≡ Menu

On Tuesday, James “Four Touchdowns” Hanson posted a great article on the support that Peyton Manning, Tom Brady, Drew Brees and Aaron Rodgers have enjoyed throughout their careers. That was Part 1, and it received over 100 comments, so give it (and the comments section) a read. Today comes Part 2. As always, we thank our guest posters for contributing.  What follows are James’ words.


Team Support by Traditional Stats and Expected Points

About 35% to 55% of all offensive plays (depending on game script, offensive philosophy, personnel, etc.) are running plays, so there is value in looking at what each quarterback’s running game produced. Even if teams tend to run more after building a lead, it’s still a key part of closing out games. I’ve included their average league-wide ranks so we can get a better idea of how many seasons they enjoyed with great rushing support.

I’ve also included turnovers minus interceptions, which I assume are fumbles from the WRs, RBs, QBs, and Special Teams – but since I can’t determine who is responsible for what, I’ve included that information here under the assumption that most fumbles aren’t from the quarterback.

I should also note that while the rushing yards and touchdowns have had the quarterback’s contributions subtracted, the rushing first downs and expected points include any first downs gained by quarterback sneaks and scrambles.

The light green indicates the leader in that category, while the pink indicates the least amount of support in that metric.


In general, it looks like Brady and Brees have enjoyed the most rushing support while Rodgers has suffered the least amount of support by conventional metrics – and remember, those TD and first down totals include ones he picked up himself, meaning his support in those areas is likely even worse than the numbers indicate. Manning and Brady have had a top ten run game in 7 seasons, while Brees has had one 3 times and Rodgers has had a top ten running game in 2 seasons.

And again, here are those numbers for the playoffs, with the YPA average from their Super Bowl winning teams (though since Brees and Rodgers have only won one each, the average represents just one playoff run) –

Other than the Super Bowl YPA ranking, Brees seems to enjoy the most running support from his teams overall while Brady enjoys the fewest turnovers. Manning and Rodgers generally seem to get the weakest running games during the playoffs.

So what about defense? Other than the traditional metrics – average rankings, points allowed, yards allowed, etc. – I decided to look at the ANY/A and passer ratings allowed by each defense as an indicator of defensive support, as well as the relative rates for each metric.

Drew Brees is passer receiving the least defensive support as his teams are ranked last among the four in every single defensive category. The leaders are spread out among the other three players – Brady’s teams are stingy with points, Manning’s are stingy with yards, and Rodgers do best by opposing QB efficiency metrics. The weird thing is how much better Brady’s defenses are at preventing points than they are at allowing yards. He’s had a top ten scoring defense in 11 out of 14 seasons, which is incredible – by contrast, the other three QBs combined have a total of 10 seasons with a top ten scoring defense! It’s odd that his defenses’ yardage numbers end up being so much worse – if you look at the other QBs, the scoring defense tends to be pretty similar to the yardages defense. I guess Belichick really does have that “bend but don’t break” mentality… are we seeing some insight into his defensive genius? Since it’s impossible to build an elite overall defense every year, he focuses on players that can best succeed in the compressed space of the red zone? Who knows, but interesting to think about.1

Going by career passer ratings, Brady’s and Rodgers’ opposing QBs average out to a Ryan Fitzpatrick or Matt Cassel. On the other side, Manning’s defenses average quarterbacks out to Joe Flacco and Eli Manning levels, while Brees’ average out to a Carson Palmer or Derek Carr.

And those numbers in the playoffs –

While Rodgers seems to have enjoyed positive support overall, his defenses don’t do as well during the playoffs – in fact, most of his numbers are comparable with Brees’ atrocious defenses. Brady enjoys the best playoff defenses by passer rating by a substantial margin, with his squads taking the lead in nearly all categories except rankings, total yards and first downs. While some are close enough that we can’t really say anyone has a true advantage since the sample size here is so small – all teams essentially average 2 takeaways a game while 20 first downs doesn’t seem worse than 19 first downs per game. But looking at what his teams do to opposing defenses is impressive –they force opposing QBs into passer ratings and ANY/A that are average in general, as you can see by their relative rates allowed. Consistent with their overall numbers, they allow the fewest points.

Okay! That was a lot of numbers over the first two posts, but I think we have gotten an overall sense of the level of support these quarterbacks have received – while Brady’s own metrics weren’t consistently higher than any other QBs overall, he’s received far more support on average across his legendary career. When you look at his winning percentage and number of titles, we can’t really see a reason for him to be so far ahead of the others when you review his passing metrics – however, when you factor in the support he receives, I think a more consistent picture begins to form.

It takes nothing away from him, of course, unless you feel the need to believe that he’s won all his Super Bowls on the merits of his own talent and performances. Having great support doesn’t mean you’re not a great player.

So all that said, what’s the deeper correlation between team support and winning and losing? Is it possible that Brady’s plays were more timely and because he put his defenses in better positions at critical times, he set them up for success? It is possible – so I wanted to look at each player’s numbers in both wins and losses. I also want to see how often team support resulted in a win and when team support failure resulted in a loss.

In other words – how often do each quarterback’s teams win when they poorly? How often do they lose when he plays well? If you can weather bad QB play to win games and not collapse when your QB is playing well, isn’t that a key part of understanding team support?

I think so and we’ll see those results in my next column.

  1. Editor’s note: From 2001 to 2016, the Patriots ranked in the top 8 in special teams according to Football Outsiders in 12 of those 16 years, and never below 16th. New England’s average ST rank over that period is 7.6, and always having good-to-great special teams likely helps in this department, too. []
  • Four Touchdowns

    Thanks for publishing, Chase!

  • Brady is a turnover-avoiding monster, which bears itself out in the defense’s stats. His offenses rarely give their defenses short fields, so opposing teams have greater opportunity to pick up yards and, simultaneously, worse opportunity to pick up points. Over their full careers, Brady and Rodgers have similar fumble rates (1.27 per play for AR and 1.17 for TB), but Brady has been particularly magnificent at holding onto the ball since 2007. Since then he has fumbled just 48 times on 5688 plays, good for 0.84 per play.

    Combine that with New England’s general superiority on special teams, and the defense is always going to have an easier time preventing scores, even if they allow a few more yards than you’d like.

    • Four Touchdowns

      Honestly, I had wondered if his super-low interception rate was the metric that provided the key to his teams’ success… but Rodgers has similar if not superior overall turnover avoidance and hasn’t seen close to the kind of dominance that the Pats have had with Brady.

      • Depends on your timeframe. For his whole career, Brady has had a fumble or interception on 2.82% of his plays. Rodgers bests him with a lower 2.58% figure. However, in what you might call the GOAT half of his career (since 2007), Brady is at a 2.14% rate.

        If we look only at fumbles lost, Brady has a 2.11% overall TO rate and a 1.60% TO rate since 2007. Rodgers has a 1.76% TO rate.

        Their numbers aren’t that far off, so when you include the consistently excellent ST play from New England, there is a clear advantage. Also, Belichick is a better coach than Capers. Capers seems to base his defensive philosophy on creating turnovers, and when the turnovers don’t happen, they are quite awful.

        • Four Touchdowns

          Interesting that post-2007 is the GOAT half of his career since he’s 2-2 in Super Bowls at that point while the majority of his rings come prior to that season.

          (Though I agree that’s when he was the better QB, just referring to the common perception.)

          How were Belichick’s defenses prior to Brady? I think he was credited with stopping Jim Kelly for the Giants in 1990, right? I don’t know much about him overall, I guess.

          • I don’t think there has ever been a time when Belichick wasn’t a highly respected defensive mind. Also, there are only three coaching gameplans ever inducted into the Hall of Fame. One of from Weeb Eubank from the 1958 NFL Championship Game. The other two are from Belichick: one as DC against the Bills in the Super Bowl, and one as HC against the heavily favored Rams in the Super Bowl.

            If he’s not the greatest coach of all time, he’s certainly the best at preparing gameplans. So you have Belichick’s ability to coach, Brady’s ability to avoid turnovers (and move the ball very well too; he isn’t Damon Huard), and the prowess of NE special teams all working in concert to make things better for each other.

            If Brady and ST aren’t as good at what they do, it’s harder for the defense. If the defense doesn’t excel at what they do, it creates more work for Brady and the offense. If ST provides great field position, that helps both sides of the ball. If ST provides terrible field position, that hurts both sides of the ball. It’s an intricate web of inextricably entwined variables.

