How Many Wins Is Manning Really Worth?

It seems that's the big question lately, with Peyton Manning doubtful for Sunday as he rehabilitates his neck. The football punditeratti say there is no other individual player so critical to his team's success than Manning, and they're probably correct.

Win Probability Added (WPA) can't do everything, but one of the things it can do extremely well is tell us exactly how much of every win or loss was due to one component of a team. In this case, it can tell us how many wins the Peyton Manning passing game can account for. Although we can't really separate Manning from his blockers and receivers, we can nail down a hard number for the Colts passing game as a whole, of which Manning has been the central fixture.

Since 2000 (as far back as my data goes), Manning played in 176 regular-season games and accumulated a total of 43.0 WPA, for an average of 3.8 WPA per season. This equates to 0.24 WPA per game, which means that Manning (and his passing offense) would give an otherwise perfectly average team a 74% chance of winning a game. In other words, he would take an 8-win team and make them an 11.8-win team.

To put Manning's numbers in perspective, Tom Brady's career WPA/G is 0.16, still extremely good, but not quite up in the Manningsphere. Drew Brees and Ben Roethlisberger have both averaged 0.14 WPA/G. Kerry Collins, who is likely replacing Manning at least for week 1, has averaged 0.01 WPA/G (since 2000). In 2010, including the playoffs, Aaron Rodgers averaged 0.23 WPA/G.

In that same period, the Colts have gone 125-51--a 0.710 winning percentage. That's not far off from the 0.740 winning percentage that WPA told us to expect. It also suggests that the Colts have been an otherwise below-average team, although not nearly as bad as many pundits suggest.

In terms of points, Manning totaled 1,617 Expected Points Added (EPA) over those 176 games since 2000. That's a 9.1 net point advantage per game. This means that Manning and his passing offense generated just over 9 points per game more than an average NFL passing game would have. It appears the odds makers agree, dropping the point spread in the upcoming Colts-Texans game by about that amount when Manning was declared doubtful.

How many wins has your favorite QB accounted for?

  • Spread The Love
  • Digg This Post
  • Tweet This Post
  • Stumble This Post
  • Submit This Post To Delicious
  • Submit This Post To Reddit
  • Submit This Post To Mixx

16 Responses to “How Many Wins Is Manning Really Worth?”

  1. Anonymous says:

    Where has the "success rate" stat gone for defensive players? I found that very useful...

  2. Brian Burke says:

    It's been replaced with 'Success Count'. There is a forthcoming post to explain.

  3. Jim Glass says:

    No ill will to Peyton, but if he misses most of the season this will be an interesting natural experiment as to how much a QB is really worth -- how much of his performance numbers are his own, how much are provided to him by the rest of the team.

    Win Probability Added (WPA) can't do everything, but ... it can tell us how many wins the Peyton Manning passing game can account for

    Yes, but how much of this is Manning himself, and how much of this famous passing game has been attributable to the other 10 players (plus situation subs) on the field with him? Some of those guys were pretty darn good. This is what's interested me for a long time. If we have no way to determine that we don't really have QB passing ratings, we have team passing ratings for which the QB gets disproportionate personal credit or blame.

    In that same period, the Colts have gone 125-51--a 0.710 winning percentage. That's not far off from the 0.740 winning percentage that WPA told us to expect. It also suggests that the Colts have been an otherwise below-average team,

    Below average minus the entire passing game, not minus just Peyton. Of course over the years the Colts have loaded up the passing game positions to take best advantage of Peyton. I can't remember any other team built around a QB like the Colts have been. This was never a balanced team with the best available QB plugged in. Subtract all the salary cap spent on all those passing-game positions and it's hardly surprising the rest of the team was below .500 quality. Polian has said that's why they play the Cover-2 on defense, so they can use plug-in DBs at minimum cost to save money for the offense.

    The interesting question is, without Peyton how much would all that other passing game talent have produced? If, say, Sorgi had had to play a full season what would the result have been? We'll never know, but....

    After the Pats' 16-0 season, the following year Vegas over-under for them was 12.5 wins. By coincidence did an analysis that concluded the average starting QB was worth 2.3 pts per game more than the average backup -- which is worth one win, the difference between 9-7 and 8-8.

    A lot of people complained that was far too little, especially in the case of Brady. Well, Brady of course was far better than the average starter and his backup, who hadn't started since high school, was presumably somewhat less than the average substitute -- yet when Brady went down the Pats still won 11, only 1.5 less than the pre-injury line.'s approximate value metric tries to divvy up responsibily for offensive scoring among all players on the field and concludes "passers get 28.3% of the passing game points and receivers get the other 71.7%. Remember, there are a lot more people that have to split the receiving points."

