Writers of Pro Football Prospectus 2008

Most Recent FO Features


» 2017 Offensive Personnel Analysis

It's a three-receiver league, but for the first time since 2010, the frequency of 11 personnel actually went down last year. Was it a blip, or sign of things to come?

13 Jan 2015

Divisional Round Quick Reads

by Vincent Verhei

Following Denver's 24-13 loss to Indianapolis in the Divisional round of the playoffs, Peyton Manning was asked point blank if he planned on playing next season.

"Uh, yeah, I guess I just can't give that simple answer, I'm processing it," Manning replied. "So, I can't say that. I could not say that."

That did not sound encouraging. Then came Monday afternoon, when word broke that Denver head coach John Fox was leaving the team, and that Manning had played the last month of the season and the loss to the Colts with a torn quad muscle. The severity and recovery time of Manning's injury remain unknown, but the news only casts his future into further doubt. Manning turns 39 in March; at that age, will he really want to go through potential surgery and rehab, only to join a new coach, in a new offense, with new teammates, and basically start over for one last shot at a championship? Or will he look at the ring he already wears, his bevy of records, and his enormous pile of money and decide that he doesn't need to go through training camp and practices anymore?

Right now, anything you read, hear, or think on the subject is mere speculation, but it's only rational to suppose that Manning must at least be considering retirement right now. And if so, it'll be the end of one of the most spectactular careers the NFL has ever seen.

First of all, take a look at Manning's page on Pro Football Reference. Seriously, go take a look. See all the bolded numbers in his statistics? Those are the times Manning led the league in something, and his page has so much black ink it might stain your monitor. Since entering the league in 1998, Manning has led the league in:

  • completions four times;
  • passing yards three times;
  • completion percentage twice;
  • passing touchdowns four times;
  • NFL passer rating three times;
  • Adjusted net yards per attempt four times;
  • sack rate five times;
  • and ESPN QBR five times. (Keep in mind that this stat has only been around for nine seasons, and in one of those seasons Manning didn't throw a pass.)

And if anything, all those stats are under-rating Manning's numerical dominance. Manning led the league in DYAR and DVOA six times each, with 14 total seasons in the top five in at least one of those categories:

Peyton Manning's Career Statistics
1998 IND 16 3739 326 575 26 28 696 12 7.7% 18
1999 IND 16 4135 331 533 26 15 1581 2 34.0% 2
2000 IND 16 4413 357 571 33 15 1888 1 38.3% 1
2001 IND 16 4131 343 547 26 23 965 6 14.7% 7
2002 IND 16 4199 392 591 27 19 1076 4 15.8% 8
2003 IND 16 4267 379 566 29 10 1891 1 37.1% 1
2004 IND 16 4557 336 497 49 10 2434 1 58.9% 1
2005 IND 16 3747 305 453 28 10 1636 1 41.7% 1
2006 IND 16 4397 362 557 31 9 2317 1 51.3% 1
2007 IND 16 4040 337 515 31 14 1721 2 37.1% 2
2008 IND 16 4002 371 555 27 12 1554 2 30.0% 2
2009 IND 16 4500 393 571 33 16 1771 3 34.0% 5
2010 IND 16 4700 450 679 33 17 1400 3 19.0% 6
2011 IND 0 0 0 0 0 0 -- -- -- --
2012 DEN 16 4659 400 583 37 11 1805 2 32.8% 2
2013 DEN 16 5477 450 659 55 10 2475 1 43.2% 1
2014 DEN 16 4727 395 597 39 15 1433 3 24.5% 5

In the interest of full disclosure, we must admit that this only covers the regular season. In the playoffs, Manning has been much, much worse -- if you judge a quarterback solely by his team's win-loss percentage, that is. If you look at what Manning has done with the ball in his hands, you see that his statistics hardly change at all -- and through the lens of DYAR, which accounts for the superior teams he has faced in the postseason, he has actually played better in the postseason.

So why the disconnect? Why Manning's reputation for postseason meltdowns? Well, some people can't see past that win-loss record, which says as much or more about Manning's teammates as it does about the quarterback. And make no mistake, Manning has had his share of hard-luck playoff losses. He has gone over 100 DYAR 13 times in the playoffs, and his record in those games is just 8-5 (a 0.615 winning percentage). For comparison's sake, there were 99 100-DYAR games in the 2014 season, and those quarterbacks went a combined 85-13-1 (0.864). So right there we find three or four games that Manning should have won, but his teammates lost.

And, to be fair, he has had some stinkers, too, finishing below replacement level five times. (Not surprisingly, his teams went 0-5 in those games.) When you compare his numbers in playoff wins to his numbers in playoff losses, the difference is striking:

Peyton Manning's Average Game, Regular Season vs. Playoffs
Regular Season 179-77 0.699 35.3 23.2 65.5% 272.2 7.70 2.1 0.9 1.1 7.3 97.5 7.27 104.1
Playoffs 11-13 0.458 39.0 24.9 64.0% 283.3 7.27 1.6 1.0 1.3 9.3 88.5 6.48 110.6
Playoff Losses 0-13 0.000 40.8 24.4 59.8% 259.6 6.37 1.1 1.1 1.5 10.5 76.3 5.26 49.3
Playoff Wins 11-0 1.000 36.7 25.5 69.6% 311.0 8.47 2.2 0.9 1.1 7.9 104.8 8.09 183.0

In some respects, this is a silly exercise. I'm sure if I ran the numbers for Tom Brady, we would find that his numbers in wins were much better than his numbers in losses, and that this is also true of Drew Brees, Brett Favre, or any other quarterback with a reasonable number of playoff games. The real question is, why does Manning get so much grief for his bad playoff games, but so little credit for his good ones?

This could all be moot, of course, if Manning decides to return in 2015. Manning was still a top-five passer this season. In fact, he led all quarterbacks in both DYAR and DVOA after Week 15, before his bum leg sent his numbers (and eventually his season) into a tailspin. Who knows? Maybe Manning will come back for one more year, and maybe he'll even lead a new crew in Denver next year to his elusive second Super Bowl ring, and only then will he call it a career.

If he does, well, one way or another, you can be sure we'll be revisiting this topic in roughly one year's time.

Tom Brady NE
Brady was at his best throwing to his left, picking on the right side of the Ravens' defense. Throwing to his right, he went 5-of-9 for 38 yards; up the middle, he went 7-of-13 for 102 yards, with an interception. To his left, though, Brady went 21-of-28 for 227 yards and all three of his touchdowns.
Aaron Rodgers GB
It took Rodgers a while to get going against Dallas. On Green Bay's first six drives, a stretch that lasted midway through the third quarter, he went 11-of-19 for 105 yards and only six first downs (including a touchdown), with one sack and a fumbled snap. On their last three drives, he went 13-of-16 for 162 yards. Ten of those completions picked up first downs (including two more touchdowns). The 11th was a 6-yard gain on first-and-10.
Andrew Luck IND
If you're wondering, his two "just as good as a punt" interceptions were worth a combined -8 DYAR, which isn't very much at all. In fact, he had eight incompletions and two completions (!) that were actually worth more negative DYAR than either interception. The biggest issue for Luck was how many very short passes he threw, and how many of them were ineffective. Throwing to receivers within 2 yards of the line of scrimmage, he went 12-of-16, but only gained 45 yards. Only one of those completions picked up a first down, and only one other was considered a successful play.
Russell Wilson SEA
First and second downs: 7-of-14 for 69 yards and only three first downs. Third downs: 8-of-8 for 199 yards and seven first downs, including three touchdowns. He was also sacked twice on third downs.
Joe Flacco BAL
Flacco's results were nearly the polar opposite of his quarterback counterpart on Sunday. While Brady was great throwing to his left, Flacco's throws to that side of the field had nightmarish outcomes. To his right, Flacco went 12-of-18 for 172 yards, plus a 20-yard DPI, with two touchdowns and an interception. Up the middle, he went 10-of-15 for 117 yards and two more scores. To his left, though, he went just 6-of-12 for 23 yards (not a typo), with as many first downs (one) as interceptions.
Tony Romo DAL
There was a stretch from the first to third quarters where Romo picked up ten first downs (including two scores) in 14 throws, going 10-of-12 for 141 yards (plus a pair of 16-yard DPIs) in the process. From that point forward, he went sack, sack, 5-yard gain on third-and-23, 10-yard gain for a first down, sack, 9-yard gain on third-and-11, incomplete. That last pass was the controversial incompletion to Dez Bryant, but the game might never have come down to that if Romo and the Cowboys hadn't succumbed to Green Bay's fourth-quarter pass rush.
Cam Newton CAR
A lot of Newton's production came in garbage time. Down by at least two touchdowns in the fourth quarter, he went 8-of-10 for 123 yards with a touchdown and four other first downs. He was also sacked once and threw a very bad pick-six.
Peyton Manning DEN
Manning's first deep pass resulted in a 32-yard gain to Julius Thomas. He threw 11 more deep passes: ten incompletions, and one 17-yard gain to Emmanuel Sanders. Throwing to his right -- at Greg Toler, and away from Vontae Davis -- Manning went 8-of-18 for 28 yards with no first downs and no successful plays. His longest gain to that side of the field was an 8-yarder on third-and-9.

Five most valuable running backs (Total)
Justin Forsett BAL
Forsett had five 10-plus-yard runs against New England, and 10 first downs on the ground, while getting stuffed for no gain or a loss only four times. His two catches included a 16-yard touchdown reception.
Shane Vereen NE
All four of Vereen's completions were successful plays, including two first downs. His only carry was a 6-yard gain on second-and-4.
Eddie Lacy GB
Three 10-plus-yard runs, and two shorter first downs, while only three carries failed to gain positive yardage.
Dan Herron IND
Only two of Herron's catches were successful plays: an 8-yard gain on first-and-10, and a 10-yard gain on second-and-7. He only gained four yards first downs on the ground, and his longest run was only 8 yards, but nine of his carries gained 4 yards or more.
DeMarco Murray DAL
Murray had runs of 26 and 30 yards, and six other first downs on shorter runs. He would have ranked a lot higher were it not for a critical fumble on what appeared to be a long gain.

Five most valuable running backs (Rushing)
Justin Forsett BAL
Eddie Lacy GB
Jonathan Stewart CAR
He had runs of 15 and 16 yards. Eight of his carries gained 4 yards or more, while only 1 failed to gain positive yardage.
DeMarco Murray DAL
Dan Herron IND

Least valuable running back (Total)
Marshawn Lynch SEA
More than half of Lynch's yardage came on two runs, gains of 25 and 8 yards. Otherwise, he averaged 2.2 yards per carry with only one other first down, with three stuffs for no gain (one on third-and-1) and a fumble. On top of that, all three of his receptions were failed plays.

Least valuable running back (Rushing)
Marshawn Lynch SEA

Five most valuable wide receivers and tight ends
Jermaine Kearse SEA
33-yard gain on third-and-12, 63-yard touchdown on third-and-7, 33-yard gain on second-and-16.
Danny Amendola NE
Touchdowns of 15 and 51 yards were the big plays, but Amendola also narrowly converted a third-and-6.
Luke Willson SEA
Willson picked up three third-down conversions, including a 29-yard gain on third-and-6 and a 25-yard touchdown on third-and-10. His other catch was an 8-yard gain on first-and-10.
Kelvin Benjamin CAR
When the Panthers were within one score, Benjamin had four catches in seven targets for only 20 yards, though that included a touchdown. His three targets with Carolina down by multiple touchdowns late: 12-yard gain on second-and-8; 28-yard gain on third-and-5; 15-yard touchdown on second-and-1.
Steve Smith BAL
Gains of 19 and 16 yards, a 9-yard touchdown on third-and-7, plus a 20-yard DPI.

Least valuable wide receiver or tight end
Emmanuel Sanders DEN
Sanders only picked up two first downs against Indianapolis. He was thrown seven passes on third down and caught two of them, only one for a first down.

Posted by: Vincent Verhei on 13 Jan 2015

197 comments, Last at 22 Jan 2015, 7:13pm by intel_chris


by PirateFreedom :: Tue, 01/13/2015 - 11:45am

Tom Brady leading the QB list in rushing DYAR, a list still containing Newton, Russel, Rodgers and Luck, amuses me greatly.

by EricL :: Tue, 01/13/2015 - 11:57am

And two Seahawks receivers in the top three. Looks more like an April Fools list than what we expected to happen last weekend.

by chemical burn :: Tue, 01/13/2015 - 12:33pm

Throw in Lynch as the least valuable RB and up is down.

by Raiderfan :: Tue, 01/13/2015 - 12:55pm

Not to mention the passer with the highest rating was Edelman.

by Bobman :: Tue, 01/13/2015 - 5:37pm

Yeah, thought Wilson having a negative rushing DYAR figure was funny, but this table is a riot all around.

by coremill :: Tue, 01/13/2015 - 12:28pm

One of the fun things about Manning and Brady's careers is that they have been so good for so long that they've kind of been a natural experiment about the value of supporting casts. Look at Manning's DVOA table: does anyone really think he actually got a lot worse in 2008-10 vs. 2003-07, before suddenly finding the fountain of youth and playing much better with Denver?

Or did he just go from being surrounded by Harrison/Wayne/Stokely/Edge James/Clark + a good line led by a pro bowl LT, to Wayne/Garcon/Collie/older-Clark/Goddamit Donald Brown/Addai + a crappy line? And then suddenly in Denver, with good teammates again, he's just as good as the last time he had elite weapons.

by Will Allen :: Tue, 01/13/2015 - 1:06pm

Manning's most impressive year was 2009. That bunch had no business getting within shouting distance of a championship, and ended up getting within an onside kick off of Hank Baskett's facemask from likely winning it. Similarly, Brady carried his roster a tremendous distance in 2011, although I like a receiving corps with not very old yet Welker, Gronk, and the homicidal maniac, better than the 2009 Colts receivers, and Brady has never had a crap line like Manning had in 2009 and 2010.

by Malene_copenhagen :: Tue, 01/13/2015 - 1:29pm

On the other hand, Brady had 2006. Which, you know.

Reche Caldwell, Doug Gabriel, Chad Jackson, Jabar Gaffney.

by Will Allen :: Tue, 01/13/2015 - 1:32pm

Yeah, you're right. Brady's 2006 was better than his 2011, and 2011 was great.

by chemical burn :: Tue, 01/13/2015 - 1:33pm

Yeah, Brady in 2006 has to rank very near the top of the best QB performances with the worst wr's. Faulk didn't even go bananas that year or anything to mitigate it.

by commissionerleaf :: Tue, 01/13/2015 - 2:45pm

Manning, 2010.