            • While true, there was an ugly period a few years ago for the Patriots defense, really from ’08 to ’13 with 2011 being a disaster. The ’07 Patriots defense was oldddddddd. It had 38 year old Junior Seau, 35 year old Rodney Harrison, 34 year old Tedy Bruschi, 32 year old Mike Vrabel, a pair of 30-year-olds in Colvin and Thomas at linebacker, and a 28-year-old Jarvis Green who didn’t do much after 2007.

              Anyway, from ’08 to ’13, the Patriots D struggled in general and in particular against the pass. Here’s how they did in DVOA each year, both overall and against the pass.

              https://uploads.disquscdn.com/images/00ec6ff97a29819d20aa20352b0fbfc66a0c0771864eefada70946a0b5c6267f.png

              Now, perhaps DVOA isn’t the right metric, but New England’s D was really good in ’03 and ’04 when they won it all, and was really good again in ’06 when they made it to the AFCCG. In ’07 it was still pretty god, but then the wheels fell off for a bit. Frankly, the Patriots didn’t hasn’t been that good in awhile, even if the team led the league in points allowed this year: http://www.footballperspective.com/new-england-led-the-nfl-in-points-allowed-in-a-unique-way/

              • This is why I think Brady’s (and the entire offense’s) ability to avoid turnovers, in concert with ST play, is so important to the overall perception of the defense. On a per play basis, the defense struggled since the undefeated season. However, even though they never ranked in the top 10 in DVOA, they ranked in the top 10 in points allowed every year but 2011 (when they ranked 3rd in ODVOA and 5th in STDVOA). I think having an offense that ranked in the top 7 in DVOA every year since 2004 has helped the defense tremendously.

                • Four Touchdowns

                  I haven’t looked up DVOA numbers but why don’t you think this applies across the board? Other great offenses by DVOA don’t always have good defensive DVOA.

                  • I’m not sure I understand your question. I think the efficiency of the NE offense has helped make the defense’s DVOA largely irrelevant.

              • Four Touchdowns

                I was surprised to see their Defensive SRS was at 5.0, which was higher than their offensive SRS at 4.3, and the third highest defensive SRS for 2016 after the Broncos (6.6) and Giants (5.4).

                How valuable a metric do you feel SRS is?

                • Paul

                  Its just points differential adjusted for SOS right? So it doeant take into account yards or any other metric but points. So as chase and gridfe have pointed out, NE points against doesnt really reflect their defense since theyre so rarely put in bad situations.

                  So its not that advanced of a metric, hence why they call it Simple Rating System. It slightly more accurate than generic point differential, thats it.

            • eag97a

              That’s why I mentioned in another thread that you need to construct a series of partial differential equations to handle the multiple variables for football. And it won’t be enough since a lot of the variables are unknowable, unquantifiable and at worst may not even be a variable. I also mentioned in that comment that even if you can construct said equation you will only find families of solutions and not exact answers due to the same causes. I’m not a mathematician but you can ask any of them and they will say that in general partial differential equations are not exactly solvable except for certain cases.

              • Richie

                Well, that’s helpful.

                • eag97a

                  What it means is analyzing real world processes like football is very hard. Not all events or situations can be reduced into a simple system like a simple harmonic oscillator (and even that example has chaotic tendencies…smh)

        • Four Touchdowns

          While it’s a fact that Brady’s overall TO avoidance is legendary and probably the key stat of his career when comparing him to other legends, you got me wondering if it’s the key to his playoff success.

          According to PFR, his INT% in the playoffs is 2.3% — just 0.1% ahead of Manning and way behind Brees and Rodgers, who are at 1.3% and 1.7% respectively.

  • sacramento gold miners

    Team support is helpful, but the Los Angeles Rams of the 1960s and 1970s are the poster child for needing that HOF QB. Only the 1979 club reached a Super Bowl, and Vince Ferragamo(spelling?), faded after 1980. Those Rams teams were loaded with talented up and down the roster, but just couldn’t produce that level of QB to get them to a higher level of success. Roman Gabriel was no Kurt Warner.

    • Four Touchdowns

      I think you’re missing the point of the team support articles — the point isn’t that you don’t need a great QB or that any of these QBs are necessarily better than the others.

      It’s simply that team support matters and not every QB is gifted with equal amounts of it. Let’s put it this way — let’s say a QB can bring his team right up the finish line of victory and defeat. Team support is either going to push him over the finish line or act as an anchor that holds him back from it.

      An imperfect analogy, but I am definitely not trying to say the QB isn’t the most important guy on the field.

      • sacramento gold miners

        I definitely understand your point, but the elite QB is more likely to make those crucial plays to get that team across the finish line. This is where the intangibles factor into the equation. And sometimes the supporting cast can make a mistake on a play, but that elite QB can still deliver the timely play for victory.

        A great example of what I’m talking about is the 1986 AFC TG, when the long snap from center bounced off the thigh of a Broncos receiver. Another QB could have been adversely affected, but Elway delivered the game-winning TD pass. Another aspect of the game which doesn’t show up in the box score, but goes down as part of the story of the game.

        • Four Touchdowns

          “…but that elite QB can still deliver the timely play for victory.”

          Sorry, but this just doesn’t make sense to me once you put some more thought into it. Let’s say the elite QB is down one possession and trying to make a game-winning drive.

          What if the QB gets the ball and no one’s open on time? What if the OL is facing a monster pass rush and can’t give him enough time for plays to develop? What if they’re playing in a system that requires the QB and WR to read the defense simultaneously and improvise routes (like the West Coast system) and the WR makes the wrong read and the QB throws an INT because of it? What if the QB makes a perfect pass and the WR drops it? What if the offense gets into field goal range and the kicker shanks it?

          This isn’t boxing where all the variables are within one athlete’s control.

          • sacramento gold miners

            Agreed, bad things can happen. But especially in the postseason, the elite QBs have that little edge, and with so many close games, it proves to be just enough. The same type of QB just seems to get it done more than others. If Kurt Warner is on those 1973-1978 Rams, I will always believe they reach a Super Bowl, and possibly win one.

            • Four Touchdowns

              You’re probably right but that’s taking away a bad QB or QBs (I assume) and replacing him with a HOF QB. Any team will get better, especially a great team.

              The real question is could you swap a great QB with another great QB and get any better results? Could Tom Brady have taken Marino’s Dolphins to greater playoff success while running the Air Coryell offense? Would Terry Bradshaw have led Tarkenton’s Vikings to Super Bowl victories?

              • sacramento gold miners

                Now you’re talking about different offensive systems, and it’s difficult to say what would happen in trading elite QBs. Brady and Bradshaw would have been poor fits in those offensive attacks, adjustments would have had to be made.

                • Four Touchdowns

                  Well, that’s sort of the point of what we’re saying here, no? QBs don’t get to control the schemes of the teams they’re drafted to — sure they have input, but it’s not like a coach who has been using the West Coast system his whole career will suddenly switch to the Air Coryell system because the QB they drafted likes it more.

                • Tom

                  And that’s exactly the point. We can’t point to the elite QB having these special powers that are his alone, and then say, well, they wouldn’t quite work if you put him over on this team.

                  • sacramento gold miners

                    Right, we only live in this reality. But I will always believe in general terms, your team’s chances for victory increase with a better QB. The non HOF QBs which have won Super Bowls were usually at least good QBs, with the exception of Trent Dilfer.

                    • Tom

                      I’m with you on that for sure (with all due respect to Dilfer).

              • Paul

                Ive never really understood the logic behind ‘can player X have been as great as player X in situation X. Id rather consider things we can largely know, not hyperbole. If Dan Marino had played for the 49ers, would he have played like Dan Marino or Joe Montana, theyre style and skill set are a result of sitautions and experience. The best qbs in history are who actually played and have results we know. Its like wondering if greg cook would have been the best qb in history if he wasnt injured. What ifs are mostly useless concepts.

                If brady hadnt been injured in 2008 he likely would have put up better numbers then 2007 since the patriots 08 schedule was laughable compared to 07, and would have won the superbowl as that years playoffs was pretty weak.

                What ifs are just wild conjecture, we dont know what would have happened.

                • Four Touchdowns

                  Frankly, the whole idea of determining the “greatest QB of all time” is just conjecture and a useless concept. The world’s greatest QB is about as important as the world’s greatest juggler — it doesn’t actually matter.

                  If we just judge everything on wins, then there was no point in writing this series of articles — the best QBs get the most wins and titles. Terry Bradshaw is better than every other QB except Joe Montana, Tom Brady, Bart Starr, and Otto Graham and we don’t need advanced analytics and sites like Football Outsiders, Football Perspective, etc. Just look up the win-loss records and who has the most championships and you’ll know who is the best. The best quarterback, coach and team win the Super Bowl every year.