    Manning is so good that 28.3% may be low for the Colts, loaded up in the other passing game positions or not -- but the numbers still shows that *a lot*, probably solidly most, of the production of a good passing game comes from players other than the QB.

    (Personally I believe that while a bad QB can kill an O all by himself via picks and sacks and stupid calls, a good one doesn't win games by himself but gets all the other O players contributing to do it -- that is, "diminishing returns" applies to QB play just as to so much else in the world. The improvement in the W-L record as the QB moves from "terrible" to "average" is a bigger than that which results as the QB moves from "average" to "great". But that's another story.)

    FWIW, the Football Outsiders season projection drops the Colts by 2.5 games (from 7.9 wins to 5.4) if Peyton is replaced with Collins. That's about 5 points per game off the offense.

  4. Anonymous says:

    The link at the end is pointing to localhost.

  5. Anonymous says:

    When looking at the difference between actual winning percentage vs. WPA-implied winning percentage, it's important to remember that in several seasons the Colts have, in essence, thrown games to protect their players. That would close your gap quite a bit.

  6. Joseph says:

    Brian, I agree with the comment before yours. Those week 17 losses are dragging down the win percentage.

  7. Brian Burke says:

    That's a good point. I agree.

  8. Anonymous says:

    I agree with Jim Glass. As posted in numerous other QB/Passing articles, you just can´t say how much it was team effort and how much QB.

    Bulger posted the same numbers as Warner in SL.
    Young in SF (a called total bust in TB) posted the same numbers as Montana.
    As Glass mentioned: Cassel (a total bust by numbers in KC, so i´d expect the same would happen to Brady in KC) posted the same numbers as Brady in NE. And so on.

    QB´s are wayyy overrated. The OL matters a lot, too. As long a QB is protected, (almost) every NFL-QB has the talent to pick a D apart. We are talking of ca. 100 QB´s out of ten thousands top tier players!

    Ok, may one has a quick release (like Marino), another is super accurate (like Montana), yet another is super strong armed (like Davey). But if those QB´s don´t have 10 highly skilled and working in sync players together, the best QB is useless.

    And Jim Glass mentioned all the money put in one player (Manning) certainly hurts in other areas.

    Without numbers (since it´s impossible to evaluate), i´d give Manning 1 more win above his (non bust) replacement. Because i also believe some wrong evaluated QB´s didn´t belong in the NFL (like Mirer, McNown, Akili Smith etc.).

    My good friend is Saints fan and always worried if Brees goes down. I just keep telling him Chase Daniel will do the same (good) job, as long this OL stays healthy and together. I am 100% sure about this.

    Greetings from Germany, Karl

  9. Anonymous says:

    "No ill will to Peyton, but if he misses most of the season this will be an interesting natural experiment as to how much a QB is really worth -- how much of his performance numbers are his own, how much are provided to him by the rest of the team."

    Not even then. Because we don´t know how the Colts would have gone with Manning under Center.
    The only way out would be like Collins starts the 1st 8 games, Manning the last 8.

    A similar "experiment" was made by FO in the 2003 (?) book. AFIR, replacement QB´s won one game less than the starters. But keep in mind that the replacements didn´t got 1st team reps when thrown in a game for the injured starter...

    Karl, Germany

  10. Jim Glass says:

    Thanks Karl, I'm glad somebody agrees with me about this, there aren't many of us around.

    Bulger ... Young ... Cassel ...

    My favorite example all-time is Earl Morrall. Long career as a journeyman bouncing around, then while the Colts' backup Unitas gets hurt, he goes in and becomes NFL MVP while leading the league in passing no matter how you measure it. Then back to the bench, can't get another starting job.

    Then he did it *again*, and even more so. After being cut out of the league Shula picked him up for $100 to be a backup for Miami. Griese went down and Earl QBed most of the 17-0 season, leading the league in passing rating and with more than 9 yards per attempt, and collecting a bucketfull of awards.

    Now I *do* think there was a big difference between Unitas and Morrall, and is between Brady and Cassell, etc. But it's really hard to quantify how much of one when *all* the passing numbers get attributed to the QB.

    For instance, Morrall in 1972 was very much the smart "game manager" QB, supported by a dominant defense and a running game that was the closest the NFL has ever seen to Nebraska's, so he was always in a position where he could carefully pick his spots to throw. And he had a HoF receiver, Paul Warfield, to throw to in those spots.

    The #2 QB in the NFL that year was Namath, who had a tremendous season, his best post-merger, while on a poorish team and throwing constantly all over the place trying to carry the team by himself.

    The larger perspective shows Namath was *much* the better QB that year. But you can't tell it by the numbers, because Morrall's are better. Which shows that the team -- top D, great running game, HoF receiver -- contributed *hugely* to his numbers. Teams do that for QBs, but it is impossible to quantify by how much.