I know he didn't make the Super Bowl, but that was a terrible team; the 2011 squad that got the first overall pick was basically a healthier version of the same team, only without Peyton Manning. His performance that year - albeit not great by rate stats - has to be one of the most amazing achievements in the history of the NFL.

by Malene_copenhagen :: Tue, 01/13/2015 - 4:59pm

Some dude called Kelvin Kight was apparently the target of 3 passes. I remember Bam Childress catching 2, but even I don't remember Kelvin Kight.

It was the year where Vinny Testaverde finished with the stupidest QB stat line ever for the Pats. 0 passes for 0 yards. 8 rushes for -8 yards.

by Malene_copenhagen :: Tue, 01/13/2015 - 5:00pm

no wait, he did have a few passes and 1 td. Anyway. Human victory cigar-rushes are still stupid.

by anotherpatsfan :: Tue, 01/13/2015 - 5:59pm

In light of injuries, murders and bad drafting Brady's 2013 receiving group was no walk on the beach, whether compared to Manning or otherwise.

IMO (totally subjective) Manning has over his career played more games with better receivers than Brady. If there is any statistical case to make in Brady's favor in the Brady v Manning debate, one would have to somehow filter the data by receiver quality (O line too?), and that would be a chore. If Scott K thought he could look at it objectively, he might give it a whirl...

by dmstorm22 :: Tue, 01/13/2015 - 6:11pm

That seems fair. I do think Brady has enjoyed better o-line play, certainly post-2007, than Manning consistently as well.

It is hard to really value what is more important for QB success.

by Scott Kacsmar :: Tue, 01/13/2015 - 7:22pm

I still think they play the game differently to the point where you'd have to break down all the video to get a true sense of receiver impact. Difficulty of throw, difficulty of catch, drops, YAC, broken tackle counts, etc.

Brady has played with two of the greatest TD scorers in NFL history (Randy Moss and Rob Gronkowski), and all of his best statistical seasons have come with those players.

Brady has played with several WRs with some of the highest catch rates and YAC rates in NFL history, and that's true by the numbers even when you remove him as their quarterback. Talking about Troy Brown, Wes Welker and Danny Amendola. I imagine Julian Edelman would be the same if he had any real sample size with another QB in NE (Garroppolo maybe down the road).

Over the course of their careers Manning has had better outside WR talent, but he's also been better at making those throws and that's the kind of offense he runs. I think the numbers Emmanuel Sanders and Brandon LaFell had this year would be swapped if they switched teams.

I think little progress has been made in separating the quality of the QB from the WR. There's still way too much focus on where a guy was drafted. There's not enough focus on offensive roles and the way receivers are used in their system.

We also run into problems with separating receivers from each other. Jordy Nelson and Randall Cobb are great players, but Davante Adams looked like a stud on Sunday against Dallas. His performance passed the eye test and backed up the stats. Would he be that good if he was the No. 1 and Nelson and Cobb were not present on the field? We don't know with any certainty. He would then have to make plays against better defensive backs, but what exactly in his displayed skill set would leave us to believe he couldn't do a respectable job of that? The same can be said for a team with three or four "good" receivers, but no stars. If the 3/4 guy can play, then he's going to have some favorable matchups to exploit.

Just seems like there's a large group of people that penalize a QB for successfully throwing to a 1200-yard receiver, who likely had to face tougher coverage, which required a lot of better throws from the QB. Then those same people would praise a QB for having two 600-yard receivers who are actually pretty good, but used differently so they didn't consistently pile up numbers.

Do you want one good player or two to share the load? Same thing happens with RBs. A 1700-yard rusher that handles the workload is someone that gets Pro Bowl credit and fame, but I'd rather have two guys that combine for 1700 with better efficiency, name recognition be damned.

by eagle97a :: Tue, 01/13/2015 - 9:04pm

@Scott Kacsmar I agree with all of this, good points! I think that you also have to factor in OL play with regards to pass pro and even the running game since the QB still has an impact on rushing and all of this ties with offensive efficiency. Last but not the least coaching and play calling, it's all interrelated and teasing QB impact out of all this is fiendishly difficult but nevertheless fascinating.

by amin purshottam :: Tue, 01/13/2015 - 11:35pm

What about weather? Manning has played the majority of his games in perfect weather conditions while Brady plays most of his in much worse conditions outside?

by SuperGrover :: Tue, 01/13/2015 - 11:54pm

Most? Really? There are the occasional poor weather games but the overwhelming majority of outdoor football games in the NFL are played in weather that has minimal impact on performance.

by amin purshottam :: Wed, 01/14/2015 - 3:48pm

you are kidding right? this is about brady vs. manning. Manning played for the colts, in a dome with no wind, rain, sleet, snow or cold. That is 8 home games per year plus playoffs, then there is another perfect weather condition game in Houston, then Jacksonville. That is a minimum of 10 perfect weather condition games PER YEAR. Brady plays all his games outdoors in the elements. Please dont tell me it makes no difference because I am sure if Manning played all his games in Foxborough, and Brady played most of his games in perfect weather, the stats would be different.

by Bobman :: Tue, 01/13/2015 - 5:40pm

That's ALLEGED homicidal maniac to you, sir.

An American citizen is afforded certain protections by the constitution and even if he is a ruthless, unrepentant killer, he is truly is innocent until prov--oh who am I kidding? I can't even keep a straight face while typing this.

by Pen :: Tue, 01/13/2015 - 12:36pm

I'd be interested in seeing Peyton's DVOA for regular and post season.

I'd also be interested in seeing each players DVOA for the Divisional Round.

by RickD :: Tue, 01/13/2015 - 12:56pm

"The real question is, why does Manning get so much grief for his bad playoff games, but so little credit for his good ones?"

He's lost a lot of first-round games. And about half of them were as the favored team, playing at home (including twice in the past three years). Also, he had a string of poor showings before his first Super Bowl run. By the time he'd established that he could play at his ordinary level and win games in the playoffs, the narrative had already been set.

What happened in that game against the Jets in 2003? A 41-0 loss to a 9-7 team? That stings.

by dmstorm22 :: Tue, 01/13/2015 - 1:04pm

Yes, 9 one-and-done's is a lot, but it also means that he's made the playoffs a ton. 5 of those losses came after a bye, so he's basically won five additional playoff games (and yes, you can say the same for Brady, who's gotten a bye, I believe, 9 times).

I do agree that Manning does not get nearly the credit he should for his good playoff games. Just look at his last three AFC Championships:

1.) 27-47 for 367 yards to lead a comeback from 21-3 down
2.) 26-39 for 367 yards against the best pass defense in the NFL
3.) 30-41 for 400 yards against hte Patriots.

It's almost like since they lost two of the Super Bowls following those games that they don't count.

When you adjust for defense, Peyton has some of the best playoff games in NFL History.

by chemical burn :: Tue, 01/13/2015 - 1:22pm

As his career winds down, the questions of judging him become entirely different I think. In like 2007, looking at the numbers was a way of swatting down the stupid "he's not actually good and he's a playoff choker" talk. When his career is over though, all of those upset losses and one-and-done appearances are his legacy, regardless of "fault" or their meaning in terms of Manning's talent. I mean, I don't think anyone in their right mind would even entertain the idea that he's not a HOFer or one of the greatest QB's of All-Time. But at the same time, all of those playoff loses are what happened - we're past the point where Manning needs to be "evaluated" and the story of a career simply is what it is. Some guys like Montana have amazing success in the playoffs, some guys like Marino struggle to even make it there and some guys like Elway bomb spectacularly early on and get improbable storybook endings. Manning's story is, unfortunately, one of consistent upset losses and stunningly early exits.

by Ryan :: Tue, 01/13/2015 - 1:30pm

"The narrative is the narrative because it's the narrative."

by chemical burn :: Tue, 01/13/2015 - 1:36pm

Not sure if you're agreeing or mocking! But yeah, that's what I'm saying. What happened, happened - there's no need for any kind of rehabilitation of his image or that he's some overlooked talent - you don't need to drill down into the numbers to find some proof of Manning's greatness. When all is said and done, his story is his story.

by Ryan :: Tue, 01/13/2015 - 2:00pm

Well, I was indeed mocking. Pardon the snark. But if you pay attention to the mainstream sports media (do so at your own peril) there is an IMMENSE need for image rehab. Choker choker choker, etc. I'm a Colts fan and I can't tell you how many fluky screw-ups cost our team a playoff game in the Manning era (he didn't screw up the snap count vs SD, he didn't give up a TD drive to Billy Volek, etc etc). The playoff losses are his TEAMS' stories, not just his, but that won't be his legacy.

Basically, I'm suggesting that Peyton's excellence should be asterisk-free, but it never will be, especially as he so tidily fits into the Brady "playoff hyper-dominance" dichotomy.

by PaddyPat :: Tue, 01/13/2015 - 2:44pm

The joke, of course, is that Brady's playoff performances have been shockingly pedestrian throughout his career, on the whole. I say this as a pats fan too. He was gutsy in the 2001 snowbowl, pretty darn good in the 2003 Super Bowl, great in the 2007 divisional game vs. Jacksonville, sure. He was also terrible in the 2007 AFC championship, and often mediocre and overmatched in other games, failing in the redzone repeatedly in the 2003 AFC championship, flopping in the 4 minute offense in the 2006 one, crumbling with a pick at the worst moment in the 2006 divisionals, completely falling apart in the 2010 divisionals. And yet, his remembered greatness is a reflection of clutch performances made possible by being part of a great team with exceptional coaching. The 2006 divisional game has always stood as a stern reminder of this for me. 2-minute drive down 8 to try to stay alive, Tom throws the pick. The defender doesn't get down and receiver Troy Brown pops the ball out, Patriots recover. Given a second chance, Tom drives for the touchdown. That's greatness and incompetence all rolled together, and it comes out with the label "clutch".

by Pen :: Tue, 01/13/2015 - 5:37pm

You look at Montans or Starr, you seeguys who got better in the playoffs. Higher TD%, Lower Int%, better AYA and ANYA.
You look at Manning and across the board his numbers dip.
His legacy will be a great QB who seldom rose to the occasion and never brought up the level of play of those around him.

by ClavisRa :: Tue, 01/13/2015 - 8:58pm

Another way to look at that though is, why were quarterbacks like Montana and Starr under-performing during the regular season, then suddenly showing more focus and drive when "it really mattered", as opposed to Manning who maintained an incredibly consistent high level of play game-in, game-out.

by Pen :: Tue, 01/13/2015 - 10:01pm

No, I tracked every HOF QB and most of them saw their stats dip in all five categories. This is simply because the level of competition got higher. The players who were the exceptions were also the players who led dynasties (except Tom Brady). Bart Starr, Joe Montana, Troy Aikman, Terry Bradshaw (and Kenny Stabler).

by Bobman :: Tue, 01/13/2015 - 5:54pm

Hey Gertrude Stein, I think it SHOULD come with an asterisk, but at the expense of his teammates (i.e. playoffs losses generally not his fault). Why did Billy Volek win? Freeney had lisfranc injury late in the season and the D had no pressure. Why did Antonio Gates romp over them? Brackett broke a arm late in the season and the D had no answer in the middle zone. Darren Sproles and Mike Scifres? Tim Dobbins and a forgotten snap count by a TE named Gijon. Lamar Smith from Miami (yes, the "immortal Lamar Smith" or legend and lore) runs for 200+! Eddie George has a late-game 68 yard run! Hank Baskett's face and of course Drew Brees going 22/24 in the second half. The center snapping the ball over his head and into the endzone before the count....

Manning's been generally good, extraordinary sometimes, and ordinary other times (truly sucky maybe once or twice), but many of those playoff losses are not to be laid at his feet, or not certainly at his feet alone. Even the 41-3 loss to the Jets; did he allow 41? I think the score was 17-0 before he had his second snap and the run game racked up about 13 yards--when you are one dimensional by necessity in the 1st quarter (before one-dimensional was a trend), on the road in January, a loss is expected. Maybe not 41-3, of course....

The narrative for those of short attention spans (or maybe just those who watch one football game a year--if you're not in the SB you must be a loser) is unfortunately about his failures. At least you and I know otherwise.

by Joshua Northey :: Tue, 01/13/2015 - 1:55pm

I think the counterpoint is that if you are actually talking about the "Best QB ever" and not the "QB who had the best career" or whatever, he is clearly number 1. Due to some bad luck, poor supporting casts, and untimely bad performances he didn't have quite as much playoffs success for his talent as you would expect.

But from an analytics perspective it is within the margin of error, and doesn't really justify diminishing his other accomplishments.

From a hot sports takes/random idiot fan perspective, sure he is a choker who cannot get it done. But who cares about hot sports takes and the opinions of random idiots?

by Will Allen :: Tue, 01/13/2015 - 2:01pm

I don't think the analytics are good enough in football to actually make any sort of definitive judgement of "Best QB ever". My instinct is to say about 5-10 guys, maybe more, have reasonable claim to that title. The game, in terms of individual performance, is just too context dependent to have very confident opinions on something like that.

by coremill :: Tue, 01/13/2015 - 2:18pm

Not only that, but the demands and skill set of the position have changed markedly over time, so it's not really even an apples to apples comparison. Just because they all played "QB" doesn't mean they were actually doing similar jobs. The hyperefficient short passing offenses that the top QBs of today have mastered might not have worked very well in the Deadball 1970s under different coverage and blocking rules. The Manning/Brady types who lack elite downfield armstrength might have been much less effective when the QB's primary job was to launch deep balls to speedsters, Namath/Lamonica style.

Conversely, none of the guys before about 1988 or so had to figure out how to deal with a zone blitz. It may be that the great QBs of yesteryear would have been totally lost trying to decipher the much more sophisticated modern pass defenses, with disguised pattern-match hybrid coverages played behind blitzes from unusual angles.

by Pen :: Tue, 01/13/2015 - 5:42pm

Seattle has arguably one of the greatest defenses of all time and they seldom zone blitz. Manning seemed to have more trouble facing a modern Pittsburgh Steelers than I suspect Bart Starr would have had.

by Joshua Northey :: Tue, 01/13/2015 - 5:24pm

I think the simple thing to think is just if I am some generic GM who is drafting a QB on day 1 of his pro career who do I want? You want Peyton Manning.

As far as changes in eras, that is a valid point, but generally as you go back the QBs are less important and less talented, not moreso.