                  We imagine hypothetical situations because they’re thought exercises that allow us to consider players and teams in different contexts. It’s not a lot different than saying, “I think Tom Brady’s stats would have been better if he had Marvin Harrison for 10 years.” We don’t actually know that but we have a feeling that it is likely true.

                  That said, if you don’t find hypothetical scenarios useful, that’s 100% fine — just ignore us when we discuss them. 🙂

    • Tom

      I don’t think James (or anyone here) is saying that you can have a team with great defense and special teams and consistently make it to the Super Bowl. A great, or at least very good, QB is still needed if you want to get there more than a few times.

  • Four Touchdowns

    Total side-note but I just discovered the Playoff Leaders page on PFR. The small sample size of the playoffs and why they’re a poor way of judging QBs has never been more evident than looking at the all-time leaders in playoff passer rating and ANY/A.

    Some of the guys with better playoff ANY/A than Brady? Tim Tebow, Mark Sanchez, Colin Kaepernick, Jay Cutler, Matt Stafford, Alex Smith and Joe Theissman. Shockingly, Brady doesn’t crack the top 50!!!

    And passer rating? The list looks more reasonable but Sanchez and and Theissman are still ahead of Touchdown Tom.

    Pretty nutty.

    http://www.pro-football-reference.com/leaders/pass_adj_net_yds_per_att_career_playoffs.htm

    http://www.pro-football-reference.com/leaders/pass_rating_career_playoffs.htm

    • Richie

      Tebow completed 19 passes in the playoffs. Six of those 19 completions went for 30+ yards.

      In the Pittsburgh game, he had 10 completions. Five of those went for 30+ yards.

      Crazy.

  • I think you can believe Brady is great but also that he receives more support than his great(er) contemporaries. Basically I’d call him Derek Jeter or Kobe Bryant. Not denying they weren’t great in their own right. But also not denying that they had benefits their piers didn’t. Anyway, thanks for posting, Four Touchdowns. Good stuff.

    • Four Touchdowns

      Thank you, Dave — you’re actually spot on. Joe Montana got more support than Jay Cutler, no doubt — but he’s also much better than him on top of it! This series isn’t designed to take anything away from these HOF worthy QBs, it’s just to see what team they had behind them.

      Or to put it another way — they’re all Superman, but some of them have Batman and Green Lantern on their Justice League and others have Aquaman on their’s, LOL.

      • I like the Montana to Cutler comparison. Drives home the point more. Also love the superhero analogy, though I’d say they’re all Batman because we all know Batman is the best ; )

        • Four Touchdowns

          No way, I’m a Superman guy! 😀

          Give me the choice between flying and super strength or having a really cool car and toolbelt, and I know which one I’m picking, LOL.

        • Four Touchdowns

          The other analogy I like to use is — what if they cloned Tom Brady? He’s a perfect double of the original in every way… and they put the cloned Brady on the Browns.

          If the Patriots and Browns played each other, who do you think is more likely to win?

          • Yep that’s a good one too.

          • Richie

            What if they also cloned Belichick and had him coach the Browns. Then the Brady Browns played the Patriots 100 times and Belichick Browns played the Patriots 100 times.

            Does the Brady Browns or the Belichick Browns win more games?

            I still lean towards Belichick being more of a magician than Brady. However, I think Belichick would have a tough time coaching against himself. His strategies probably wouldn’t work against himself. But, Brady should still be able to make plays against the Patriots and maybe steal a few more games than Belichick would.

            • Four Touchdowns

              LOL, then the Belichick with the better roster would probably win. His brilliance is fitting players to his gameplans — the better the players, the more freedom he’d have in his schemes.

            • Tom

              Belichick would have a hard time doing his Jedi Mind-Probe on himself…not sure what would happen here…

          • Tom

            This works pretty well…and the reason we know that using an analogy like this works is that if you did the same thing with Aaron Rodgers, the answer isn’t so clear.

            • Dave Archibald

              Supporting cast definitely affects Rodgers, too. The 2015 Packers finished 15th in points, 23rd in yards offensively.

    • Tom

      Agreed. And I think comparing him to Jeter or Kobe is pretty much right on. Total legends…all on great teams.

  • Richie

    From 2001-2016, the Patriots rank 2nd in points allowed, but 17th in yards allowed; a difference of 15 ranking spots.

    (Kansas City has a differential of 11 – but they rank 18th in points and 29th in yards. Chicago has a differential of 9 – 10th in points, 19th in yards.)

    At the other end, Buffalo ranks 26th in points and 16th in yards for a differential of 10.

    Even though some other teams have differentials that are pretty large; the fact that New England has the 2nd best points defense, but an average YARDS defense, sure seems like there is something in their scheme that is designed to prevent points, but not worry about yards. And, of course, this is really the way that defense should be designed because points is (are?) what matters.

    Anybody have any theories as to how a scheme (or personnel) could actually maximize points allowed, while not being as good against yards?

    • Richie

      Answering my own question a bit. It looks like starting field position might have something to do with it. I only went back a few years, but the Patriots consistently pin their opponents back about 3 yards more than average. If you assume 175 drives per year, that works out to (incredibly) 8,000 more yards that their defense is able to give up. If you subtract those 8,000 yards, then the defense ranks 2nd in yards allowed. (But I only subtracted the yards from New England, so surely other teams would benefit as well.)

      • Richie

        Incidentally, the Chiefs had the largest differential in 2016. They were 7th in “points percentage” but 21st in “yards percentage”.

        (Sorry to hijack this thread.)

        • Four Touchdowns

          Hijack away, I love this kinda stuff!

        • Dave Archibald

          They’re a little more straightforward than NE – KC’s D led the league with 33 turnovers.

      • Tom

        This is good stuff Richie. Not surprised to see that starting field position plays a part. The more we look at this stuff, the more we uncover what the hell is going on with the Patriots – basically, they are consistently very good at EVERYTHING. Think about it, in James’ last post he provided the info that showed that Brady has a starting field position edge of about 2 or 3 yards over the other three guys. This amounts to around a point per game. Sure, not a lot. But now we have the fact that the Pats gain a few yards on defense as well, which, for simplicity sake we can is about a point per game. The Pats could be gaining about 2 points per game on field position alone. This, in my mind, is a substantial advantage, it’s almost like a built-in home field advantage.

        • Richie

          Yes, 1 point per game on each side of the ball would have been worth 0.6 pythagorean wins to the 2016 Patriots.

          0.6 wins for 16 years is nearly an additional 10 wins for the 2001-2016 Patriots. That’s got to have a significant impact on expected Super Bowl wins over that time period.

          But how the heck do they do it? There have been multiple coaches who have gone to other teams. Were they not able to implement these schemes? Are there certain schemes/plans that Belichick uses that he doesn’t even share with his assistant coaches?

          Is it all Ernie Adams?!?!?

          • Four Touchdowns

            Perhaps Belichick isn’t just great at finding good players for less money, perhaps he’s also got some ability to evaluate how these players will *play together* within his scheme. Who can cover up whose mistakes?

            Also, maybe he understands matchups with opposing teams better and adjusts gameplans better than most, who may be in love with their schemes?

            He also seems to put a premium on discipline more than physical abilities — see how they cut Jamie Collins, a really good player, for not doing what he wanted him to do?

            • Richie

              Yeah, it takes gigantic balls and comfort in your job security to get rid of a player like Jamie Collins during a season that you have a chance to go to the Super Bowl.

              Even if one of Belichick’s disciples sees Belichick do something like that, and understands why he did it – would they have the confidence to do it themselves, given the opportunity?

              Jon Gruden tried punting Keyshawn Johnson in 2003. But the Bucs were struggling at the time (4-6 with Keyshawn), and it didn’t seem to help much (they went 3-3 without him).

              • Tom

                Yes, massive, gargantuan balls are needed in a situation like that (I’m sorry, I could not help myself).

      • Dave Archibald

        NE also doesn’t give up a lot of big plays. There aren’t a lot of 2-play, 66-yard TD drives against the Patriots. Perhaps “bend but don’t break” shouldn’t be looked at as doing a particularly effective job in the red zone, but doing a particularly good job preventing long scores outside the red zone.

        • Richie

          Interesting (to me) numbers on that:

          In 2016, New England gave up 17 plays of 30 yards or longer. That was 5th best in the league. The Falcons were best with 14 long plays. (Oakland was worst, with 35 plays.)

          The correlation between “Long Plays” and points allowed in 2016 was 0.46. Interestingly, the correlation was also .46 between Long Plays and Total Yards allowed. Those aren’t super strong correlations.