    Which means QBs by default get hugely over-rated generally, IMHO. Not that they aren't the most important single player on the team, by a lot, tney are. But if a QB is worth three other starting players -- which is a huge valuation -- that's still only among 21 other starters and 30-odd regulars who make a difference.

    It also means QBs are seriously mis-rated in contemporary and historical rankings. One thing that stands out about the HoF QBs of the last 50 years is that they *all* played on good teams. Not one spent his career on a stinker team. That's not true of any other position -- or even in any other sport. One can make a list of HoF RBs, linemen, LBs, (MLB players, NBA players) who spent their careers on stinker teams. But no HoF QBs.

    I see no way to avoid concluding that QBs who get to the HoF do so in large part by getting credit due to the rest of their teams, and that excellent QBs on bad teams can't get that credit, so they are seriously under-recognized and never get to Canton except by buying a ticket.

    Until somebody with a supercomputer straightens this out, all we have is team passing records, not QB passing records, IMHO, FWIW.

  11. Anonymous says:

    "Until somebody with a supercomputer straightens this out, all we have is team passing records, not QB passing records, IMHO, FWIW."

    Exactly the same way i see it. All the good work Brian does with EPA/WPA etc.; i "only" read it as team efficiency stats.

    And i agree that QB is the most important position in Football. I mean if i give Manning 1 game above his "virtual" replacements, thats a lot for one player. No doubt about it.

    Karl, Germany

  12. Anonymous says:

    By losing Manning aren't the Colts also losing a big part of their offensive coordinator as well as their quarterback? It is hard to imagine another player calling audibles like Manning. The pacing of the offense and play selection will be different with a back-up. That is a much bigger loss than a quarterback who merely executes the plays called by someone else.

  13. Anonymous says:

    Manning is a great QB and OC combined into one. Those adjustments at the line, they killed Belichek over and over.

    Unfortunately he pretty much wasted his potential with incompetent timid coaches. One of those is in a studio booth giving poor analysis of games.

    One could speculate that had Belichek and Manning paired up, they would have done far greater things over the last decade or so.

  14. Anonymous says:

    "One could speculate that had Belichek and Manning paired up, they would have done far greater things over the last decade or so."

    Like frequently under-perform in the playoffs?

  15. Andy says:

    Would be interesting to do a followup now that the season is over and the Colts went 2-14.

  16. Anonymous says:

    ^Agree. Adter watching Indy all these years, my take:
    -Clear there was some transitional QB issues. With Orlovsky + camp under his belt, my feeling was they would have been a 6-10 team last year.
    -Colts won 10 games the year before, but what hasn't been pubbed until recently was that Manning had the nerve thing last year too and he struggled by his standards. Their decline from a 12-13 win team to 10 wins is due in part to that. Call them 11-5 with a healthy Manning last year.

    So that's a 5 win decline to a replacement level QB and probably 4 wins difference between an adequate starter and Manning, which is a bit higher than the article. What often gets underestimated is just how integral Manning was to personnel/style selection on both sides of the ball for Indy.

    -receivers: premium on hands and work ethic--not athleticism. Manning teaches you the routes and drills them into you via repetition. Easier to find these guys than a supremely naturally gifted receiver. No coincidence the exeption to the receiver corps had a good year last year: Garcon.
    -o line: always poor run blockers. Never got Manning sacked much, but I think we see how deficient they are now. Quick release masked the weaknesses.

    -Defense: undersized and built to play w/ a lead Manning gives them. Edge rushers on obvious passing downs (more of them with Manning) + weak secondary that gets masked thanks to DE QB pressure and obvious passing downs allowing Colts to play pass more. In a more balanced scenario, the secondary was burned and the D suffered. Also didn't help that Colts O had trouble sustaining drives to give the D a rest.

    My takeaway is the Peyton Manning is probably more important to his team than any QB in the "modern" era (Unitas or so on). This is different than saying he is the best QB in that era (which he may or may not be). The difference is the level of influence his skill set had on the selection of his teammates. His teammates have excellent skills at extremely specific things that Manning is good enough to hone in on. Other QBs tend to have teammates are good or very good in more things without being excellent in a very specific set of items. Better mix of possession vs. separation vs. jump up and get it receivers, better mix of run d vs. pass d, o line pass block v. run block, etc and probably a better line to begin w/ because Manning is unique in his ability to read the D and get the ball away quickly.

    All of these things allow the team to be more adaptable w/ respect to a replacement QB than Indy. We'll never know for certain, but that's my take on it after scrubbing the stats (DVOA and similar items) and watching the games (looking for play calling, strategy, and execution differences w/ and w/out Manning).

Leave a Reply

Note: Only a member of this blog may post a comment.