Football is weird in that in some ways it barely makes sense to compare today's players to 1970s players. It is just a different game.

by Raiderfan :: Tue, 01/13/2015 - 6:56pm

I would rather have Johnnie Unitas, Steve Young, Joe Montana, or Otto Graham, not to mention the Mad Bomber.

by anotherpatsfan :: Tue, 01/13/2015 - 7:04pm

Not mentioning the Mad Bomber prolly helps your case. But all time QB is very subjective...

by Will Allen :: Tue, 01/13/2015 - 2:02pm

delete repeat

by deus01 :: Tue, 01/13/2015 - 2:03pm

Reading the comments on NFL.com is almost as bad as YouTube, it's pretty depressing and should be avoided at all costs.

by Otis Taylor89 :: Tue, 01/13/2015 - 9:50pm

The one thing that keeps Manning from being the All-Time greatest is his total lack of ability/will to leave the pocket, when needed to keep the defense honest in the playoffs. My GOAT would be Joe Montana, a great athlete, who didn't have the speed of a Wilson, but was able to pick up a 1st down on 3 and 6 when the defense was covering man to man. Even Brady, who is about as slow as Manning, wasn't doing that in the past 4-5 years but has done it several times this year. That play Sunday, when Manning had 20 yards in front of him, when he only needed 5, could have put some lfe back into the team. Look at someone like Luck - Patriots are saying he's like a 6th receiver and has to be accounted for, especially playing man to man. Even Rivers will leave the pocket to pick up a 1st down, if needed.

by beargoggles :: Wed, 01/14/2015 - 2:34am

His good playoff games have been insane.
Which makes me wonder about his variance in the playoffs. I wonder where he ranks?
I'm not trying to flame, I basically think he's the best QB ever. And most of the "narrative" ignores all the factors beyond the QB's control. But I do wonder if he's padding his playoff--advanced stats if you will--with a few fantastic games.

by Will Allen :: Tue, 01/13/2015 - 1:24pm

A one and done tournament formula really skews things, especially if people focus, in a somewhat silly manner, the w-l record. Change the long field goal performance of the Patriots early in the Belichik era slightly, and Brady and the Hooded One's playoff record has a decent chance of starting 2-3, with zero Lombardis, instead of 9-0 and three trophies, and the narrative changes completely, to the point that those two indisputably first ballot HOFers are now labeled by the meatheads, talking into microphones, or to their bartenders, as perhaps the Biggest Chokers of All Time.

by MightyMackHerron :: Tue, 01/13/2015 - 1:49pm

One reason that Manning may get so much grief for bad playoff games (or bad games in general) is the hyperbole that accompanies his team's wins. I can't remember how many times I have heard that Peyton (and to be fair, most upper echelon QBs) "willed his team to a victory". My opinion is that if a QB can will his team to a victory, he is entirely responsible for not willing his team to victory any time that it loses.

by Will Allen :: Tue, 01/13/2015 - 1:56pm

Never let one manner of meathead influence your tendency to become another manner of meathead. Favre's 2009 in Minnesota had a lot of that. The idiotic hype about Favre, which people had been subjected to for years, influenced a lot of those people into spewing a lot of idiotic anti-hype, to the point where I started call Stubbleface The Zombie King, since all humans who came into contact with him, even if only through pixels on a screen, completely lost their powers of reason.

by Joshua Northey :: Tue, 01/13/2015 - 1:56pm

This happens for every single QB. Look at Cardale Jones last night. You are just having confirmation bias of your opinions.

by dmstorm22 :: Tue, 01/13/2015 - 2:02pm

I've never heard that (apart from maybe the 2006 AFC Title Game). I've always heard excuses, like he's only beaten Jake Plummer, Trent Green types, or that his one Super Bowl came against Rex Grossman (no mention that the Bears were the #2 defense, and had great Special Teams, and it was a monsoon in Miami).

by MightyMackHerron :: Tue, 01/13/2015 - 2:35pm

"Peyton Manning wills his team to success on a yearly, weekly and daily basis"
Greg Huston, New Castle News

"It was typical Manning. With the Colts playoff hopes fading after a 3-4 start in 2008, Manning almost single-handedly willed an injury-riddled team to nine straight wins and a wild-card berth".
Michael Marot, AP.org

"Manning willed the Broncos 80 yards in the final 59 seconds to tie the game"
Mark Kiszla, Denver Post

Those are just from the first page of a google search.

by dmstorm22 :: Tue, 01/13/2015 - 2:43pm

So three isolated articles from a targeted search? That'll hold up as good evidence

by MightyMackHerron :: Tue, 01/13/2015 - 2:54pm

How many, exactly do I need to post here? Enough that I get banned for filling up the board and wasting everyone's time?

And seriously, a TARGETED search is a problem? How else do you search for something when you are looking to point out that the phrase has been used, in response to someone who has never seen it? Should I search for best fabric softener for under $2.49, and hope that I come across something?

by Ryan :: Tue, 01/13/2015 - 2:56pm

"How many, exactly do I need to post here? Enough that I get banned for filling up the board and wasting everyone's time?"


by dmstorm22 :: Tue, 01/13/2015 - 3:01pm

No, but you can pretty much defend anything with a Google Search.

Plus, you picked three articles, two of which were written last decade.

Manning is not media coddled at all for dragging his team, and if he is, that is far outweighed by the people who try to strip successes from him because he had the audacity for his kicker to miss field goals, and his TE to forget the snap count, and his safety to not play safe.

by MightyMackHerron :: Tue, 01/13/2015 - 3:14pm

I don't care either way. Originally, I tried to suggest a reason for the negative perception of Manning in the playoffs. If that gets you all butthurt, deal with it. They probably have amazing new prescription drugs for that. I posted the three quotes SPECIFICALLY in response to someone saying they had NEVER heard the "willed his team to a win" thing. This isn't my dissertation, and I won't apologize for the lack of scientific rigor. I realize now why other people have told me that this message board is full of contrary, officious d-bags. Thanks, and I leave you with this.

dmstorm is clearly too high on himself because he needs to seem smarter than me in the eyes of his fellow footballoutsider readers. Offing himself is way better than this. He is the contrary officious d-bag I mentioned earlier.

by Ryan :: Tue, 01/13/2015 - 3:26pm

Internet pro tip: Always take people who use the term "butthurt" seriously.

by duh :: Tue, 01/13/2015 - 4:26pm

Possibly offset by the 'he knows enough to have use the screen name of an obscure pretty darn good for half a year player' rule.

by Steve in WI :: Tue, 01/13/2015 - 7:42pm

My number one pet peeve regarding Manning is the people who attempt to devalue his Super Bowl victory by saying "he beat Rex Grossman." No, he led an offense which scored 22 points (should've been 25 if not for the missed FG before the half) against a great defense, and that arguably could have scored more if not for running out the clock with a 12-point lead, in adverse weather conditions.

Like it's his fault that the Bears had a less-than-stellar offense in 2006.

by beargoggles :: Wed, 01/14/2015 - 2:22am

Funny. Do those same people mention that Brady led the Pats to 13 offensive points against the Rams?

by Bobman :: Tue, 01/13/2015 - 6:07pm

I think we've all seen games where this is the case, sometimes for a QB, sometimes for a RB, or DE/LB who is just everywhere and seems to make every possible play.

QBs most often, because of their unique spot handling the ball on every offensive down. And QBs with questionable teams have a higher percentage of those. (Probably multi-threat QBs and RBs more often as well, sometimes a guy like Darren Sproles with a potent and unusual skill set.)

So I often agree that one hard-headed football demigod can take over a game every once in a while and will his team to victory. On the defeat side, maybe I am being kind, but I think it's usually "they lost despite his heroics". Sometimes he's the reason, but usually he's the reason they lost by 3-7 instead of 21. Can't blame him for THAT!

And here's a really weird example game: Colts-Chargers in the SD rain in 2007 or so. Manning throws 6 picks IIRC, one returned for a TD. Sproles returns two punts for TDs. Freeney breaks his foot and leaves the game, so the D is toothless, and still, they are a midded 45 yard FG away from tying in the end. They lose by 3. You can see six picks and say "Manning lost it" or you can see three freak returns and say the ST or random chance lost it. If not for Manning, they surely don't score as much as they do or make that late game drive to almost tie it. Had they won would anybody say he willed them to victory? I doubt it, but he certainly overcame a crap-ton of obstacles to get them there (some of his own making, some freakish god-like one-hand leaping INTs). Did Sproles will SD to victory? Well, he did all he could, but he only touched the ball maybe 20 times, maybe less.

Bottom line, I think there are games where a guy or two do "will their team" to victory, and it falls on QBs the most. But if they are generally good, as opposed to having high variance, then the losses are probably despite their best efforts, not because of them.

by Dired :: Tue, 01/13/2015 - 1:12pm

Unfortunately, the last thing we've seen of Manning was that playoff game. By the end, he wasn't just not-good; he looked helpless. I have to think that's not how he wants to go out, but at the same time he might just not have a 17 game season in him anymore. Can a team pay a player like him for a deliberate 6 games worth of play?

by deus01 :: Tue, 01/13/2015 - 1:19pm

It depends how much of that is injury related and how he thinks he can come back from that injury. Sometimes players get unlucky and get injured regardless of age, he was playing great through most of the season so I don't see a reason why he couldn't come back next season and potentially stay healthy (especially if the Broncos realize he shouldn't be throwing 50 times a game).

by chemical burn :: Tue, 01/13/2015 - 1:25pm

There's also the question of how much he wants to come back. The total house cleaning that happened with the coaches and the impending free agency issues are going to influence his desire to battle through injury. There's a not unreasonable chance he can still play at a high level - does Manning himself want to take that chance? And the additional chance his team won't be in chaos?

by Will Allen :: Tue, 01/13/2015 - 1:36pm

Well, you need to have a roster which can expose Manning to less contact, and if you pay Manning what his contract calls for, you likely don't have a roster which can do that. I'd never tell a guy he should get less money than what he can command, but Manning has to think hard about some unavoidable trade-offs.

by deus01 :: Tue, 01/13/2015 - 3:41pm

Elway stated that they want Manning back and that the OLine is priority, how they work all that in with the other free agents will be interesting.

by dmstorm22 :: Tue, 01/13/2015 - 3:44pm

My guess, Welker is gone; JT will probably go as someone will overpay. That moves Latimer to starting lineup (assuming he can play, and the reason he never saw the field was depth). Knighton resigned, Ware re-structures deal if possible, and then go o-line heavy in draft. Not to many areas of real need. Another pass rusher would be nice, as more quietly than Manning, DeMarcus Ware really slowed down in the 2nd half of the season.

by deus01 :: Tue, 01/13/2015 - 3:49pm

Yeah that would mostly be my guess as well. Not sure what they do with Franklin, sometimes he plays great and other times is a mess. Keeping JT would be great but I agree that someone will likely overpay and it won't be worth keeping him.

The pass rush seemed to slow down a lot in general over the second half so if they can beef that up or come up with some better schemes that will help.

by SuperGrover :: Tue, 01/13/2015 - 11:57pm

Agreed. Manning was playing as well as last season for the majority of the season. I really think his injury played a big part in his performance down the stretch.

by carljm :: Tue, 01/13/2015 - 1:22pm

In the comment for Dan Herron: "He only gained four yards on the ground" - eh?

by billsfan :: Tue, 01/13/2015 - 1:33pm

I have nothing substantive to add to this conversation, but am fascinated by the fact that p-f-r needs to give three decimal places for a player's age. Peyton Manning is 38.295 right now, I'll have to check back in 8.76 hours to see if the thousandths digit has changed.

by Paul R :: Tue, 01/13/2015 - 1:39pm

I think the three-digit number is days. He's actually 38 295/365ths.

If you want to see impressive numbers, check out his mother's success rate for raising NFL quarterbacks!

by Bobman :: Tue, 01/13/2015 - 6:14pm

Yeah, but early in her child-rearing career she was a miserable 0-1 and Archie was considering trade offers for Olivia, even though she was "under contract." Her second chance brought her to a .500 record (whew, the pressure's off) and on her third shot she also succeeded, putting her into the NFL-mothering HOF with a .667 average of bearing a SB-winning QB son. Not too shabby; welcome to Canton, here's your gold jacket, madam.

But for a few years there, it was touch and go. A less patient husband might have cut her and she'd be known as an 0-1 failure for the rest of her NFL life.

by anotherpatsfan :: Tue, 01/13/2015 - 6:22pm

Maybe they had the same "unlimited free agents" arrangement PM and his wife are rumored to have...hard to cut her in that case...

by Paul R :: Tue, 01/13/2015 - 1:36pm

I put Marshawn Lynch on my Loser League team for the second half, taking a chance that he might flame out. Instead he beast-modes all over the place.
So I put him on my playoff fantasy team and he does this. Ugh. Should have picked Richardson.

by Paul R :: Tue, 01/13/2015 - 1:43pm

I think Manning should retire and coach a high-school team. He'd be great at it.
Plus, I'd love to see him teach the boys' health class.

by eagle97a :: Tue, 01/13/2015 - 1:45pm

Isn't it a bit to early to write PFM's career? He still might have enough for 1 last run next season. But having said that, seeing how his season ended and reading about all the news and rumors swirling about the Broncos FO one way or another this is the end of an era. Even TB and Breesus have all shown signs of decline so the Big 4 will soon be ARod and whoever of the young guns ascend and show enough consistent production, wins, "eliteness" to be worthy of it. Just a bit sad that a new era of football is dawning kind of like when MJ retired from the NBA it just wasn't the same for me.

by Joshua Northey :: Tue, 01/13/2015 - 1:58pm

It really is about to be a new era in a way that hasn't happened in a while.

by Mike B. In Va :: Tue, 01/13/2015 - 2:04pm

Three of the next round of "greats" are in the conference championships, most likely.

by jacobk :: Tue, 01/13/2015 - 2:06pm

If Peyton Manning's teams were obviously talentless shells that he had to drag to the playoffs all by himself, how were they favored in so many of his playoff losses? I sense a smidgen of revisionist history going on.

We've progressed from the Manning hype (Best QB prospect in a generation! The next Joe Montana, only better!) to Manning backlash (Chokey McChokerson in the playoffs) and now to a counter-backlash led by analytical types. Sneer all you want, the fact remains that his playoff record is quite bad.

by Will Allen :: Tue, 01/13/2015 - 2:13pm

We could use any category you choose, to obtain about 10% of a qb's total number of games, like "2nd and 4th Sundays in October", and some HOF qb will be "best" and "worst", and those designations won't very much less meaningful, it at all, from the perspective of understanding the quality of performance, than the category of "games after December 31st", or "playoff games".

by jacobk :: Tue, 01/13/2015 - 2:33pm

And when they start handing out trophies for those categories and teams start gearing up to do their best work on those days then we might care about them. Until then we'll just have to stick with the playoffs being treated as a meaningful thing.