          Some outliers:
          The Giants were the anti-Patriots, and ranked 26th in Long Plays allowed, yet ranked 2nd in points allowed and 10th in yards allowed.

          Philadelphia ranked 30th in Long Plays, but 10th in points and 12th in yards.

          Atlanta was at the other end. There were 1st in Long Plays, but 27th in points and 25th in yards.

          I find that fascinating. The two Super Bowl teams were both very good at avoiding long plays. For the Patriots, it resulted in low points and yards allowed; but for the Falcons it resulted in high points and high yards.

          And then there was the 49ers. 31st in long plays; last in yards and points.

          • Four Touchdowns

            Can you look up short plays? Like plays inside the redzone? I wonder how they do in that space.

            • Richie

              What would you want to see? The percentage of red zone plays that DON’T turn into touchdowns?

              • Four Touchdowns

                Yeah, just how many points they’ve allowed in the redzone. Maybe I can figure out how to look it up.

          • Dave Archibald

            That is interesting. I think the Falcons suffered from playing from ahead so much. They faced 655 pass attempts, the most in the NFL. In the second halves of games, they allowed scores on 44.2% of drives (3rd-worst in the NFL) and their 31.2% TD rate was NFL worst. But in the first half those figures were 37.2% and 23.4%, respectively – pretty much middle of the pack.

            Of course, the Patriots were ahead a lot, too, and their D didn’t drop off as much, but it still did to some degree.

          • garymrosen

            “And then there was the 49ers. 31st in long plays; last in yards and points”

            Another great “achievement” for the Yorks, the worst sports team owners since Harry Frazee.

    • Four Touchdowns

      Where did you get the total points and yards numbers? Or did you add it up the hard way?

      • Richie

        I just used the PFR Play Finder: http://pfref.com/tiny/hrZJz

        (I also discovered that PFR has added a nice “Export to Excel Workbook” feature, so I don’t have to export CSV any more.)

        • eag97a

          Oh man you are a lifesaver. This will make it so much easier getting data off pfr. Cheers!

    • Paul

      Gridfe had a pretty good explanation. The patriots have consistently been great at avoiding offensive turnovers and have ranked very high in special teams, so they often put the opponent in bad field position and limit putting their defense in bad situations.

  • Paul

    The nfl is a buisness. With the salary cap, its not much of a coincidence that bradys superbowls wins have come at the beginning of his career, and at the end, when he hasnt been paid in the top 3 qb salaries. Part of supporting your team and bradys greatness is how he takes less money so the team can be better. When he took up a huge amount of cap with his megastar salaries they didnt win superbowls. Now hes paid what backups are. The patriots obviously will have better teams when the can afford to pay better players.

    Drew brees is terrific, but he is paid like a megastar, so i dont think its resonable to blame the saints 100% for not giving him as good of support as the patriots do brady. They cant afford to.

    • Four Touchdowns

      Probably not Brady support, but surely they can do better than they have been, LOL.

    • Terry

      Exactly right. Support to the QB is inversely proportional to their salary cap hit. The less money you spend on the QB, the More you have available to spend on the rest of the team.

  • Matthew Penney

    You have Brady led Patriot teams with a rushing yards per attempt ranking of between 11-12, where is this data coming from? Is it just from regular season play, or does it include post season. I just cannot understand where that number comes from since historically Brady’s patriots have been terrible in YPA.
    Year

  • Matthew Penney

    Where did you get 11.7 from in average YPA ranking? Using Pro Football Reference for the regular season he is on average rank 19.375. Historically Brady has had some pretty bad rushing efficiency from his teammates. Gonna have to know where you are getting your data from before I believe any of it.

    • Four Touchdowns

      I pulled it from ESPN and PFR — I strived to be accurate but I am only human. If a season is incorrect, let me know.

      https://uploads.disquscdn.com/images/0709e1db66d4cc31abb81a45ad6285c8d59c82a66405a44ba6cbffa0accdc148.jpg

      • Tom Goodrich

        Looking at PFR for ’01-’16 but without ’08 I get these as New England’s rushing Y/A rankings: 24, 27, 30, 8, 30, 18, 14, 21, 10, 24, 17, 9, 22, 29, 25

        Without bothering to average those out I feel VERY confident that the numbers don’t average to 11.7 or anything close to it, and 19.375 as an average doesn’t sound far off.

        I also get different results as well for the Colts: you have the 1998 and 1999 Colts as ranked 14th and 20th in rushing Y/A, PFR has them ranked 17th and 16th.

        Perhaps you took the rushing Y/A from ESPN rather than PFR and they count slightly differently that PFR does (perhaps one site excludes kneeldowns and the other doesn’t).

        That could explain some small differences, but it’s extremely difficult to imagine that that’s all that’s at work here: you have the 2003 Patriots with the 3rd best Rushing YPA in the league, PFR has them with the 30th.

        • Four Touchdowns

          Clearly I must have been looking at the wrong fields or something because I just checked ESPN and they don’t match my numbers.

          My apologies… ugh.

      • Matthew Penney

        It appears your YPA numbers are actually reversed, like in 2001 the brady-led patriots had a YPA of 3.8 putting them 23rd overall. On your sheet you have their ranking as 9th. The same is for Manning, Brees, and Rodgers. So instead of having the best YPA from his RB’s, Brady has historically had the WORST YPA from his RB corps.

        • Four Touchdowns

          Ah, that must be it. I really apologize for this, I did my best to be accurate but I must have double-clicked or did something wrong when I sorted the ranking tables.

          • Reagansmash

            No big deal, Credit for acknowledging the mistake. You would be surprised how many ppl wouldn’t.

        • Four Touchdowns

          And THANK YOU for noticing and letting me know.

  • Tom

    More info for this great post…

    I looked at Special Teams DVOA for each of year of these guys’ careers, below shows the average ST ranks:

    Brady, 8.4
    Rodgers, 20.0
    Manning, 21.7
    Brees, 22.1

    That is ridiculous. Throughout their careers, Manning, Rodgers and Brees have been on teams with below average special teams…and not “slightly” below average, *solidly* below average. Manning has had three seasons – 2003, 2014 and 2015 – in which his special teams were above average, he’s never had a Top 10 special teams unit. Brees has never had a Top 10 unit, Rodgers had one in 2011 when the GB special teams ranked #8.

    How about Brady? During Brady’s tenure, the Patriots have NEVER been below average in special teams and they’ve been in the Top 10 NINE times. This agrees with what James has been showing, so I think we can feel comfortable in saying that Brady has enjoyed a significant advantage in this area (special teams) over the course of his entire career, something the other three have not had, and it may not even be close.

    How much of an affect this has on him winning we can’t say, but I’m of the opinion that, over the long haul, it’s been significant.

    • eag97a

      You have to factor in how much Bradys’ play and his offense affect ST performance. It goes both ways and is the heart of what BB terms as complementary football. Units playing well influence and help other units play well as well. Good ST performance might have had a significant impact on TBs’ performance over the years but its equally true that TB and offense has had significant influence and good impact on the ST and defensive performance of the Pats as well.

      • Tom

        Totally agree. But shouldn’t we see the same type of affect for a Manning led offense? In 2006, the Colts fielded a historic offense, but their special teams were ranked 25. If there was just a few ranking points difference here and there, fine. But look at those numbers…they’re not even close. Brady has had a significant advantage in this area and I don’t think it’s because the Pats offense is good…we should see at least somewhat similar results with Green Bay, New Orleans, etc…and we’re not.

        • Richie

          Specifically in 2006, the main contributor to ST DVOA difference between New England and Indianapolis is kickoffs and kick returns.

          I know that Tom Brady has magical powers, but I refuse to believe that he had any impact on his team’s ability to kickoff or return kicks.

          (Indianapolis was actually much better at FG/XP than New England that year.)

          • eag97a

            He and the offense has an impact on punts and FGs. And it doesn’t need magic.

            • Tom

              And again – then why aren’t we seeing this with other great offenses? Do you have an answer to why the great Saints offenses don’t also show this affect? Or the great Colts offenses? Is there something Brees and Manning are doing wrong?

            • Richie

              How does he impact a punt, when he’s on the sideline, in a way that is measurable by DVOA?

              And he couldn’t figure out how to help the field goal team in 2006. That year the Patriots were below average at XP/FG and the Colts were above average.

              http://www.footballoutsiders.com/stats/teamst2006

        • eag97a

          It was a historic offense with regards to scoring but maybe not when it comes to clock control? If it was historic then they should have gone undefeated? They had 4 losses in which one was against the Jags 44-17. For me a definition of a historic offense is a juggernaut with a record of 16-0 or at worst 14-0 like the 07 Pats, 89 Niners etc. If you had a good scoring team but struggled with a few matchups and had more than 2 losses and few shellackings it isn’t historic IMO.

