback# DB01234567891015202530
Drew Brees687278777644221-1-10
Matt Ryan6583555333343320-1-1
Matthew Stafford6450-2-2-1-2-1-1-1-1-1100-1-1
Andrew Luck643300-100122323321
Ben Roethlisberger642144555666642021
Ryan Tannehill636-13432212101-2-2-2-1
Eli Manning62933233221221100-1
Derek Carr6212-4-4-6-6-6-6-8-7-6-7-6-4-2-1
Peyton Manning614477776344232222
Tom Brady605344443332210-1-1-1
Philip Rivers60403455332323310-1
Jay Cutler600043330100-1-2-3-2-1-1
Joe Flacco5742111223432220-10
Aaron Rodgers547254544545552211
Blake Bortles529-4-7-7-7-7-7-7-6-7-7-5-3-2-10
Colin Kaepernick528-2-2-1-1-10-1-1-2-2-2-200-1
Alex Smith507-21110112222-1-1-1-1
Andy Dalton5043333322-1-2-20-1-2-10
Russell Wilson492-2-1-2-2-1-2-100111211
Cam Newton487-1-4-4-2-20013211-1-1-1
Kyle Orton480-200-2-2-2-4-4-4-5-4-1-100
Tony Romo465088876567754200
Brian Hoyer4632-6-5-4-3-2-102353211
Teddy Bridgewater442-4-1000-1110-1110-1-1
Geno Smith3911-2-2-2-2-2-2-2-2-2-2-2010
Josh McCown362-2-6-5-5-4-4-4-4-3-5-5-1-1-10
Ryan Fitzpatrick333000111332352311
Mark Sanchez330-122343433234200
Nick Foles321401-1-1-1-1-10-10-100-1
Austin Davis314-3000000-2000-2-200
Drew Stanton2512-6-6-7-7-6-5-3-2-113321
Shaun Hill246011223210-1-1-1000
Robert Griffin244-72200-1-3-3-3-3-4-1-101
Carson Palmer233200-1-1-1-121321-100
Mike Glennon2171-3-4-3-3-3-2001-11211
Kirk Cousins2123112221002-11433
Charlie Whitehurst203-1-5-6-6-4-2-1-2-20-30001
Zach Mettenberger196-3-4-3-3-2-1-20-1002221
Jake Locker160-1-3-4-6-6-3-1-2000-1-2-1-1
Colt McCoy145-534554440211001
EJ Manuel1374-3-3-2-4-6-8-8-7-6-5-2120
Michael Vick137-4-11-11-9-11-9-11-12-12-13-14-10-5-2-1
Average925654535047433934312715953
  1. This is essentially a proxy for percentage of times the quarterback wasn’t sacked. []
  2. This is a decent proxy for completion percentage, or, frankly, an improvement on completion percentage. []
  3. Note that I took an average of the averages approach; in other words, I am placing the same weight on Brees’ dropbacks as Michael Vick‘s when calculating league average, rather than count Brees’ numbers five times as much as Vick’s because he had five times as many dropbacks. []
  • Mido

    Kinda interesting. Thanks.

  • Tom

    Derek Carr does not look good here…of the QB’s with over 600 dropbacks, he’s easily the worst…why am I remembering him being better last year? Maybe I’m equating “showing promise” with “being good”?

    • Clint

      Yeah, he played smart, but not necessarily extremely well.

  • Bryce

    Would be interesting to see the same data for QB runs. Might shed some light on who the best scrambling QB is.

    • Perfundle

      Last year it was Rodgers and Wilson. The issue you would have is that most QBs average a couple of non-kneeldown runs per game, so small sample sizes will distort the heat maps.

  • Richie

    Except for Michael Vick, the heat map almost looks like an age chart. The more blue he has, the older he is. The more red he has, the younger he is.