Are you honestly baffled by the emphasis placed on playoff performance? Every team's goal is to win the Super Bowl. Racking up passing yards and wins in the regular season is a sub-part of that goal. It may be that maxing out regular season numbers is the best way to win championships, or the only part you can really control. That seems to be Billy Beane's theory and has been adopted by a lot of analytics people. But when we see a single person's career showing a marked disconnect between lighting up the regular season leaderboards and bringing home championships it's not crazy to question the causal relationship there.

by Will Allen :: Tue, 01/13/2015 - 3:00pm

I'm honestly amused by people who can't delineate between enjoying a tournament for what it is, and making meaningful judgements about individual accomplishments in a game with 43 other starters, not including special teams, based upon roughly 10% of HOFers' games. It's pretty silly.

by Lance :: Tue, 01/13/2015 - 11:00pm

I think the "enjoy a tournament for what it is" is key here. The reality is this: pick you team postseason event, and look at the number one seed. From a betting perspective, 99 times out of a hundred, you are better off taking the FIELD vs. the #1 SEED. That means that more often than not, the number one seed is not going to win.

2013: #1
2012: #4
2011: #4
2010: #6
2009: #1
2008: #2
2007: #5
2006: #3
2005: #6
2004: #2
2003: #1
2002: #2

And those are just conference seedings-- several of those #1s may have been inferior from a "power" or "gambling" perspective to the other conference #1s (i.e. they may have been #2 overall). Point is, the top team usually doesn't win the tournament. This makes tournaments interesting, but it also means that we need to re-think how we think of "best" or "great" or whatever. Manning (or whomever) can have a sub-par game in Week 4 vs. a rival team that gets all the hype and lose. But it's OK because he plays again the following week and we take those performances with a grain of salt. But in the play-offs, with their one-and-done style, such performances are less-than-erasable.

It's easy from an outsider's perspective to say "well, if he can't perform in Big Games then [he's no good-- however you want to phrase it]" but my sense is that there really isn't that much difference. I think you'd be hard pressed to find a professional player who doesn't approach each week with an incredible desire to win-- manifested in hours of film study, working with receivers on routs and plays, with the OL on pass protection calls, etc., etc., etc. That doesn't mean it's always a success, but I'm not convinced that "play-offs" means that there's some new "level" of skill that has to be achieved and that some people have and others don't. I suppose some people get nervous and choke under pressure. But by the time you're 30 and have played in countless big games in junior high, high school, college, and the pros, some of those nerves are going to have worn off.

Anecdotes: I ran track in high school and college. I got nervous the night before each race. But I never went into a week's practice thinking "I know the people I'm racing against this week, and my personal best is better than all of theirs by 2 seconds, so I'm going to coast this week." Never. Even when I knew that to be true. This was true at the first meet as in the conference or regional meets, and true in high school and in college (though, let's be real: I wasn't beating too many people in college).

I teach and give talks at conferences in my field of study. The first time I did this-- as a grad student-- I was totally nervous. I didn't flub and handled questions, but inside I was a wreck. Now, 10 years later, when I give talks, I'm at ease. I know what I'm talking about, I know the scene, I know the people. I don't take it easy, and I get nervous (sort of), but it's not a big deal like it was early one.

These anecdotes are to suggest that "big games matter more and if you can't win those it's because you can't focus" might not be correct. It also is to suggest that after awhile, when you know what's going on, you just aren't going to be intimidated by big events like you used to. So a 30+ Maning is just not going to enter a play-off game at 35 years old with nerves that get the best of him. Been there, done that, as the kids say.

I think FO did something some time ago-- perhaps related to Manning-- about the whole increase in play-off competition and how that plays a role in judging one's success (or, "What do we make of the fact that Player X is 9-3 in the play-offs but 30-27 in the regular season, while Player Y is 8-9 in the play-offs but 50-10 in the regular season?"). I should re-read it, whomever it was. But I don't think that "being not very good" or "he just chokes [defining 'choke' as the tautological 'doesn't win play-off games']" were the reasons arrived at in the article.

by nat :: Tue, 01/13/2015 - 2:42pm

I wholeheartedly agree.

If we think playoff games are no more important than games in September, Peyton has a damned excellent legacy.

by Will Allen :: Tue, 01/13/2015 - 3:52pm

Yeah, sure, that's what my assertion was.

by Johnny Socko :: Tue, 01/13/2015 - 3:53pm

So now its clear who has actually read the article and who hasn't.

"If you look at what Manning has done with the ball in his hands, you see that his statistics hardly change at all -- and through the lens of DYAR, which accounts for the superior teams he has faced in the postseason, he has actually played better in the postseason."

by nat :: Tue, 01/13/2015 - 4:54pm

Jeez, man. Get a grip.

I wasn't responding to the article directly, and certainly not to any specific sentence in it.

I was responding to Will's very specific claim that playoffs and (cherry-picked in his hypothetical scenario) early season games are equally unimportant in judging a career's quality.

If Will were right, neither he nor you would have any problem and would take no offense at the idea that "If we think playoff games are no more important than games in September [October, I should have said], Peyton has a damned excellent legacy."

For the purpose of judging average performance treating all games as equal, he'd be right, by definition. For the purpose of judging career accomplishments treating playoff games as more important, you'd be wrong. It really depends what you think a "legacy" is for an NFL quarterback.

by Will Allen :: Tue, 01/13/2015 - 5:08pm

Well, I overstated a bit, but the more specific point is that judging performance from the metrics taken from 25 games, when we have 10 times as many to use for the task, is very unwise.

by blan :: Tue, 01/13/2015 - 4:05pm

This argument only makes sense under the assumption that playoff conditions are the same as the regular season conditions. In that case looking at a player's playoff numbers would be the same thing as looking at a random subset of his regular season numbers.

It seems clear to me however, that there are a few ways that playoff conditions are not like regular season conditions. The competition is better, the stakes are higher, the weather is usually worse, and more people are paying attention. I think these factors are probably enough to make postseason numbers different from a random selection of regular season numbers even in the limit of an infinite number of games.

We weigh postseason games higher than regular season games (postseason record is how we decide the "champion") so it makes sense to me that we could weigh postseason numbers more than regular season numbers.

Now with the specific example of Peyton Manning, from casually watching most of his postseason games and many of his regular season games over his career, it appears to me that the variance of his play per game is significantly higher in the playoffs than in the regular season. His regular season games appear almost uniformly great, whereas his postseason seems filled with unbelievable performances (e.g. the Colts-Broncos game after the 2003 regular season) and disappointing ones.

by Will Allen :: Tue, 01/13/2015 - 4:12pm

No,that isn't my assumption. My argument is that a sample of 25 games or so is wholly inadequate for the the task, and using data wholly inadequate for the task is unwise.

by blan :: Tue, 01/13/2015 - 4:46pm

I didn't mean to ascribe incorrect assumptions to your argument. The last time you had said that 25 games is not enough to evaluate a player, I made the point that 25 games is a season and a half of data, which seems to me large enough to draw some conclusions. You replied saying something about how selecting 25 games randomly from the regular season may give you a different result than looking at a player's whole career. That along with your comments above made me think that your working assumption (which may very well be true although it is unproven) is that the postseason is equivalent to a random sampling of the regular season.

Now of course any difference between a sample and some subsample may be due to randomness. If there's no reason to expect a difference based on how the sample is selected (e.g. picking out every 10th game) then the difference is almost certainly due to randomness. However, the argument that I'm trying to make is that playoff conditions are different enough from regular season conditions that we might expect a player to have different outcomes in those two sets of conditions (maybe even corrected for the strength of opposition).

by Will Allen :: Tue, 01/13/2015 - 5:10pm

Yes, but 25 games doesn't give you enough data to have confidence to judge to what degree the variance we are seeing is random, as opposed to a true measurement of skill.

by blan :: Tue, 01/13/2015 - 5:52pm

Well, we'll never know for sure. But we could calculate a p value that tells us the probability that the variance difference (if it even exists) would occur by chance. Right?

by Will Allen :: Tue, 01/13/2015 - 6:16pm

Sure, ignoring the problem of using a subsample in a one and done format, whereas the larger sample consists of 16 game seasons played no matter what. The issue I have with the way most people see things like this is that they view 100-1 odds, for instance, as something striking, and I've spent enough time at horsetracks and other venues to know that 100-1 ponies coming home is nothing earthshaking.

by blan :: Tue, 01/13/2015 - 6:35pm

Yeah, that's reasonable.

The traditional statistical significance of 0.05 has always seemed too high to me, but if it were a lot lower social scientists wouldn't be able to publish most of their results.

by Will Allen :: Tue, 01/13/2015 - 6:44pm

I was always amused that Econ Nobel Laureate Gary Becker said he left the field of sociology, as an undergrad, for what he became a Swedish Medal Model in, because sociology was too damned hard. He forgot to add the addendum "if done in an intellectually honest fashion".

by blan :: Tue, 01/13/2015 - 7:20pm

Yeah, he's quite right.

by Will Allen :: Tue, 01/13/2015 - 7:27pm

It's really a shame that the hacks have tarnished the reputation of a field of inquiry that's just mind-boggling in it's difficulty, when pursued with real rigor. Of course, that mind boggling difficulty is in good measure why it attracts the hacks.

by Flounder :: Tue, 01/13/2015 - 11:59pm

Canterbury Park? I used to go with friends on dollar Thursdays. Good times. They used to have these amateur boxing matches between races. Some of them were quite amusing.

by Will Allen :: Wed, 01/14/2015 - 1:59am

Oh, I've been there, and had fun, but my favorite tracks have been Del Mar and Santa Anita.

by Pen :: Tue, 01/13/2015 - 5:58pm

Everybody gets to pad their stats in the regular season vs weaker opponents. Cone playoff time you face the best teams. C%, TD%, ANYA, etc should all drop for most QBs. But an elite few actually rise to the occasion. Bart Starr was phenomenomally better. Montana got better. Brady and Mannings playoff stats are worse across the board. Even direct comparisons those two lead in only number of pass attempts. They didn't rise to the occasion, they got worse.

by Will Allen :: Tue, 01/13/2015 - 6:19pm

Yeah, that Mark Sanchez. What a Clutch God with Swagger, Rising Up When It Really Mattered.

by anotherpatsfan :: Tue, 01/13/2015 - 6:19pm

You are saying QB stats should drop in the playoffs (for all but the clutchiest) but criticize PM and TB for their stats dropping? Unless you quantify whether their stats dropped more or less than the theoretical average QB should drop, not sure of the point you are trying to make, unless it is just to negatively compare them with the clutchier dudes...

by Johnny Socko :: Tue, 01/13/2015 - 6:24pm

Is anybody actually reading the article? The whole point of the article is that Manning's performance is NOT worse in the playoffs. Of course the W-L record is not better, but if that's the only judge of an individuals performance, why we are even here reading about advanced statistics on FO?

by Will Allen :: Tue, 01/13/2015 - 6:31pm

Yeah, but if he had a better clutch, he'd shift into higher gear, When It Really Mattered, or something.

by deus01 :: Tue, 01/13/2015 - 6:35pm

Just like those racing movies where apparently everyone is just driving around in low gear just waiting to shift up so they can speed ahead at the end.

by deus01 :: Tue, 01/13/2015 - 6:33pm

Ars Technica had an interesting article recently on a study about how education plus ideology exaggerates rejection of reality. I'm sure some look at advanced stats only when they show what they want to believe and ignore them otherwise.

by anotherpatsfan :: Tue, 01/13/2015 - 6:49pm

His conventional stat performance is worse in the playoffs, but only slightly so (his DYAR increases but he on average throws more in the playoffs). It certainly does not drop off a cliff.

by turbohappy :: Tue, 01/13/2015 - 3:57pm

Far from talentless, but as a Colts fan there was only one year (2005) where I felt like there was a great team around him. The other years it was always a decent team (or worse) with an incredible quarterback. But with Peyton I felt like we had a chance in every game.

We can't totally ignore though the fact that Peyton's extremely high salary affects the team you can put around him.

by Bobman :: Tue, 01/13/2015 - 6:45pm

You understand that teams are "favored" by network talking heads and Vegas sports books. One group is roundly mocked for their ineptness and the other group is primarily interested in making money by having an even number of dollars bet on each side--to that end they take the pulse of the game and the betting public.

A Vegas line of -7.0 has no bearing on the play on the field--it's all about where the money is going.

Next, that favored perception is generally fueled by records, dominant wins, and (to steal a Jim Irsayism) Star Wars numbers on offense. The league rules favor offense. People like exciting offense, etc. So when your team puts up 30 pts a game, a popular mystique grows and you are deemed to be (in public perception) better than you really are, especially if your D allows 26 PPG. Late in the season, when weather creeps in (and you take your indoor team on the road to the blustery northeast) and you cannot maintain your 30 PPG offense and your D still allows 26 PPG, you end up losing.

Actually, the games in NE the Colts were never favored, prob the same for the 41-3 blowout vs the Jets. Off the top of my head I can recall three playoff games where they lost when favored at hme: SD twice and Pitt (which eventually beat everybody, so there's no actual shame in losing to the team that wins it all, though if Nick Harper's wife had not stabbed him in the leg the night before the game he might have outrun a back-pedaling Big Ben and won the game with a 99 yard fumble return. Had they won on that freak play, would they have won it all like Pitt did? Would Manning have won redemption? Would they have repeated the next year--which they DID win, cementing him as all-time postseason winner? Who knows.). He played well in that loss, but not great. H was sacked like 5-6 times and hit many more. Pit won because they were better, regardless of who was favored.

In the two SD losses, Manning's performance was superior (please check the numbers) to Rivers and Volek. One game in SD went to OT and was won by Mike Scifres (5 punts inside Indy's 10) Darren Sproles and refs who never heard of illegal hands to the face (Mc'Neil blocking Freeney). Also a TE who forgot the snap count and let a blitzing OLB run right by him for a sack inside the final 2-minute warning. How is any of that Manning's fault? It's generally referred to as the Mike Scifres game. There was nothing Manning could do about the opponent's punter playing the single best punting game ever. Also Sproles racked up what was at the time a record for total yards (return plus scrimmage), 328. The other SD loss was a loss to Billy Volek--it's fuzzier in my memory, but Indy was down their two top pass rushers and could get no pressure on Rivers/Volek. Colts punted ONCE and Manning's 400 passing yards and 3 TDs went for naught. Not sure what MORE he could have done to win that game.