          • Tom

            Then take the word historic out of it. Let’s just say the 2006 were rated #1 in DVOA and leave it at that. And now…let me get this straight…the Colts were good at scoring, but not too good at clock control? Is that the secret?

            • eag97a

              Take a good look on my statement and you can see the question mark, I’m putting it out there that scoring is not the end all and be all of football games, it’s winning and after scoring points and denying the opponents from scoring points the next most important thing is clock control and protecting leads or denying possession. I’m not a coach but these things should be evident watching the game.

              • Richie

                That’s not what this branch of the thread has been about. We’ve been talking special teams performance – specifically special teams DVOA.

                • eag97a

                  The conclusion was TB was helped by team support (ST and defense) more than the others. That conclusion can be tentative or even wrong if it’s proven that the TB and his offense has helped his ST and defense more than the others. That will invalidate or at the very least question this articles’ conclusion.

                  • Tom

                    I’m just lost as to how to answer this…I don’t know where to begin. Yes, that’s the conclusion because it makes sense. Great special teams means better field position, for one. We’ve already demonstrated that Brady has at least a few yards edge in starting field position. Does that make sense to you? He is closer to the end zone, that’s a significant advantage, can you accept that? Now how in the hell does the offense help the special teams in as meaningful a way as that? And AGAIN, shouldn’t we see this in other teams?

                    • eag97a

                      The offense can help the ST in many ways. One example is running safe plays during 3rd and long to shorten the opponents’ punt return which helps the defense pin the opposing offense which leads to turnovers and short possessions. The point is the offense can help the STs’ performance indirectly by making their jobs easier which in turn helps the ST performance. It can be a virtuous cycle except for big plays happening and luck. BB and the Pats have always stressed preparation and execution to minimize the effect of opponent big plays and luck impacting their own performance. Of course its not perfect and things happen but WE SEE this happen to other teams in the reverse because sometimes (not everytime) their offense doesn’t help their ST and defense a lot when they score quickly, when they don’t drain the clock properly or when they pass up on safe plays on 3rd and long to help the ST and maximize the potential number of possessions still available etc. That’s the importance of proper game planning and even then the Pats haven’t won everything since the league is very much parity driven unlike the other sports leagues. Once again I respect the way your reductive logic and compartmentalization evaluate football, I just subscribe to another way of looking at things.

                    • Tom

                      “One example is running safe plays during 3rd and long to shorten the opponents’ punt return which helps the defense pin the opposing offense which leads to turnovers and short possessions.”

                      Great point. Should the punter be pretty good at punting so that he can pin the other teams offense? And what evidence do you have that this bad field position leads to turnovers?

                      “…their offense doesn’t help their ST and defense a lot when they score quickly…”

                      And when the Pats score quickly, it’s the right time to do so? What can we make of the Pats in 2007 when they were scoring so many points people couldn’t keep track of it? Where they not helping their special teams unit that year?

                      “Once again I respect the way your reductive logic and compartmentalization evaluate football, I just subscribe to another way of looking at things.”

                      I’m not being reductive at all, and I’m fine with the way you view things. Do you have any other evidence to back up what you’re saying besides your examples? Any numbers to show how many times the Pats have run safe plays on 3rd and long compared to other teams? Anything at all to help prove what you’re saying? I’d like to just take your word for it, but it would help if there was at least something to demonstrate what you’re saying.

              • Tom

                See Richie’s comment below. We’re not talking about all the elements of the game. We’re talking about ST performance and how you’re convinced that we need to give Tom Brady credit for how good the Pats special teams are.

                • eag97a

                  See my comment above about the difference of our approached, mine is holistic and inclusive, yours and the others reductive and compartmentalizes. Neither approach is better just different.

                • eag97a

                  I said qbs and their offenses not Tom Brady alone. My point is qbs and the offense have a significant effect and impact on ST and defense. Stating Tom Brady alone is inflammatory and I have always argued and maintained that 3 phases of the game influence each other.

      • Richie

        I don’t think Brady has any affect on ST DVOA. (Unless I am misunderstanding how DVOA works.)

        DVOA is a comparison of how a team compares in a given situation compared to league average. So, Brady putting his punter or kicker in a “good” position may help with the result; but it’s not going to help them in their performance compared to league average for the same position.

        • eag97a

          If QBs help with the result then they help with the performance. DVOA might not capture this but it doesn’t mean that there is no significant effect. DVOA is a good metric but it doesn’t capture everything.

          • Tom

            Agreed. But the question is: why aren’t we seeing this affect with other offenses? Does Tom Brady have magical special teams powers? Can’t we just say that Belichick is a brilliant coach who recognizes that special teams are important and it has helped his team win games?

            Peyton Manning consistently led offenses that were at the top in DVOA (and other standard metrics), and yet his special teams are consistently below average. We’re not talking a few times, we’re talking always. What’s the difference here? Can we look at other teams with great special teams and point to their great offense? What about the Bears in the 2000’s with Dave Toub as ST coach and Devon Hester? Should we look at that Bears offense? How is it that when Toub went to KC, their special teams play skyrocketed? They were #22 in 2012 and then went to #1 in 2013 when Toub got there.

            I think I’d rather just give this one to Belichick and leave Brady out of it.

          • Richie

            What I meant was that if Brady can help set up an easier field goal, then the 3 points will be more likely than a QB who leaves his kicker to hit on a 53 yarder.

            But DVOA doesn’t care. It compares the 24-yarder that Brady set up to the 24 yarder that every other QB set up.

            Yes, Brady’s team has a higher win probability because he got 30 yards closer to the end zone. But DVOA is going to isolate his kicking team’s field goal performance from what the offense did.

            And that’s Tom’s point. Brady’s special teams have (apparently) helped his teams out much more than the other guys’ teams.

            • eag97a

              And my point is does anybody out there measured the contribution of the offense to ST and defense? It’s all interlocking and complementary. If ST and defense has helped TB through the years more than the other elite qbs then its also possible TB has helped his ST and defense more than the other elite qbs. We always forget the other side of the equation.

              • Richie

                Do you have any theories as to what Tom Brady may have done that made Stephen Gostkowksi significantly better at kickoffs in 2006 than Adam Vinatieri for the Colts in 2006?

                Or something he did that made Laurence Maroney (and other Patriots) significantly better than Terrence Wilkins (and other Colts) at returning kickoffs?

                • Paul

                  Its simple, brady has that ‘it’ factor. Certainly alot of ‘it’. Far more then the colts, they have less ‘it’. But indy certainly has more ‘it’ than the browns, they have so little ‘it’, its more like ‘not’.

            • Paul

              Does dvoa account for how impactful something is? I realize special teams is an often underappreciated aspect of team play, but does dvoa have a division of how much each unit provides. Im pretty sure defense and offense are far more important then special teams, as they account for the majority of plays a team runs.

              • Tom

                I don’t think so, at least as far as I’ve seen. As far as each unit’s overall impact, this is an interesting read that kind of addresses what you’re talking about:

                http://www.footballperspective.com/scoring-is-60-of-the-game/

              • Richie

                Yeah, football outsiders has a ratio. I think it is about 45% offense, 45% defense and 10% special teams.

      • Tom

        Yes, this is at the heart of what Belichick is saying, complementary football and he is absolutely correct and it has led to many, many wins. But unless Brady has some kind of “special teams helping magic”, the Pats offense isn’t having that much of an affect on ST. No…that unit is good because BB makes it a priority and he’s right in doing so.

        • eag97a

          See my comment below. A qb and offense consistently scoring and executing long drives will help both ST and defense with regards to field position, clock control etc. This are the parts of football that cannot be reliably be quantified and I disagree that we can just dismiss the effects as trivial. Someone has to mathematically prove it first before I change my opinion on this. It doesn’t make sense to put walls between 3 phases of the game and say they don’t influence each other to some significant degree. Doesn’t make sense IMO.

          • Tom

            Again, I totally agree, the units are connected. But as I’ve asked a few times, if we are in any way to believe that the Pats offense is having this significant impact (and it must be significant otherwise, you wouldn’t mention it) on the Pats special teams, simply because they are good, shouldn’t we see it with other teams? Is there a way that Manning with the Colts just wasn’t doing it right? Is Drew Brees just not understanding that his offense, which consistently plays at a high level, year after year, was also supposed to be improving his special teams? Can you explain that?