There were two home meltdowns with Denver, one where is may well just be too old and broken down (this week) and two years ago where his team was forced to OT via a last-second 80-yard TD hail mary by the eventual SB champs. Again, no great disgrace losing to the team that is best/peaking, but Balt won in a freakish way. Manning threw for 290 and 3 TDs but was somewhat outplayed by Flacco who was on a SB MVP run that saw him throw 11 TDs and 0 INTs. He could have played better but hardly cost his team the game.

Please tell me of other playoff games where his team was favored and lost. I'd like to know how you think he choked the game away.

by Anonymouse :: Tue, 01/13/2015 - 7:35pm

"A Vegas line of -7.0 has no bearing on the play on the field--it's all about where the money is going."

If you think there is no correlation between winning percentage and point spread, you are out of your tree.

And per CHFF (http://www.coldhardfootballfacts.com/content/no-debate-facts-show-tom-br...), Manning has been favored 9 times against the Pats, and has lost 5 of those times (I assume this is regular season and playoffs combined).

by dmstorm22 :: Tue, 01/13/2015 - 2:07pm

Personally, I think he's coming back. This is a guy who came back from a serious neck injury that limited his arm to the point that he was embarrassed to throw in public. He was also the guy who loved the game so much he tried to convince Colts brass to let him play just red zone opportunities in 2011.

If most of the decline was due to his thigh/quad injury, I don't see how he doesn't return. He'll be healthy by the start of the season, and a healthy Peyton Manning was the best, or second best, QB in the NFL through 10 games this year.

I don't think he cares too much about losing John Fox, as there were reports he wasn't a fan of the conservative playcalling. Losing Gase may be a different story, but Peyton's won with a lot of coaches. One factor in his direction: he's had a 13-3 season with four different head coaches (Mora, Dungy, Caldwell, Fox).

by deus01 :: Tue, 01/13/2015 - 3:32pm

Elway mentioned in his press conference that they want Peyton back and that he'll also have input on the head coach search (with Gase and Del Rio both being candidates with no interviews yet). He also mentioned planning to bring both DT and JT back.

Peyton could still decide to retire, but given the talent likely to still be on the team, input into the coaching and his competitiveness and love of the game I think it seems unlikely he isn't back for at least one more season.

by dmstorm22 :: Tue, 01/13/2015 - 3:38pm

I tried to listen to the conference at work, couldn't get the feed to work. I was following various twitter feeds.

Did he say PM will have input in coaching search? That wasn't in the feeds I was reading, but makes sense, and a strong indication that he'll likely be back.

Still, it is a tough spot for a coach, especially a coach who isn't established yet. Your long term success really depends on how you do when PM leaves. If I was any coach in that interview, I would definitely try to get assurance, and make my case, that if we go 4-12 the first year after PM, that it isn't a failure, but a natural change. I wouldn't want to be Caldwell-ed.

by deus01 :: Tue, 01/13/2015 - 3:45pm

I couldn't get sound at work either but it showed up in #Broncos (following a hashtag for news is like a special level of hell) but it appeared in lots of tweets following the conference.

I agree its a hard position for the coach and I'm sure that will come up with whomever they interview. The plus side is that (pending resigning FAs) the new coach will have a lot of talent on both sides of the ball.

by blarneyforbreakfast :: Tue, 01/13/2015 - 2:23pm

I think the problem with the perception Manning is that he never had a great postseason "run" as a QB. When he won the Super Bowl in 2006, his numbers were actually pretty rough for the postseason (3 TDs, 7 INTs). His best run was probably 2009, when he lost to Brees in the Super Bowl.

Some would claim that it's just chance that Manning has only strung together 1-2 good games at a time. But in a sense he's similar to McNabb, who always seemed to come up short in the games he lost.

by The Powers That Be :: Tue, 01/13/2015 - 2:33pm

Wow, quite a change from the WC round. Six QBs over 80 DYAR this week: one last week. One negative QB this week, five last week.

by PatsFan :: Tue, 01/13/2015 - 2:40pm

Out of curiosity, what were Edelman's total, receiving, and passing DYAR numbers?

by Vincent Verhei :: Tue, 01/13/2015 - 3:08pm

I'd be interested in seeing Peyton's DVOA for regular and post season.

This was actually in there originally, but I wasn’t able to calculate it for his playoff wins and losses, so I took it out. But, for the record, his regular-season DVOA is 32.5%, and his postseason DVOA is 28.3% - - and it was 31.3% before Sunday’s loss.

In the comment for Dan Herron: "He only gained four yards on the ground" - eh?

Oops. Four FIRST DOWNS. I’ll fix that.

Out of curiosity, what were Edelman's total, receiving, and passing DYAR numbers?

Passing: 41 DYAR
Receiving: -24 DYAR
Total: 17 DYAR

by dmstorm22 :: Tue, 01/13/2015 - 3:15pm

32.5%, my word, man.

Peyton was basically as efficient for his entire career as Aaron Rodgers was this year, at least by DVOA.

It is weird that we have to adjust stats for even Manning, but years like Peyton's in '99 and '00 were in a far less offense-heavy environment as 2014.

by Bobman :: Tue, 01/13/2015 - 6:51pm

dmstorm, I LOVED your point about 13-3 seasons with four coaches. you see a lot of tv screen graphics about coaches taking multiple teams to the playoffs, SB, etc, but how many QBs can say they did THAT?!?!

I think he'll be back too, esp if Gase is and his injury was the main factor this week, and not rapid aging (which happens).

We would do well to remember that Manning was a pioneer (along with Warner, Favre...) in the transition from the pass-heavy-but-balanced-O of the 90s to the run-game-may-not-matter O of the 2010s. His first 5 years maybe were a world where balance was needed, and then the transition hit, numbers got freaky in 2004, and never reverted--a whole new world and he was one of the primary architects. Brady came to the party a little late (07) but fit right in, and guys like Rodgers and Stafford spent their whole careers in tis atmosphere.

by Scott Kacsmar :: Tue, 01/13/2015 - 5:00pm

We ran postseason DVOA and DYAR leaders after SB 48 - http://www.footballoutsiders.com/stat-analysis/2014/nfls-best-playoff-qu...

Haven't added this year yet, or new qualifiers like Romo/Luck/Dalton/Wilson. Yes, I'm sure you're all dying to see Dalton's numbers.

But if Manning dropped from 31.3% to 28.3% because of a lousy game he played injured in, then that likely drops him from 4th in playoff DVOA to 6th or even 7th (tied with Sanchez!). Maybe even 8th, though I'm not sure Wilson's been that good. And that's for a guy that played his 24th playoff game, the equivalent to 1.5 regular seasons. Seems kind of silly to get hung up on playoff numbers. We'd think of 4 vs. 7 as a sizable difference on a regular-season list, but one game made that change in the playoffs.

by Pen :: Tue, 01/13/2015 - 8:58pm

Thanks, I recently did a study on which HOF QB's saw their stats rise in the playoffs compared to their career regular season averages. The theory is that QB's face tougher teams in the playoffs and should see a dip in their stats. I compared Completion%, TD%, INT%, AYA and ANYA. Wanted to compare DVOA because I felt it was more meaningful for comparison than DYAR. The vast majority of QB's saw those stats go down in every category. It wasn't really a surprise to see the names of the ones that gained in more areas than they dropped, but for the most part, even the great ones dropped off in Completion % and had worse INT%. Bart Starr stood out as not only getting better in every category, but getting VASTLY better. Joe Montana gets better in AYA, ANYA and TD%. Peyton Manning gets worse in every category, but he has plenty of company.

Completion %, as you probably already know, varies so little that it's pretty much meaningless. I just wanted to see that it DID, in fact, drop off slightly against playoff caliber teams and it did, with a few exceptions.

I was hoping to see career vs playoff DVOA to see if that trend also showed up there. Glad to get the playoff DVOA data. Wish I had the career DVOA data to go with it.

Btw, only two other QB's with at least 150 pass attempts besides Bart Starr increased in every category and one of them is Russell Wilson, who leads in 2 of the 5 categories, so it's very probable that Peyton drops to 8th.

by Bobman :: Tue, 01/13/2015 - 9:38pm

I know it's playoffs, and some guys only have 60 attempts their whole careers, but 150 attempts is a pretty small sample. That could be three games for Brees. Say, on average, it's 5 games--not sure I'd put a lot of stock in DVOA for less than a third of a season.

Of course, these days, a team on a roll can get in three playoff games in one season (Wilson's already had 6 in three years and counting), whereas back in the day (since you bring up Starr) they had no wildcard games and considerably fewer playoff chances--was it just an NFL/AFL championship game before the SB in the late 60s, or was it purely the best record from AFL/NFL with the SB as the only playoff game? So if you are bridging eras, you have to keep the minimum number low. But it does suffer from small sample size-itis.

by Pen :: Tue, 01/13/2015 - 9:51pm

The table Scott linked to uses 150 attempts. I was only interested in HOF QBs, plus Brady, Manning, Rodgers, Brees, Wilson and Luck. I also got interested seeing if anyone besides Starr and Wilson improved in all 5 categories. Kurt Warner was the third player who did so.

Bart Starr had 213 attempts in 10 games. That's almost an entire season's worth in his day. You'd be hard pressed to find many QB's with a season's worth of playoff games. The games, however, are spread out over a QB's career so the averages should be the same, only slightly lower due to level of competition. They prove out to be slightly lower across the board except for some exceptions. Those exceptions come as no surprise. Joe Montana, Brett Favre, Bart Starr. Kurt Warner isn't in the HOF. I only mention him and Wilson because besides Bart Starr, they're the only other two QB's with at least 150 attempts to improve in every single category.

If I was to include every QB who ever played a playoff game that would be stupid for all the one and dones. But after 150 attempts, the trend to see your stats drop off due to level of competition starts to show, except for the truly great QBs.

They rise to the occasion and probably raise the level of their teams play as well.

by Will Allen :: Tue, 01/13/2015 - 10:23pm

Where do you get one of them finely calibrated occasionometers, to determine who rose to it, as opposed to having his teammates rise to it, or his opponents fall to it?

by jacobk :: Tue, 01/13/2015 - 11:44pm

You realize the playoffs are designated as important before they are played, right? It's not like "playoff performance" is some kind of ad hoc category people made up in order to cherry pick stats to trash Peyton Manning.

by Will Allen :: Wed, 01/14/2015 - 2:12am

Yeah, your misplaced sentiments aside, I really don't care about Peyton Manning, but you just go ahead and think that, since it seems to comfort you. I'll forgo for now a detailed discussion of the relevance of the concept of pre-designated important game, to the concept of 25 games in many years of one and done tournaments, compared to a sample tens times larger, of games which are played no matter what happens in the previous games.

Look, you really, really, believe that metrics obtained in playoff games are a useful tool with which to evaluate qbs who have played 10 times as many games in the regular season. Fine.

by Pen :: Wed, 01/14/2015 - 2:48am

Yes, I do, because the odds that over 90% of the QB's who had at least 150 attempts in the playoffs had lower Completion %, TD%, AYA and ANYA and worse INT% means it's enough of a sample size to infer that the quality of the opposition in the playoffs has a measurable effect on their stats. If I took the time to do so, I could calculate how much the Average Playoff QB declines in the playoffs and use that as comparison to see how far above or below average each QB played in his career.

That certain QBs, who just happen to be the ones known as the greatest QB's of all time, Bart Starr, Joe Montana, Brett Favre, as well as the QB of one of the great all time dynasties, Terry Bradshaw, ALL did much better in the playoffs than during the regular season says that these guys shined when most played as well as they could.

Manning, however, while doing no worse than others in the playoffs in most categories (ie, he got worse in all the stats I measured), did noticably worse in one interesting stat: His TD% plummeted far further than the norm. For whatever reason, he didn't score TD's at near the pace he did during the regular season.

by Will Allen :: Wed, 01/14/2015 - 9:23am

Oh, I'm sure you can infer, many, many, things, which say many, many, things to you.

by Pen :: Wed, 01/14/2015 - 2:43pm

Like you're doing right now?

by Will Allen :: Wed, 01/14/2015 - 2:49pm

Actually, I'm not. I fully admit you might be right. I'm just asking you to actually prove it, with an very empirically rigorous approach to what it means to "prove" something.

by Jerry :: Wed, 01/14/2015 - 7:53am

[W]as it just an NFL/AFL championship game before the SB in the late 60s, or was it purely the best record from AFL/NFL with the SB as the only playoff game?

Each league had its own championship game before the Super Bowl, in both senses of "before". Until the NFL went to four divisions in (I think) 1967, the two division champions met in the championship game. If there was a tie for the division lead, there was a one-game playoff to break it. (That's why the Divisional round is called that.) While the AFL always had two divisions, they added a first round of first-place team vs. other-division's-second place team for the last couple years of the league. Starting with the 1966 season, the two league champions met in the Super Bowl.

by dbostedo :: Tue, 01/13/2015 - 10:23pm

"The theory is that QB's face tougher teams in the playoffs and should see a dip in their stats."

Shouldn't you look at the quality of the defenses faced to look for whether it will dip or rise? I mean, in general playoff teams will probably have a better than average defense, but not always. And the "not always" might be enough to be significant.

by Pen :: Wed, 01/14/2015 - 2:53am

Which is why I was interested in seeing each QB's DVOA for career and playoffs. Unfortunately, the only one I have acess to is Mannings and like every other stat I measured for him, it dropped off.

DYAR isn't as good for this study because it is more a raw measure, like pass attempts, or yardage. It's a great stat, but it would reward QB's who played in soft divisions and thus got to play a lot of playoff games. I think DVOA is more suited to this kind of study.

by Red :: Wed, 01/14/2015 - 1:04am

I don't think it's hyberbole to say that Peyton Manning has faced the highest expectations of any athlete in American sports history. If he doesn't win the Super Bowl every year, people will tear him down and attempt to minimize his accomplishments. I really don't get it.

Case in point: Listening to local sports radio since Sunday night, the majority of callers want to run Peyton out of town. There's just no perspective whatsoever. In three years, the Broncos have gone 38-10 with three division titles and a SB appearance. There are 25 NFL fan bases who would kill for a run like that, but since it's Peyton Manning, that's considered a failure here. He's brought so much to the Broncos and the state of Colorado, but hardly anyone seems to appreciate it.

In general, sports fans are delusional about the potential for future success. Even if Manning declines sharply, and is only the 15th best QB next year, who else is going to play better than him? Brock Osweiler? Kirk Cousins? Brett Hundley? On some level I almost hope Peyton retires and the Broncos go 5-11 with the next "savior," just so people realize how great he really was. Rant Over.

My favorite nugget from those tables: 11 straight seasons ranking top 3 in DYAR.

by Duff Soviet Union :: Wed, 01/14/2015 - 6:03am

People just get complacent. Once a player / coach reaches a certain level but doesn't win a championship, they just assume that anyone can reach that level and that it's actually the player / coach who is holding them back from winning more.

Then the player / coach leaves and things go bad. And yet nobody ever learns their lesson.

by Scott Kacsmar :: Wed, 01/14/2015 - 2:19am

150 pass attempts is what NFL uses to officially qualify for rate stats in the playoffs.

As for comparing regular season to playoffs, I think you have to exclude games against teams like the 2-14 Raiders and make it closer to apples to apples. That's why last year I looked at regular-season games against playoff teams vs. playoff games. In using that, Manning looks like the same guy, though I don't think I used DYAR or DVOA for that one. I'm still building an absurdly detailed QB game log database.


by Pen :: Wed, 01/14/2015 - 2:59am

One would expect that against playoff caliber opponents his stats would remain the same. That's why I don't concern myself that his, like the vast majority of QB's, stats get worse in the playoffs. What I find interesting is the ones who actually get better in the playoffs.

Someone argued that such a thing would mean they coasted during the season. I disagree. In war there are guys - it's been shown to be a common phenomena - who can fight just as well as the other guy for most of the war, but when things hit a crisis, this person will suddenly fight like a superhuman and kill everyone around him, not on single occasions, but in every occasion. There are simply humans who rise to the occasion. It doesn't make Manning a choker that he didn't, but it sets Montana and Starr apart.

At least, it would seem to. Knowing their DVOA's would add more insight into this phenomena.

by Duff Soviet Union :: Wed, 01/14/2015 - 6:05am

And yet Montana had a 3 game playoff stretch, smack dab in the prime of his career, where not only did he lose all three games, but the 49ers didn't even score a single offensive touchdown with him on the field in any of them! I'm guessing nobody thought he was rising to the occasion then.

by Will Allen :: Wed, 01/14/2015 - 9:25am

He forgot to set the timer om his occasionometer.

by Pen :: Wed, 01/14/2015 - 2:52pm

That's just proof that he was facing quality opponents. Is the only barometer of a player that he go undefeated? Stupid argument by people who claim to be into statistics then use flawed logic when said statistic would show their pov to be inaccurate. (no offense to FO, but remember The Asterisk?) Pittsburgh went two years sandwiched between 4 Super Bowls where they lost in the playoffs as well. I guess they weren't really a dynasty.

Go on back to worshipping your phallic statues of Brady/Manning. All I asked was if there was a way to get DVOA data on QB's. I didn't ask for your opinion.

by Will Allen :: Wed, 01/14/2015 - 2:53pm

It hasn't occurred to you, that it might be "proof" that the quality known as "Joe Montana playoff performance" is capable of considerable peaks and valleys, has it?

by Duff Soviet Union :: Wed, 01/14/2015 - 4:29pm

"That's just proof that he was facing quality opponents."

According to the public / Vegas, 2 of those teams weren't as good as the teams Montana was playing for. And again, the 49ers didn't score a touchdown with him on the field in any of those games. You can't just wave that away by saying, "yeah, but the opponents were good".

by Will Allen :: Wed, 01/14/2015 - 4:48pm

Sure ya' can, when you start with Faith that your assertion is True. This really isn't a debate with it's foundation in observable reality. It really is more of a Belief, given a thin veneer of statistics to give it the appearance of rationality. We all tend to give short shrift to what it means to actually prove something, and engage in confirmation bias, when we want that something to be true.

by intel_chris :: Wed, 01/14/2015 - 10:35pm

@Will (RE: 151 and some related posts)

I understand your point about doing rigorous proofs v. illustrating confirmation bias. However, and this is going back to the 25 game sample size, I've seen proofs that show that given random data surprisingly small sample sizes yield results that generalize well to larger sets. In light of that, I'm not certain that a 25 game sample size (i.e. the playoff records of some QBs) is so biased that it cannot lead to insight.

Moreover, the fact that that small sample shows different characteristics than a large sample, may be indicative of there being something "interesting" about that sample. If I were to throw a coin and it came up heads 20 out of 25 times when flipped by some specific person, I would worry about it being a "fair coin" (and the person having some insight as to exploiting that unfairness) despite the coin coming up 145 times heads out of 250 times when flipped by a variety of people, and thus seemingly fair "in general". Now, that wouldn't be rigorous proof about an unfair coin, but it would be enough to make the coins fairness worth discussing, and enough to suspect (not convict) that one person of "cheating".

Therefore, although I'm certainly in the camp that takes PM's overall record as being more indicative of his skill than his playoff record, his playoff record certainly seem different enough than his general record, the stats in this article not-withstanding, especially through the coarse grained lens of W-L records, that it is worth discussing. The stats in this article do tend to show that on closer examination PM's performance in the playoffs are not sufficiently different to explain the W-L record discrepancy. However, I think the issue is not just sample size (nor cherry-picking), and thus worthy of discussion.

We are, however, left with insufficient data to "prove" our hypotheses and speculations and thus they must remain beliefs coated by a thin veneer of statistics to cloak them in an air of credibility. That doesn't mean that they are not right, just not proven.

In physics, Newton's theories worked pretty well until we had to explain Lorentz contraction, a special case (someone throwing 20 heads out of 25 throws) that violated what we knew about the other cases we had measured (the other 125 heads out of 225 throws). Gravity is pretty fair until we approach the speed of light. PM wins most games until he plays in the playoffs.

Is there a reason or is it just noise? I don't know, but I appreciate hearing theories. Some theories like phlogiston and ether, we will eventually discard when the evidence against them becomes too ponderous. Other theories will become part of the narrative that supports our ever growing need to have a narrative that confirms our own personal narrative. None of it will ever be proven.

by Will Allen :: Wed, 01/14/2015 - 11:05pm

Is it possible? Sure, and if the statement that had been made was "There are are a select few HOF qbs who might display that nebulous quality of being "clutch", that is, they can deliberately raise their performance in playoff competition, whereas the others have their performance decline, but we really just don't have enough data to make that assertion with a high degree of confidence", then I would not have bothered to disagree. What I disagree with is the claim that such a statements can be made with high confidence, or even more ridiculously, that the proposition is established fact.

by anotherpatsfan :: Thu, 01/15/2015 - 1:12am

It seems one main reason behind the clutch-related propositions being advanced is a desire to make "Russell Wilson is a super-clutch makes-teammates-better all-time great" a thing... A little early for that IMO.

by Will Allen :: Thu, 01/15/2015 - 8:38am

I think Russell Wilson is a terrific player, and may be on a HOF path. That's pretty good for a guy's fourth season.

Yeah, I'm just a simple Pilgrim trying to make my way through this Vale of Tears, and thus have the no doubt bumpkinish view that all the games after Labor Day are competed for with great energy, so if you want the clearest notion of how well a guy has done, ya' oughta give all those games equal weight.

by LyleNM :: Thu, 01/15/2015 - 12:42pm

I think you mean third season.

by Will Allen :: Thu, 01/15/2015 - 12:49pm

Well, we are about two weeks, or, if we go by league rules, about 6 weeks, away from Wilson being into his fourth year, and I don't think we are going to gain that much more data by the end of the first Sunday in February.

by LyleNM :: Thu, 01/15/2015 - 12:53pm

Nor, in fact, will there be much more data until they play an actual game in his fourth season which won't be until September.

by Will Allen :: Thu, 01/15/2015 - 1:31pm

OK, I should have written "third".

by intel_chris :: Thu, 01/15/2015 - 12:44pm

Yes, I can understand that. And, I know that concepts like "clutch", "choker", and "momentum" have been investigated and noone so far has been able to come up with data the proves that any of them exists, much less can be measured by a metric. However, words are invented to describe patterns and stick if they are patterns that other people recognize also. Thus, those three concepts do convey information, not scientifically rigorous information, but things which color are expectations. Those expectations may simply be leading us to confirmation bias, but they do influence our perception of the game.

Rationally I understand the gambler's fallacy, but when the other three top seeds had won their games, I had more trepidations about the Broncos winning also, it would balance the odds and reflect the norm that not all top seeds advance. Scientifically valid no; the odds (if you accept the premise of DVOA) should have favored the Broncos. Similarly, those trepidations were heightened because the Broncos and Colts had met before in the season, and the Colts winning would have caused the teams to split their games. Again, my internal narrative says it is hard for a team to beat another team twice in the season, so expecting a split was more likely--the gambler's fallacy in simply another guise. (And, yes, I wanted the Bengals as the opponents, just for that reason.)

You can add to those additional trepidation caused by the Broncos ever declining DVOA numbers. It just felt like the Broncos had been slipping ever since playing the Patriots. I wanted to scream in Zlionsfan template form that the Broncos were rated too high because the early season numbers were exaggerated by luck, unsustainable home fields advantage, or early season over-performance.

We also have the narrative that suggests that PM was injured much more seriously than was admitted. Something equally hard to "prove".

Moreover, we also have the general narrative that the Broncos are late season/playoff chokers. That's a hard narrative to shake. One, for those of us who endured the first five SB losses, carries weight. My Broncos version of that narrative suggests that the team expends too much effort getting to the SB and then simply arrives to lose. That narrative persists even though none of the players or coaches are carry-overs from that time, and so there is nothing to suggest a causal effect, just a correlation at best.

All of those did color my perception of the game though. All of those "facts", none of them very scientific, set my expectations. My optimism in the Broncos' and PM's ability to rise to the occasion and stage a come-back eroded as the game wore on. Those nagging doubts were of course confirmed and their narratives reinforced. The Colts 8 minute drive in the 3rd quarter seemed to "steal all the momentum" reinforcing another "discredited narrative" that still seems to resonate even if we can't validate it.

Watching this irrational debate over PM's career, where people use whatever stats they can find to express their own narratives, just shows that we cannot easily escape our own conceptual biases. The various sides will never agree. And while Chemical Burn's comment (#13) over this being now essentially (if not entirely should PM retire) in retrospective is true, it is clear that we can't even agree on the "facts" that form that retrospective. Perhaps that is endemic to being a "fan".

For this "fan" three years of getting close but falling short weigh heavily, no matter how many other team's fans are envious. I don't need a narrative to tell me that "we" have fallen short. I just need one to help rationalize the pain away.

Unfortunately the narrative that says all games are equally important doesn't fix that for me, as only the SB makes the team the champions for the year. In a few weeks the fans of the Seahawks, Patriots, Packers, or Colts will get to enjoy that privilege. It may only be 1 game, but it is the most important, and that's why it defines careers. Very few people try to claim Marino was a better QB than Montana for that very reason. If PM can't lead the Broncos to a SB victory, there may be some who won't want to even credit him as the greatest QB to play for the Broncos.

by anotherpatsfan :: Thu, 01/15/2015 - 1:00pm

Some? If PM does not win a title for Denver my guess is well over 90 percent of Denver fans would (IMO illogically) choose Elway as best ever Denver QB.

by Will Allen :: Thu, 01/15/2015 - 1:29pm

Now, now, as far as starting qbs FOR Denver, Elway indisputably is the Broncos' best ever.

by Will Allen :: Thu, 01/15/2015 - 1:26pm

The most pertinent indisputable fact about Peyton Manning at this point is that he is very, very, old by NFL standards, and anybody who doesn't adjust expectation of what an old fart is going to do in a playoff game, after a season's worth of brutality, just doesn't understand athletics generally, or the NFL specifically, at all.

One of the Clutch Rising Up Playoff Gods mentioned in this thread was Bart Starr. He was 33 when he started his last playoff game. Another one mentioned was Terry Bradshaw, and he was was 34. Montana was the same age as Manning and Favre is the real outlier, and even he proves the rule, in that the opposition in his last playoff game (in which he outplayed his opposing HOF qb), specifically targeted him for brutality, often illegal, because that d-coordinator had confidence that an old fart would be more affected by it. So now we're down to two guys to support this narrative of some HOF Fame qbs being clutch. I just don't know what to say about that, other than I won't argue with anyone for having their Beliefs, as long as they don't inaccurately claim that that their Belief has been proven.

As I said earlier in the thread, I think 5-10 guys, maybe more, can can be reasonably said to have been the greatest qb ever. I just think people aren't satisfied with that, for some reason.

by Grendel13G :: Thu, 01/15/2015 - 6:41pm

Hey now, there were only four early SB losses (assuming "early" means before the team's first win in the '97 season). Unless you want to count that 49ers steamrolling as two losses...

On the whole, I think the Broncos have been one of the top 5 teams to root for in the past 3 decades. But ever since I was old enough to remember, all of their playoff losses (except to the Bills in '91) have either been hide-the-children blowouts or crushing home losses to worse seeds. It makes for tough finales to otherwise good seasons.

by intel_chris :: Thu, 01/15/2015 - 6:57pm

My fault. In my over simplification, Craig Morton lost one and Elway lost four before 97/98. I guess I gave Elway an extra loss.

by Pen :: Sun, 01/18/2015 - 10:28am

Hardly. I went into the research blind. I only looked at Wilson out of natural curiousity as a Seahawk fan. I chose my parameters. Completion %, TD%, INT% AYA and ANYA because I thought those were the most pertinent and meaningful stats I had at my disposal.

THEN I started looking up the HOF QB's. I posted my results on another board and was asked to look up other, non HOF QB's and I did them too.

It wasn't any bias that caused Bart Starr to have such spectacular stats. Nor bias that Russell Wilson was one out of only three to rise in all those stats.

Nor was it physically possible that Joe Montana had great stats and Peyton Manning didn't because the person looking up the stats had any preconceived bias or any personal expectations. Joe Montana's stats are what they are whether I measure them or not.

To infer that I found what I was looking for BECAUSE I was looking for it is silly. I could have just as easily found that Peyton Manning and Tom Brady stood apart from all the rest.

However, what the stats say DID match my expectations. That the QB's the common fan define popularly as "clutch" do in fact have better playoff stats AND PRETTY MUCH ONLY THOSE PLAYERS, while those defined as "chokers" in fact have some glaring holes showing come playoff time is a FACT based in undeniable statistics. You can make of them what you will, but you CAN'T say it is because those stats exist solely due to my bias or what I was seeking to find. I didn't cherry pick them, I chose them before even looking up what each QB had done because those were the pertinent stats.