            • eag97a

              Probably from a game plan or execution perspective. PM and DB might not be going for the safe run or short pass from 3rd and long (like 3 and 20) to shorten the 4th down punt and they go to long incomplete passes for a boom and bust result? This is all speculation but how many times have we seen TB and Pats run a draw play from 3rd and long? I’m pretty sure DB and PM have mastered their offenses but it could be a question of offensive philosophy and/or play call preference by the OC and the qb? This question might be ultimately unanswerable but dismissing the effect of the offense on ST and defense isn’t really a good way to analyze a team game like football properly IMO.

              • Tom

                Yes, we know the units affect each other, I’m fine with that, we don’t need to discuss that. But there is just no evidence to suggest that the offense is having a *significant impact* on special teams, none. A little? Sure, that makes sense. But it’s not significant, because we’d see with other offenses…we’d actually notice that when an offense is doing well, the special teams do well, and I’m just not seeing. Your comment above suggests that you do think, somehow, that the Pats are doing something “special” with their offense that helps out special teams and I don’t buy it. That’s why Richie and I are using the word “magic”. Because somehow, other teams can’t figure out this special way of doing “safe runs” and “short 4th down punts” and “draw plays”. And somehow Manning and Brees just can’t figure out this secret…

                The reluctance to even suggest that Tom Brady isn’t the reason for every great thing that happens with the Pats is just simply mind-blowing. There’s just no way you’re going to budge on this…Brady MUST be somehow involved, even to the point where we have to give him credit for special teams play.

                • eag97a

                  Sigh…. we are going around in circles. I said it’s all complementary and you agreed with it and then the next statement you dismiss it as trivial. I have always noted that it’s not only Brady but its him and the offense and I always use that clarification. It actually includes coaching and the scouting arm of the FO as well. I’m in the camp that stats are well and good but it won’t capture the intricacies of football otherwise we would have been able to predict with some regularity wins and losses and even statistical performance.

                  And saying that I said that TB is the reason for every great thing that happens with the Pats is outright disingenuous if not false. I never said anything of the sort and you can backtrack my statements. We have different ways of looking and evaluating, my way is holistic and includes the 3 phases of the game, yours and the others reductive and compartmentalizes the different phases. I’m not saying its wrong we are just different from an evaluation perspective.

                  • Tom

                    OK, you did not say that Tom Brady is the reason for the Pats being great, and to be particular, I didn’t say that you actually said that. But yes, I did suggest that you’re thinking that and it was unfair. So I apologize for suggesting that and I take it back.

                    And we’re not going in circles, at least I’m not. I agree that the units are complimentary, and I’m not compartmentalizing the various units, and I’m not contradicting myself on that. What we disagree on is HOW MUCH. I do not believe that the offense has a *SIGNIFICANT* impact on special teams play, I believe that the impact is MINIMAL and not worth noting. We see a HUGE difference between the special teams that Brady has had and the special teams that Manning and Brees has had, and I do not believe that it is because of something the Pats offense is doing…I just think that the Pats special teams, on their own and with MINIMAL impact from the offense, are very, very good. That is why we keep saying “magic” because it just doesn’t make plain sense that the Pats offense would *SIGNIFICANTLY* affect the special teams, but somehow, the Colts and Saints don’t have this power (again, if they did, we would see it). If you want to say that the Pats have great special teams because Belichick coaches wiht a more holistic approach, I agree 100%…but I don’t think that the offense is somehow intricately involved. The special teams are called “support” for a reason – they are “supporting” the offense, it’s not the other way around.

                    Now – I have provided evidence of why I think my opinion is right…you may not agree that that evidence proves anything, but I’ve at least provided something other than “it’s my opinion”. If you have any kind of evidence that supports what you are saying – namely, that the offense has a *SIGNIFICANT* impact on special teams play – then provide it. I’m alright if I’m wrong on this. Otherwise, we can agree to disagree and that’s cool too.

    • Dave Archibald

      I wouldn’t credit Brady for the success of New England’s special teams and defense, but from a roster construction standpoint the Patriots have chosen to invest more in these areas, sometimes (but not always) at the expense of Brady having the kind of weapons some of these other guys have had, particularly early in his career. This is especially true with respect to Manning, who had years of Harrison and Wayne and Clark and first-round running backs (not all of them good) and, accordingly, a top-heavy roster that sacrificed on defense and special teams. The Patriots have chosen to invest differently; we shouldn’t be surprised that they’ve been better in these areas

      • eag97a

        I feel the need to clarify things here. I’m not crediting TB for the Pats ST performance, all I’m saying is the Pats offense have a significant impact on ST and defense and the reverse is true as well, that ST and defense significantly impacts and affects the offense as well. People might misconstrue my position here.

        • Richie

          Do you have any theories as to how an offense can improve a team’s kickoff performance?

          • It can create the opportunity to have more kickoffs by scoring on a greater percentage of drives. Not much more than that.

            But…It can give FG units better position to make kicks. It can give the punt team better field position, which, in turn, should give the defense better field position.

            Then the defense can force a three and out, giving the punt return team better FP, which should give the offense a better shot at scoring. Or the D can allow a longer drive before a punt, which would be to the detriment of the return team and offense. Or the D could allow a FG attempt, which could either result in a score and kickoff (and meh field position) or a miss and great field position. Or the D can force a turnover and give the offense great FP, or even score themselves.

            The O can help the D, which will, ultimately, benefit the O again. And vice versa. They can both help ST, and ST can help both of them. Of course, as you mentioned, the O and D can’t do much to affect kickoffs, aside from increasing or decreasing the number of them.

            I hope that was as confusing to read as it was to write.

            • Tom

              Bryan – all of the units interact with each other and can help each other out. I’m in agreement with that, I would think most of us are. That being said, do you think that the offense has a *significant* impact on special teams play? In other words, in light of what you just described above, should we expect teams with excellent offenses to at least have “average” special teams? Is there an explanation as to why Manning’s, Brees’ and Rodgers’ special teams all rank in the low 20’s even the offenses led by those guys would be considered, for the most part, excellent? Are those offenses doing something wrong, that they can’t help they’re special teams more, year after year after year? My explanation is simple: those special teams units just weren’t/aren’t very good.

              • I don’t think a great offense can somehow make a 12% DVOA defense play like a -6% DVOA defense, but I do believe that an efficient offense can put that 12% DVOA defense in better position to succeed. On the other side of the ball, teams like the 2015 Panthers come to mind. They didn’t make the offense play better, but they put them in situations that didn’t require a high level of play in order to score.

            • Richie

              More kickoffs does not mean better success rate on kickoffs. The Patriots had a better rate in 2006.

              Again, DVOA isolates for field position, so the Patriots are punting better than other teams from the same spot.

              How does offense contribute to those things?

              • Tom

                The whole thing seems simple to me: the Patriots have great special teams and if the their offense has anything to do with that, the effect is minimal. We’ve seen great ST on teams with mediocre offenses (the Chiefs in 2013) and as I’ve shown, we’ve seen incredible offenses with below average special teams. Interaction, yes. Significant interaction? I’m thinking no, until I see some numbers. Heck, something as simple as seeing a correlation between teams’ offensive DVOA and special teams DVOA would help I think.

              • I don’t believe I stated or implied that more kickoffs means better kickoffs.

                I also don’t believe the offense makes the defense or special teams better, and I don’t think I said so. I simply believe that a unit can make things easier or harder on other units, and thus make them look better than they actually are, without ever actually improving the quality of play of any of those units.

                • Richie

                  Apparently I misunderstood your comment.

                  I agree with your second paragraph.

                  • A misunderstanding in a comment thread? Get outta here, you!

            • eag97a

              My point exactly. Thanks Bryan for pointing that out. A Biology teacher one taught me that Characteritics =Genes + Environment + Synergistic effects/Interaction of genes and enviironment. If I have to write out an equation about football it would be something like this Team performance = offense + defense + ST + coaching + FO/scouting + Synergistic effects/interaction of other variables.

              • Tom

                Yes, I agree with that. All of those things interact with each other and affect each other. You could add opponent, weather, all kinds of stuff.

                • eag97a

                  Our only difference is the significance of the impacts of the various phases of the game to each other. To you its minimal, to me its a bit stronger. Just a question of degree.

                  • Tom

                    That is exactly correct, that’s where we disagree.