Popular belief that Manning chokes in the playoffs didn't cause his TD% to plummet dramatically from 5.9 to 4.1. Popular belief that Bart Starr was one of the most clutch QB's ever didn't cause his TD% to shoot from 4.8 to 7.0. Manning simply threw less TD's in the playoffs per attempt while Starr threw significantly more.

That your confirmation bias makes you decide to read into those stats differently or dismiss them outright is YOUR confirmation bias, not mine, and certainly not the stats.

You can argue all you want that it was hideously bad luck that Manning had 9 one and outs. You can point to those missed FG's or all the opposing FQCB's. But what you can't deny is that his TD% dropped vastly further than any other QB I tracked. That's nearly TWO TD's per 100 pass attempts and just MAYBE his lack of TD production had way more to do with his team being in a position to lose due to missed FG's and FQCB's than people care to admit. Because a 30% drop off in TD production during the playoffs, by definition, is choking.

by Scott Kacsmar :: Mon, 01/19/2015 - 12:27am

So...update those Russell Wilson numbers yet?

by intel_chris :: Thu, 01/22/2015 - 7:13pm

I think you can get a 15 yard taunting penalty for that one. Or perhaps the penalty for hitting a defense-less player.

by Pen :: Wed, 01/14/2015 - 3:01am


by SFC B :: Wed, 01/14/2015 - 7:15am

I swear to God reading through these comments makes me feel like I've been teleported back to baseball message boards on AOL in the 90s or listening to WEEI discussing Jim Rice's HOF candidacy.

by ChristopherS :: Wed, 01/14/2015 - 1:55pm


by Todd S. :: Wed, 01/14/2015 - 2:21pm

Bart Starr was the MOST FEARED QB of his time!

by nat :: Wed, 01/14/2015 - 2:59pm

So right there we find three or four games that Manning should have won, but his teammates lost.

That's actually at odds with history. In Manning's 13 playoff losses, his offense scored more than 20 points just two times: the 28-24 loss to SD in 2007, and the 38-35 OT loss in 2012.

It's harsh to lay a 28-24 loss all on the defense when the QB threw 2 picks. It's almost as bad to lay it all on the defense in that 38-35 loss, when you consider that the seven of those 38 points came on a pick-six, and the game was lost in OT after a dismal showing by the offense ending with another pick.

Are you really claiming that Manning led his offense to a win-worthy 18 points (wow! so many points!), but his defense blew the game he had rightly won? Because in 11 playoff losses, his offense scored 18 or fewer points.

One final analysis note: comparing 100 DYAR games in the regular season to 100 DYAR games in the playoffs doesn't really work to give an expected winning percentage. 100 DYAR losses are more common in the playoffs, especially indoors. Over the last three playoff seasons, 100 DYAR QBs have a 15-8 record (65.2%). Peyton's 8-5 record in such games is about as close as you can get to the expected value. Considering the percentage of games he plays indoors where DYAR is easier to come by, that's exactly ZERO games that he should have won, but his defense lost for him.

Which pretty much matches what we see in actual game logs.

by dmstorm22 :: Wed, 01/14/2015 - 3:49pm

It ins't harsh to lay a 28-24 loss on the defense. Neither of those picks hurt the Colts in terms of field position. It is also hard to lay the picks on Manning. One was on a well-thrown screen pass that Kenton Keith batted up into the air for no reason.

I'll blame the defense, the #1 scoring defense in the NFL, for letting Billy Volek drive down the field for the game winning TD.

I'll also blame the defense for the Ravens loss. They had to protect 80 yards over 1 minute up by 4 to win. They blew a hail mary type pass. That is unforgiveable. Yes, Manning threw a bad pick to eventually lose the game, but that's after his defense lost it multiple times over.

by nat :: Wed, 01/14/2015 - 5:00pm

In that Ravens loss, the Broncos' offense netted just 14 points over five quarters.

You might blame the defense for the loss. But you certainly can't reasonably call this a game where Manning's offense played well enough to earn a win. Thanks to excellent special teams play they might have lucked into one. But earned? Never.

Really, to think that Manning is some kind of hard luck QB whose teammates saddled a playoff paragon with a below average record is ludicrous. He lost the games where his offense failed to deliver the goods. He lost one game where his offense put up 24 points, which isn't bad, but hardly a sure victory or a surprising loss.

Since his teams were mostly built around an overloaded offense and a more mundane defense, that's about the result you'd expect. The only surprise is that his overloaded offenses failed to deliver the goods so often.

by dmstorm22 :: Wed, 01/14/2015 - 5:01pm

Yeah, a pick six that should have absolutely been called DPI.

The defense gave up 31 points. That is not good enough in any measure. 28 if you take away the field goal off the Manning pick.

Whatever it is worth, the Broncos had a lead with a minute to go and the Ravens backed up, and Rahim Moore made the worst play a DB has ever made.

I'm not absolving Manning from blame from the losses in 1999, 2002, 2003, 2013 and 2014, but all those other one's really aren't on him (or there were a lot of people that effed up worse than Manning did).

by Scott Kacsmar :: Wed, 01/14/2015 - 6:45pm

"I'm not absolving Manning from blame from the losses in 1999, 2002, 2003, 2013 and 2014"

I'd throw in the 2004 game (20-3) even though the big mistakes weren't on him. But take those six games and those are the ones where Manning didn't do enough for his team to reasonably win the game. That's six games out of 24 (and six out of 13 losses). That's not that bad.

by nat :: Wed, 01/14/2015 - 8:06pm

It sounds like "scored 14-18" points on an offensively stacked team equals "Peyton did enough to deserve a win" to you. That's an insanely low bar, even in the playoffs. You can win games like that now and then. It happens more in bad weather or with defensively tilted rosters, of course. But it's not hard luck to lose them. Not in the least.

Those are games where near perfect defensive play might save the day. They are not offensive gems spoiled by defenses.

To sum up: if we consider leading an offense to 14-18 points in a losing game a win, Peyton wins a lot in the playoffs. If only the defense had held the opponent to 13 or so points, it was a sure win! He was robbed by his awful teammates.


by Scott Kacsmar :: Wed, 01/14/2015 - 8:34pm

You really underestimate just how many QBs have loaded up on playoff wins because their teams allowed 0-16 points, which is a situation where Manning has yet to lose in his career in games he finished (81-0 I believe). Since Manning's entered the league, teams are 91-7 (.929) when allowing 0-16 points in the playoffs. That also accounts for 49.5% of all playoff wins since 1998. If we go to 0-20 points, so that a slightly below average total of 21 is good enough to win, then that accounts for 71.2% of all playoff wins since 1998.

You also need to move the PM range up to 16-18 points, because none of the games we're talking about had fewer than that. Some also had more, but I'll just point out a few games.

2000 MIA, L 23-17 OT: What looks like a 17-point game to you is a game where Jerome Pathon dropped a wide open TD on third down (+4 pts), Jim Mora tried a fake FG instead of kicking a 46-yd FG (+3-7 potential points) and Vanderjagt missed the game-winner in OT (+3 pts and the win). Not to mention Peyton had a 17-10 lead with 4:55 left and the defense let Jay Fiedler drive 80 yards to force OT.

2005 PIT, L 21-18: He got some 4Q breaks in this one (Polamalu INT overturned, Bettis fumble), but he should have done enough to at least get to OT in a 21-21 tie. But Vanderjagt cemented his legacy at the end.

2008 SD, L 23-17 OT: Only scored 17 because Scifres had an all-time punting night and the Colts had some of the worst field position in playoff history. Had a chance to win game with a 3rd-and-2 conversion, bit Gijon Robinson forgot the snap count and his man had a free path to Manning for a sack. SD tied the game and won in OT without Manning getting a possession.

2010 NYJ, L 17-16: Only had 9 possessions. Colts kept getting stuck on 3rd-and-1 in the first half and didn't have a running game capable of easily converting. Put his team ahead with 0:53 left. Jets had a big kick return and Caldwell called a dumb timeout. Jets won on last-second FG. Only the 2nd playoff game in NFL history with two lead changes in the final minute.

There's four one-and-done losses right there. Manning had 0 turnovers in those games. He had a 4Q lead in the final minute of three of them. These are usually wins for any other QB, but it's hard to win when your special teams and defense make those kinds of mistakes.

And I've studied every playoff game in the NFL. No other QB has this stuff happen to him as often as Manning. He's the only QB to have two playoff losses where his kicker missed a clutch FG. He has the most 4QC against his teams in playoff history. Very few QBs make the playoffs 9+ times, so that alone explains a lot of the record one-and-done's.

by nat :: Thu, 01/15/2015 - 10:55am

I checked. Since Manning entered the playoffs, teams scoring less than 20 points in a playoff game are 15-114. Manning is 1-11.

His low scoring games are getting rescued by great defense about as often as you'd expect.

I'm sure as a dedicated Peyton Fan Club member, you spend a lot of time compiling a catalog of his teammates letting him down. That's what Peyton Fan Club members do.

We could do the same for any QB. Brady's 5-0 in the Super Bowl using your brand of what-if analysis. Who cares? It's not what happened.

There could be a similar catalog of times Peyton let his team down. Pick-sixes, other picks, incomplete passes or poor play calls that killed drives, fumbles, you name it. Again, who cares?

"IF ONLY" is a loser's game.

The reason Peyton has lost 13 playoff games is simple: In 12 out of 24 games, his team could not generate the points usually needed to win. He won one of those anyway, thanks to great defense and Adam Vinatieri. He lost two others, because you don't win every single game where you score 20, or even 30 or more points.

by dmstorm22 :: Thu, 01/15/2015 - 11:21am

Yes, and they key word in your last paragraph is his 'team' could not generate those points.

That's my overall point at least. Here we have evidence that statistically he is quite good in the playoffs. When you adjust for the general level of defense he's played, he is very good (100 DYAR a game). He has played very well, but yet they've gone 11-13. As you mention, they haven't scored that much. That's really down to two factors:

1.) Awful field position in multiple games (2008 @ SD and SB XLIV being the two worst)
2.) The rest of his team effing up

Manning is one end of the spectrum. Brady is the other. I believe Brady is around ~85 DYAR/game (probably went up after last week). His team is 19-8. That's because Brady's team doesn't eff up as much.

by Will Allen :: Thu, 01/15/2015 - 11:50am

You just don't get it. It was obviously Brady's well manufactured clutch which caused the flight of the ball to go long and true off of the Vinatieri's foot from early 2002 through early 2005, resulting in Brady's team going 9-0 in the playoffs with three trophies, instead of possibly 2-3 and zero.

by Scott Kacsmar :: Thu, 01/15/2015 - 4:13pm

Re: 161

Actually everything I said is what happened. You seem to want to believe that things are the way they are for no other reason but QB failure instead of, you know, what actually happened. Life doesn't work that way. Cause and effect. Go watch a game instead of bringing up the box score.

by intel_chris :: Thu, 01/15/2015 - 6:51pm

Scott (re: 181)

This is a forest-trees argument. Many people don't care why those games were lost (trees), only that they were lost (forest). Now, perhaps that sample size is too small, and Vince's writings above certainly point to an interesting counter-argument. However, what actually happened is that the team lost with the QB at the helm. The QB's fault or not, it is still his record. The DYAR stats always emphasize the team nature of play, i.e. that the RB is dependent upon the line and the QB selling a fake, etc. but in the end the stats belong individually as well as collectively to the team members, so do W-L records. You cannot have collective responsibility without also having individual responsibility.

Yes, box scores may not be as precise a metric as DVOA, but they are a metric, and they are the metric by which teams advance. Dismissing box scores does not improve the sport. The box scores are what the game is played for. Taking away the scores is like giving kids all A's just on attendance in class. Unfortunately, scores, like grades, mean some teams (kids) fail. It's nice to have a metric that helps understand why your kid is failing, but in the end if the kid is failing, you don't send him to the next grade. You hold him back, get some tutoring, break his drug habit, whatever it takes to turn the kid around.

by Will Allen :: Thu, 01/15/2015 - 7:20pm

Why on earth would one rationally use a metric designed to measure the output of 53 people, working together on a task, to measure the output of one person?

by intel_chris :: Fri, 01/16/2015 - 12:26am

Part 1:

For the same reason we say "Buzz Aldrin" landed on the moon and neglect the thousands of NASA, Boeing, McDonald-Douglas, IBM, et al engineers that got him there. It is human nature to over-simplify. In fact, it is evolutionarily important to over-simplify. In most cases an answer that is reasonably approximately right under some restricted set of criteria but in-expensive to calculate (or communicate) is far better than the actual correct answer. It is so much better, that our minds don't even let us conceive what the truth looks like. Glass is transparent but solid; clouds are opaque but gaseous. Illusions abound.

Using a QB as a proxy for a whole team is not even a particularly bad simplification. It's not like as fans we have any ability to substitute in different line-men or receivers. The QB and his team are not separable items from an honest fans perspective. I couldn't replace PM with Brady if I wanted to, nor Sanders with Hilton. Fantasy Football is just that--fantasy. I can't impact the play calls nor the audibles. I don't even know who is really hurt nor how bad.

Next, if it takes you five seasons to begin to collect the data to isolate QB performance from confounding factors and the goal is to win the SB this year, you are four years late, and at least some of the team mates have been lost to free-agency or retired. After one season (or worse in the course of a season) I don't have the ability to evaluate which parts of QB fault and which parts aren't except at the grossest level, and if the teams performance changes during the course of a season, I'm even less likely to know the actual cause, especially as the opponents, field conditions, temperature, and even playbooks can change over that same period.

You like football because it is a complex team effort. However, to deal with that team effort one needs to make decisions on a play-by-play or game-by-game basis with very incomplete data. That requires condensing and simplifying complex patterns into short memes that one can communicate under less than ideal conditions.

No one knows for certain why the Broncos beat the Colts at the beginning of the season and lost to them in the playoffs. We have speculations that we need to give names to. Those named items become our narrative. When they appear to repeat, the narrative is reinforced due to confirmation bias. However, if before the game you asked me to list my fears and expectations, the narrative is all I have to distill those thoughts into. Is it an over-simplification? Yes. Is it rigorous? No. If we were in the African savannah and I saw a rustling in the bush and I said run without checking to see if it really was a lion, might it save your life to do so? Possibly.

Part 2:

Because it is the relevant metric. Football games and seasons are played to see who wins and who loses. Everything else is to help us better understand why the team won or lost, but winning is the goal, not DVOA, not DYAR, not yards, not first downs, not even points, just winning. Everything and every player on a football team is there to help the team win. Thus, in the end, that's what they all should be judged on. The broken out stats, the watching film, that analysis is simply to determine which players seem to be helping the team win the most. But, in the end, we need to hold them all individually and collectively responsible for winning. That's what they are there to do.

If you take the cynical view that it's actually about making the team money, I won't argue, but I will say that DVOA is really the wrong metric in that case.

by Will Allen :: Fri, 01/16/2015 - 2:06am

I think you have it reversed. You are using the team as a proxy for the quarterback. That's an extraordinarily bad oversimplification, assuming you are actually curious about the answer to the question "Which qb has played best", because the output of 53 football players is as different from the output of 1 player as the output of a carpenter's supervisor is different from the output of the several dozen other people who built the house. When the wiring catches fire and burns the house down, you seem to be saying it would be good simplification to blame the guy who supervised the nails being hammered. Now, if you aren't really interested in the question "Which qb has played best", then I guess we're in the wrong thread.

If you really are truly curious about why a team won or lost, then using the win or loss as a proxy for qb performance is a tremendously bad idea, because the assumption is so frequently in error. A quarterback isn't a rustling sound behind a bush, and if we think it is, then we end up not trading for Fran Tarkenton, or Jim Plunkett, or Steve Young, or Brett Favre. This seems to be a bad way to go about things.

by intel_chris :: Fri, 01/16/2015 - 2:40pm

I won't "debate" the reversal aspect, except to say, "yes, teams win games, but many people associate those wins with the QB (rightly or wrongly)." My point is simply if one wants to have football discussions outside of this forum, one had better be prepared to defend a QB's W-L record. Sometimes, you even need that inside this forum.

It may not be a good way to scout a QB. In fact, it is probably a terrible way to scout a QB, since any QB that does well enough to have good W-L stats is unlikely to be available. You are certainly right there, but if you want to argue that your favorite QB is "elite" and the QB has a lot of playoff losses, expect a lot of justifiable criticism.

My point is that the criticism is justifiable. We pay QBs (and all their teammates) to win games. If they aren't doing that, then something is wrong, and blaming the QB maybe an over-simplification, but if it is a continuing problem (like it was for the Broncos at the end of the year on the road against good teams), you need to start looking somewhere and the QB is a reasonable place to start, as he is usually the player with the most leverage on how the team performs.

And, I'm not trying to say that one shouldn't use advanced stats to suss out the actual cause, just that most people don't get that far, and if you want to get their insights you had better realize the model they are starting with, e.g. QB's W-L records and narratives picked up from sports announcers and writers.

I, for one, am not willing to claim my insights are that much more sophisticated. I may understand the concepts of field position, successful plays, expected wins added, and so forth. However, my narratives all came from something I read. I don't know enough to determine whether what I've read from the writers here are more likely correct than King, Tanier, or Estherbrook.

For example, take the criticism of Cutler's laziness in improving his mechanics. I had to take those at face value, because I don't watch enough of his play to judge, and even if I did, I'm not versed enough in throwing mechanics to see flaws in them. On the other hand, I could understand the effect that his passing was having on the W-L record. However, you can see how that combination has led me to accept a narrative I can't rightly prove. Cutler may be an above average or even elite QB, but I have accepted a judgement of him in part because his teams W-L record isn't stellar. In that, I'm no better than the average drunk in a sportsbar.

That was my point.

by Will Allen :: Fri, 01/16/2015 - 3:29pm

How divorced from measurable observation must criticism be before it isn't justifiable? I ask, because in my observation of football, using team w-l record as a proxy for qb performance, especially among playoff teams, constitutes a significant break from measurable reality.

by intel_chris :: Fri, 01/16/2015 - 6:29pm

I don't know and since I'm past the point of contributing anything useful on the topic. I will sit down and shut up.

by intel_chris :: Thu, 01/15/2015 - 6:53pm

duplicate post

by EnderCN :: Thu, 01/15/2015 - 11:21am

Go back and watch the Packers vs Giants 2012 playoff game and tell me stats are a reliable way to judge a QB. Rodgers played a pretty good game that game and the team was credited for 6 unforced drops and about 4 more should have been catches. The stat line looks awful but the Giants defense wasn't the reason.

I appreciate all the work you guys do but this isn't baseball and you can't really know what happened in a game without watching the actual game. These next level stats are certainly better than just using the more common stats, but they still aren't all that reliable.

by dmstorm22 :: Thu, 01/15/2015 - 11:23am

Sure, but over the course of a lot of games, those instances generally even out. The actual game also will let us know that the two GB touchdown drives came off of two awful calls, first Bill Leavy overturning a clear fumble by Greg Jennings, and then a terrible Roughing the Passer call with the Giants up 30-13.

by Will Allen :: Thu, 01/15/2015 - 12:23pm

Your point is about 75% right. Film evaluation, in detail, is always part of a sound judgement of a player's performance, and the smaller the sample size of games, the more important it becomes. Now, if the sample gets large enough, which, in my view, means, at the very least, 5 seasons worth of games, then you begin to approach having enough reps where you can approach having enough confidence that random stuff is being evened out, so film evaluation becomes less important, but even then, there are just way too many variables independent of the qb performance, which affect qb metrics, to use stats alone. Football simply is never going to be as amenable to statistical evaluation of individual performance in the manner of baseball or basketball. It is literally impossible. Which is one of the reasons it interests me; it's just plainly more difficult to understand what is happening.

I'll go with my favorite counterfactual to illustrate the point. Take two guys pretty close in age, named "Terry Bradshaw" and "Archie Manning", reverse their birthdays, and the teams they were drafted by, and I have a very strong suspicion that the latter would frequently touted at the greatest qb of all time, and the former would be thought of as a regional celebrity who had some moments in an otherwise unsuccessful NFL career.

Nobody is ever going to have played in enough playoff games to have enough even advanced metrics to usefully separate from the regular season. The use of playoff w-l record as a proxy for qb performance is silly in the extreme, to say the least.

by intel_chris :: Thu, 01/15/2015 - 1:00pm

While your point is probably right and if football were a dissertation topic would probably be unassailable. However, most football conversations hardly rise above the "talking to a drunk at a sportsbar level". (Even here we are not immune to sportsbar arguments.) It would be really nice to be able to be convincing in those situations also. Actually, it would be much nicer. Thus, one must understand and be able to justify those W-L records. They are the "ultimate metric" at least at the sportsbar level. And, now I return you to your talk show host on WEEI.....

by Will Allen :: Thu, 01/15/2015 - 1:46pm

It's funny. I think the game is much, much, much, more interesting, if less emotionally energizing, today, than I did 20 years, ago, and even much, much moreso (sigh, for I am old now) than thirty or forty years ago, because I don't care too much who wins any longer, unless I place what is now the rare wager. The one great thing about the Vikings stadium welfare debacle (and some other factors, like, sigh, getting old), is that I now even care much, much, less if that favorite team from childhood wins. Being more detached (never completely, of course; who can't enjoy a camera shot of Jerry Jones, as the otherwise mostly likeable Cowboys experience a bitter defeat?), from the outcome really can make the game more fun on some levels.

by duh :: Thu, 01/15/2015 - 2:09pm

I really identify with the sentiments expressed here. It is why, I think, watching folks who know what they are talking about break down tape has become about my favorite part of following the game.

If I wasn't so adamant about directly giving the NFL money it is why I'd love to get game rewind so I could watch the All-22 on a regular basis.

by Will Allen :: Thu, 01/15/2015 - 2:18pm

I think there are enough gearheads like us out there that a guy like Parcells, if he were so inclined to spend the time, could add to the estate he is going to leave his heirs, by putting together an on-line tutorial on how to break down film. I think it'd really advance the game as well, in that a lot of young people with energy would have the opportunity, that they would otherwise never get, to immerse themselves in the topic.

by Will Allen :: Thu, 01/15/2015 - 2:19pm

delete repeat

by Johnny Socko :: Thu, 01/15/2015 - 3:05pm

W-L records are the "ultimate metric" for what? For an individuals performance in what many people call the ultimate team game? If it were that simple, I don't think any of us would waste our time reading FO. We here because we are trying to delve further into the roots and causes of what makes one player and/or one team better. If we thought wins and losses are all that matter it would be too easy. Russel Wilson is the best QB! - until next year - when Joe Flacco wins, immediately making him the best QB!

Its strange how QB's are forever tied to their teams W-L record, while talent evaluation for every other position is viewed independently of wins or losses. Has anybody ever made the argument that Barry Sanders is not one of the greatest RB's of all time because he lacks a super bowl ring?

by intel_chris :: Thu, 01/15/2015 - 6:13pm

W-L are the ultimate metric for "sportsbar" arguments and I think they are used beyond the QB position also. If you want to have football conversations in the real world, you need to understand those arguments and what they mean to the holder's opinion, otherwise you can't really converse with them. Even here we get people regularly advancing their equivalents.

All the advanced stats (or normal stats) don't change the outcome of the game, and in the end that is what goes into the W-L record, that's the one metric no one can argue with. You can argue that the refs made a bad call, or the something was just luck, but in the end one team wins and another loses (or they tie). No one is required to accept your argument that the call was bad or the fumble recovery was lucky. They can also argue the relevance of any stat, is DVOA or DYAR or EPA or WPA or .... a better metric? They can't change the outcome of the game (except by wishful thinking).

We wouldn't have the concept of "garbage time" and "garbage yards" if there weren't plays that occurred where the team was unable to come-back and the yards they were making on the field were irrelevant to the W-L outcome of the game. Thus, plays are only important in so much as they impact the game. Even FO doesn't track plays that were nullified by penalties. Those plays since they don't affect the outcome of the game don't count.

The same thing goes for whole games. The Bills end of season win over the Patriots is little more than a footnote. It was a "garbage game" which didn't affect the season outcome for either team. There have been arguments here as to whether they should be included in a teams DVOA. Now, Will might rightfully argue that he's interested in who is the truly the better team, but that is still defined by the team's ability to win, and he's just trying to remove some noise from that measurement, and if the better team always won (exactly in proportion to there better-ness), he would be less dependent on advanced stats.

And all that is because, it is the W-L record, that gets a team into the playoffs, and gives them a chance of winning a SB. That is the goal of playing the season. Winning teams, teams we expect to play (and hopefully win) the SB, are the ones we care about.

In the movie Flashdance, they didn't use a bet on the Steelers/Cowboys game because they were two closely matched .500 ball clubs. They were picked because those were regular SB contenders near the time and teams everyone heard of because of their W-L records. Heck, we still remember those teams, at least if we were old enough to see them. If something other than W-L records (and in particular SB victories) mattered, I would know who Archie Manning was beyond what I do today, which is pitifully little. The Saints and their QB never made it onto my radar. The Steelers did.

Similarly, I just barely know Barry Sanders and the narratives around him in my head are not all favorable--not bad enough to not call him a top RB, but enough to know Walter Payton and OJ Simpson better. If the Lions had won a SB, I'm certain that would be different. I'm sure the narratives that I heard and internalized would have been different.

Otis Armstrong is similar in my book. I know him because he played for the Broncos and went to their first SB. I'd doubt that many others remember that he was the league leading RB the year (or two) before that, besting OJ. Moreover, no one cares.

Yes, if you want to know more about football, you can study the stats and the various breakdowns of how players have played on this site. However, if you want to talk football to the "common man" you had better know who last years SB winner was and why the defense is called "legion of boom" and why Romo is considered a choker and what "the Patriots' way" is. That's the fodder for everyday football conversations and that's why things like W-L records matter--they influence that conversation.

Do you seriously think many people would have expected the NE/BAL game to go how it did if they hadn't seen the 2012 Ravens? It was their previous "getting hot" just in time for the playoffs (and taking that all the way to the SB, including beating NE that year) that made them fearsome. Otherwise to most of the world they would have been just another 6th seed who just barely got into the playoffs.

Perceptions based upon over-simplifications are the norm. That's why we have narratives. Those are simplifications we can remember. Similarly, so are W-L records and SB champions. Even stats are simplifications. We don't consciously remember every single running play a RB had. We remember average yards per carry, percentage of time the player got stuffed, and similar metrics--that gives us only a few numbers to remember, not the details of all the snaps the player was on the field.

So, W-L records are too simple for detailed analysis. They are in contrast much easier to remember and objective. You can talk about the number of consecutive home (or away) games each team has won in the playoffs. That's something everyone can understand. We may not agree on its importance, but we know what is said. W-L records are the ultimate "sportsbar" metric, because they are something everyone can understand and remember and no one can dispute.

Thus, the world will remember PM's playoff loss record, because it is a simple fact that is easy to remember and not disputable. Quibble-able perhaps, but not disputable. His average yards per catch in the playoffs and whether it was better than his regular season record--well some here may remember, but most won't, and few would even care, unless it supports some narrative they are attached to.

by EnderCN :: Thu, 01/15/2015 - 11:51pm

Keep in mind that 2012 NE team was poor defensively and had a number of injuries on the offensive side. The Ravens had never faced a healthy Gronk in the playoffs as an example. The Packers will likely lose to the Seahawks this weekend and it will at least partially be due to Rodgers being hurt. Last year Rodgers and Mathews were both hurt and key reasons why that team lost.

To use my earlier example of the Packers dropping about 10 passes to the Giants how do you differentiate that from the Saints vs 49ers game of the same season where the 49ers hit the Saints WR over and over until they were dropping passes just from fear of getting hit.

This game is really really hard to judge on an individual basis.

Another good example is Romo this year. I don't for one minute think he changed very much, he had a great offensive line and a good running game which turned him from just a good QB to a great one. He didn't feel he had to carry the offense so he didn't force so many passes.

by DisplacedPackerFan :: Thu, 01/15/2015 - 6:04pm

I would love to be able to get data on what play the offense and defense called, then you could tell if that hole in the zone was by design or mistake, if the corner back screwed up or if it was the safety. I'm sure all the NFL coaches would love to have complete information like that too, since it would require having the offensive and defensive playbooks for the the teams involved.

But even just for a few games, say 5 - 10 years later. I know NFL films did similar type stuff for Super Bowls and that, with interviews with the coaches and players who could talk about the plays called and what not, and I loved those. But having it all there in data form so that people could look at and quantify it, and then you could really apply some individual grades to players.

Even with that, trying to evaluate it on the level of baseball or basketball would be tough. But it would be nice to get some more of the insight that teams presumably are using, and then some.