                  • Tom

                    Alright, let’s just wrap this up. We both agree that the various units interact, I’ve provided some evidence of why I think it’s minimal between the offense and special teams, I haven’t seen any from you, but admittedly, this is a hard thing to untangle. So let’s end this this way (because I need to go to bed):

                    I’m thinking that the special teams performance is mostly due to the coaching, the players, front office, etc., etc. that is specific to the special teams unit. I’m going to say that accounts for 92% of why they are good. I’ll say that offense is contributing 4% and defense another 4%. Basically, I’m saying that the Pats special teams are good because they’re just good, and the greatness of the Brady offense is having little affect.

                    On the other hand, I do believe that special teams have a greater impact on an offense. For that, I’m going to say that an offense’s performance is 82% due to how good it is intrinsically, and 10% credit goes to ST and 8% to defense. So here I’m saying that Brady and the offense deserve most of the credit for why they’re so good, but special teams, and to a lesser degree their defense has had a modest impact.

                    This has the general effect of me saying if two offenses are rated the same, I might give the edge to the offense that had the “worse” special teams.

                    Now you tell me what your percentages are, then we’ll know where we stand and just end it like that? Cool?

                • Tom

                  (mistaken post, should have replied to eag97a, see below)

                  • eag97a

                    Gun to my head I’d say the offense can contribute 10% or more, same with the defense. There are 6 different special teams (punt, punt return, kickoff, kickoff return, FG, FG defense). I must stress these numbers have no basis at all and are just off the top of my head.

                    • Tom

                      Alright, well, we’re not that far off then. I was thinking you’d be more in the 30% range! And yes, I’m guessing on my numbers as well. Until next time, I’m out for now!

                    • eag97a

                      When you have time, check out the discussion here because some of the contributors have thoughtful and insightful comments; https://www.reddit.com/r/nfl/comments/6bwrkd/support_for_manning_brees_brady_and_rodgers_part/. Ciao

                    • Richie

                      I don’t see anything in there talking about the impact of offense on special teams…?

                    • eag97a

                      Read again and take particular attention on PureOrangeJuches’ comments. They are hidden because the guy he is replying to is downvoted. It’s a long thread you might have overlooked it.

                    • Richie

                      OK, I looked again. I still can’t find anybody giving any specific examples of how a quarterback/offense is going to affect his team’s kickoff and kick return teams.

                    • Four Touchdowns

                      That was a good thread, thanks for sharing that. I was avoiding reddits and other things like that because people can get so nasty and personal.

                    • Richie

                      I think 10% is still high. (Again, I need to stress that I am talking about the DVOA of special teams, which neutralizes for similar game situations.) I think it’s somewhere between 1% and 0%.

                      Now, going back to the 2006 DVOA that Tom originally brought up. The Patriots had +3.2% ST DVOA. The Colts had -4.3%. Even if we assume the offense has a 10% effect, and we assume that New England’s was good 10% and Indianapolis’ was bad 10%, that would bring their DVOAs to +2.88 and -3.87. This is still a large difference.

                      New England regularly has good DVOA on special teams (and VERY good DVOA on kickoffs). Indianapolis usually does not.

                      This is part of the reason that the Patriots regularly have a 2-4 yard per drive advantage in starting field goal year after year.

        • Tom

          No one is misconstruing anything. You think the Pats offense has a significant impact on their special teams play and I, for one, think that the Pats offense has a MINIMAL impact on their special teams. And while I wouldn’t say that special teams has a SIGNIFICANT impact on the offense – because if they did, we would see all teams with great special teams have great offenses – I do believe that impact from special teams on the offense is more significant than the reverse.

          • eag97a

            Impact can be positive or negative. Other great offenses might not have positive impact on their ST because they might not be constructed to help their ST/Defense. That is one good reason why good offenses by TRADITIONAL METRICS don’t necessarily have great ST/Defenses. The Pats might have a different definition of a great offense, they might define it as an offense that helps out their ST/Defense much more than the others etc. There are just loads of possibilities why there isn’t much correlation between the two.

            • Tom

              I’m not using “traditional metrics”, I’m using DVOA, which isn’t traditional…and I chose DVOA specifically because they have one of the only comprehensive ways (that I’ve seen) of measuring a teams’ ST performance.

              I get the gist of what you’re saying, I totally get the interaction thing and I agree that Belichick must preach this whole approach to the team. But I disagree that their offense would be defined as one “that helps out their ST/Defense”. Their offense is designed to score points and to score them as often and as quickly as possible, and there probably hasn’t been another offense in history that has been so good at doing that consistently from year to year.

              Yes, there are probably a lot of possibilities as to why there wouldn’t be a correlation, between a great offense and great special teams, but don’t you think *any* kind of correlation would at least somewhat bolster your position? If there is no correlation or even a negative correlation, do we just ignore that? Should we even bother looking at any numbers and just kind of go on our anecdotal thoughts of what we think happens? I don’t have all the answers, but I’ve at least provided evidence of why I think the way I do.

              • eag97a

                The evidence is the NFL results. Having a great offense doesn’t necessarily translate to wins and championships. DVOA is designed to compare units to the same units of other teams for baselining and measuring performance above or below average. It wasn’t designed to measure interaction effects. Its a great tool but just another tool with its own set of limitations.

                • Tom

                  How is the the fact that teams with great offenses don’t always translate to wins evidence of what you’re trying to prove? How does that prove that offenses have a significant impact on special teams. When did we start disputing that great offenses always win championships? When did we start saying that DVOA is anything but a tool? I’ll ask again – do you have any evidence, anything at all, that supports your position that a team’s offensive performance has a significant impact on that team’s special team’s performance?

                  • eag97a

                    So I’m guessing you have iron-clad proof that offenses have minimal impact on ST/Defense? You said in one of your comments that its simple for you and credit good ST performance of the Pats to BB alone? I presume that you have incontestable proof of that assertion as well?

                    • Tom

                      Nope, I don’t have iron-clad proof, but I least offered something that we can look at – I showed you that great offenses are not always accompanied by great special teams. And those were solid examples with long time lines…Manning went for 17 years without a single ST unit in the Top 10, and only 3 times were they above average. His offenses have been great, by whatever the hell measure you want to use. If your position is right, than we should have seen some kind of interaction. Same with Brees…he is a HOF QB, the Saints offenses have been great. And yet, year after year, their special teams are crap. You yourself have said that these units have an interaction, and I agree that there is *some* interaction. What is it that Tom Brady and the Pats offense is doing to make to make their special teams so good? Not just anecdotal, “well, on 3rd down they play safe”, etc. Heck, I can just tell you, “Well, the 2015 Texans played it safe on 3rd down and their special teams stunk”.

                      And while I did credit the Pats ST greatness to Belichick, I meant more that his coaching style, his emphasis on the importance of that unit (which we both agree on) should be given the credit, not the offense. If we want to put numbers to it, I can take a guess: I believe that the coaching, front office and scouting and the skill and determination of the ST players themselves accounts for at very least 92% of the Pats ST success. At most the offense accounts for 4%, the defense 4%. What percentages would you give?

                      No, don’t have proof of Belichick’s input. Since I don’t believe that the Pats offense is responsible for the majority of the ST success, I’m just using common sense that it must be the coaching. Do you disagree?

                • Tom

                  Here is an example of evidence. Below is a link to an article where Chase discusses the various importance of the the three units (but mostly offense and defense). It proves neither of our points (at least I don’t think so, haven’t read it in a while, heck, it might prove your point), but is interesting:

                  http://www.footballperspective.com/scoring-is-60-of-the-game/

                  Or here’s another one that touches on a lot of stuff, but has stuff about the three units:

                  http://www.footballoutsiders.com/info/FO-basics

                  Have you found anything that says that special teams performance is heavily impacted by the performance of the offense?

      • Tom

        Exactly. The Patriots should be praised for their holistic approach to building their teams. They’ve emphasized each phase of the game and it has paid off tremendously.

    • Four Touchdowns

      I’m not sure you guys will get anywhere for this argument but just to provide some context since the PFR EP box scores break them out —

      Kick-Off EP

      Manning: -2.41
      Brady: -2.22
      Brees: -2.38
      Rodgers: -1.25

      Kick Return EP

      Manning: 1.16
      Brady: 2.04
      Brees: 1.84
      Rodgers: 0.18

      Punting EP

      Manning: 1.00
      Brady: 1.20
      Brees: 1.64
      Rodgers: 1.29

      Punt Return EP

      Manning: -1.80
      Brady: -1.25
      Brees: -1.82
      Rodgers: -1.65

      Field Goal / Extra Point EP

      Manning: 0.33
      Brady: 0.63
      Brees: -0.42
      Rodgers: -0.20

      So it looks like the numbers are similar enough over all but the big stat to take away is field goals.

      • Richie

        Did you add those up for every game? The only place I’ve found EP for special teams is the individual box scores.

        I just took a quick look at the individual kickoff EP for the Patriots in 2006. PFR gives the Patriots a negative EP for most games. This fascinates me, because Football Outsiders gives the Patriots a +4.2% DVOA on kickoffs for 2006. I know DVOA and EP are different calculations, but I would think they should correlate a bit, since they are both based on similar principals. The one difference is that DVOA will take opponent strength into consideration, but (I don’t think) EP does. But it doesn’t seem like it should make a ton of difference on kickoffs, because I assume (but could be wrong) that touchbacks are a large component of kickoff success, and shouldn’t be affected by opponent.

        Just looking at one game. Week 1, 2006 Buf at NE. PFR gives NE -0.34 EP for Special Teams. Football Outsiders gave NE +3.9% for Special Teams. Interesting.

        http://www.pro-football-reference.com/boxscores/200609100nwe.htm
        http://www.footballoutsiders.com/dvoa-ratings/2006/week-2-dvoa-ratings

        • Four Touchdowns

          I copied and pasted the EP chart from every game, yes. It was insanely tedious, which is why I get freaked when people find mistakes — I don’t want to have any credibility issues.

          I’m guessing both kick offs and punts inherently have negative EP because you’re giving the ball to the other team, which increases their odds of scoring. The lesser the negative EP, the better your team did with field position but unless the other team turns the ball over, I bet it’s impossible to have positive EP on kicks to the other team.

          This is all speculation, of course.

          • Richie

            A positive EP is possible if the receiving team starts deep in its own territory. In the linked game, when Buffalo started on its own 9, New England gained 0.38 EP from the kickoff.

            • Four Touchdowns

              Interesting — then I don’t know why FO and PFR EP don’t sync up.

              ???

  • Tom Goodrich

    Something is rotten with the data set here, and rotten enough that I don’t think it warrants just a small tweak to the article but probably a full rewrite.

    The author was kind enough to provide the table he used to compile the average ranks here: https://uploads.disquscdn.com/images/0709e1db66d4cc31abb81a45ad6285c8d59c82a66405a44ba6cbffa0accdc148.jpg?w=800&h=402

    To take two of many, many examples of the rotten data set:

    – the table indicates that the 2003 Patriots had the 3rd best rushing Y/A in the league. PFR has them listed as 30th, and their two lead RBs (182 and 178 carries, while no one else on the team had more than 50) averaged 3.5 and 3.6 Y/A… suggesting that the error here is not PFR’s.

    – the table indicates that the 2001 Colts had the 28th best rushing Y/A in the league. PFR has them listed as 4th with a 4.5 Y/A average. That year Rhodes rushed 233 times for a 4.7 Y/A and Edge rushed 151 times for a 4.4 Y/A… again suggesting that the error here isn’t PFRs.

    I admire the effort, and would love to read the article with corrected data, but until then, garbage in, garbage out.

    • Four Touchdowns

      I wonder if I somehow mixed up rushing defense or something? Damn, this is annoying.

      Chase, my apologies.

      Have you noticed any of the other ranking stats to be incorrect or just the rushing?

      • Tom Goodrich

        No worries, mistakes happen, and I admire the willingness to accept criticism and take a second look.

        It looks to me like the rushing YPA ranks may be backwards– that they may be correct but with the team ranked “1” having the WORST rushing Y/A and the team ranked last having the BEST rushing Y/A.

        At a quick glance based on the 2003 Chargers, it looks like the same may be true for the total rushing yards column as well– though it wasn’t clear to me that that column ended up impacting your article directly as I didn’t notice a “total rushing yards rank” row in your graphs, just a rushing yds/game row.

        Hopefully it’s just a matter of checking and correcting the datasets and you can somewhat revise the article based on that– but if so I’d strongly urge you to start from a clean slate when making conclusions, as the differences may be significant enough to warrant different ones.

        • Four Touchdowns
          • Tom Goodrich

            TL;DR: Better, but still way off in a place or two for some reason.

            I have, for PFR’s rush YPA for New England (the only team for which I checked a full column worth of data): 24, 27, 30, 18, 30, 18, 14, 21, 10, 24, 17, 9, 22, 29, 25.

            One or two of those could be an error of me reading the wrong column (in typing this up I noticed I originally had “8” for 2004 when it was in fact “18”)

            That set of NWE rush YPAs is not quite the same as yours (often a rank or two different: e.g., you have NWE 2007 as 11th I have it as 14th) but it’s close enough that the difference could probably be explained if your numbers, e.g., don’t include QB rushing numbers or don’t count kneeldowns (or if PFR’s numbers don’t count kneeldowns but yours do).

            HOWEVER: on spot checking I’m still seeing at least one major oddity which makes me assume there are others: you have the 2006 Chargers with the 25th ranked YPA… I recall that as being a god mode LdT year and, indeed, PFR lists them as having the 4th best offensive rushing YPA that year with a 4.9 average… and that’s not a discrepancy that seems explainable by the inclusions/exclusion of QB rushing numbers or kneeldowns.

            • Four Touchdowns
              • Tom Goodrich

                Interestingly, then, we have a case of PFR on PFR violence: take a gander at the PFR team page for the 2001 Pats:

                http://www.pro-football-reference.com/teams/nwe/2001.htm

                There under the “Lg Rank Offense” row they list the Pats as 24th in rushing Y/A. Interestingly this is on the same number of rushing atts and yds, so it seems the numbers are being calculated the same way.

                At a quick guess your PFR looks like the wrong one: its numbers have the Pats as (1793/473) 3.791 Y/A and the Titans as (1794/468) 3.833 YPA… yet it ranks the Patriots (3.79 Y/A) 23rd and the Titans (3.83 YPA) 24th. I’m guessing there’s some glitch in their rank sort coding and that the ranks on the given team pages are the correct ones (which would suck as those would be a bitch to harvest).

                Since the rushes and attempts in your table seem to be correct there might be a way, if you’re clever enough with Excel, to rip the rush and attempt columns, calculate YPA within excel, and then rank the results. That also might be a colossal pain itself, though.

                • Four Touchdowns

                  Geez, that’s weird. Honestly, this was already WAY too much work as it is, so I won’t be doing all that — if there’s an error in my new chart, it belongs to PFR.

                  https://uploads.disquscdn.com/images/5f4a35fb20c20d75f0ece610e2c8b194ca7d6e32e6fa20da2a412edbafb9350c.jpg

                  • Tom Goodrich

                    I wouldn’t blame you for that. I’d recommend disclosing, though, that some of the ranks may be off slightly due to an error in PFR’s approach to table sorting and how it treats teams that round to the same decimal place.

                    Generally speaking that error probably doesn’t advantage any particular QB so it probably evens out. But definitely seems like best practice to just disclose that niggling detail within the revised article.

                    Not sure what’s up with that 2015 Green Bay discrepancy though, that doesn’t seem like just an issue of table sorting.

                    • Four Touchdowns

                      I mixed up the Saints and Packers numbers — they should be reversed for 2015.

                • Richie

                  Are you just talking about the difference between the Patriots being ranked 23rd or 24th in 2001?

                  Both the Patriots and Titans Y/A rounds to 3.8, so that’s probably why they are flip-flopping.

                  • Tom Goodrich

                    Yeah, though it’s weird as it’s not like they have all of the teams that round to the same number in alphabetical order which seems to suggest that the table is sorted based on additional, hidden decimals.

                    As I suggested to the author in another comment I think he can probably just reasonably proceed using the table ranks and including a brief note that there may be slight discrepancies due to quirks in PFR’s table sorting, but that it seems unlikely that they systematically benefit or hurt any particular QB.

          • Tom Goodrich

            I could easily be getting confused myself at this point but actually I think I notice at least one discrepancy that seems like it might be a significant error.

            2015 Packers are listed as having the 2015 rush YPA. I’m seeing 11th on both ESPN and PFR. Rodgers helps a team’s rush YPA significantly so if your numbers are ex-QB running that might explain it…. though I’m getting a 3.984 rush YPA when I remove Rodgers’ yards and attempts, which still seems a bit high for 28th (albeit since most other teams probably have their rush YPA hurt by including QB numbers, it’s not out of the question that removing QB numbers might make the difference).

            However this seems unlikely to be what’s going on here, as the 2016 packers are listed as 7th in rush YPA in your table and 7th in rush YPA in PFR, despite Rodgers again boosting their rush YPA considerably in 2016.

  • Zahidul Hasan

    Nice article keep on writing. Thank you, wonderful job!

    sports rings
    Gameday attire