≡ Menu

Checkdowns: Carson Palmer and MVP Voting

Since twitter limits me to 140 characters, and I’m having fun debating with guys like Bryan Frye (@LaverneusDingle) and Adam Hartstad (@AdamHarstad), I thought I’d crunch some numbers here.

Is Carson Palmer the best choice for MVP this year? Let’s put aside the Cam Newton argument and just focus on Palmer’s place in post-merger history.   As Adam pointed out, 2015 Carson Palmer is currently 7th on the list of Adjusted Net Yards per Attempt leaders, and the six players in front of him all won the AP MVP (that would be Peyton Manning 2004, Aaron Rodgers 2011, Dan Marino 1984, Tom Brady 2007, Manning 2013, and Rodgers 2014).

Of course, ANY/A is biased in favor of modern players, so let’s look at ANY/A+. Here, he doesn’t drop as far as you might think: Carson is still tied for the 11th best season since 1970, and a few non-AP MVPs sneak in there ahead of him (Mark Rypien 1991, Randall Cunningham 1998, and a quarterback who lost to another quarterback a having historic season: Montana ’84 and Ken Anderson 1981).

So we know Palmer is having an awesome year; how does it compare to the average MVP-caliber season? Before the 2014 season, I looked at just that question.  Here is how 2015 Palmer relates to the average MVP-caliber year:


Here’s how to read the table:

  • An average MVP has a completion percentage that’s 112% of league average; Palmer’s is only 102%.
  • An average MVP puts up about 25% more passing yards than league average; Palmer is at 118%.
  • Even in passing touchdowns, Palmer doesn’t stand out as great: yes, he has 144% of league average, but this is where MVP quarterbacks usually shine, at 159%.
  • Similarly, Palmer looks below-average — relative to MVPs, of course — in interception rate.   Palmer’s interception total is good, but only a little above league average.
  • In terms of sacks, sack yards, and pass attempts, Palmer pretty much is in line with what an MVP quarerback has.  That gets us to the big category….
  • As far as ANY/A goes, Palmer doesn’t stand out here as much as you might think.  Of the 31 quarterbacks to win the MVP award since 19701, 23 have had ANY/A averages higher than Palmer, at least as measured as a percentage of league average.
  • As for wins, Palmer is on pace to win about 13.5 games, slightly above the average MVP.

There are many ways to measure league average, and given the rising averages, you could argue that percentage of league average is the wrong way to go.  With higher averages, it’s harder to stand out: Palmer would look better if we subtracted league average, rather than divided by it.  That’s a mathematical theory question that we could debate.  But I was a bit surprised to see that Palmer doesn’t quite stick out here as much as you might think.

If we look at Relative ANY/A — which would be Palmer’s ANY/A minus league average — he would be at +2.32.  You can find some pre-2014 RANY/A ratings here, albeit in a somewhat unorganized (for this purpose) manner.

Of course, it goes without saying that Newton — with a ANY/A as a percentage of league average of 113% and a RANY/A of +0.83 — doesn’t fare particularly well in these metrics.  But Newton’s case isn’t built on RANY/A (thankfully, for him) — it’s built on his contributions to a team that’s devoid of offensive weapons, an undefeated record, and what Newton adds on the ground.  That’s precisely the sort of thing a quarterback would need, though, given that no quarterback since 1970 has won the MVP award without having an ANY/A average at least 20% above league average.

And it should be obvious to those of you who understand the way these formulas work that Palmer looks “worse” here because of the rising league average. This season, the NFL average ANY/A is currently 6.30. It could still dip over the last three weeks, and Palmer could have a hot finish: those two factors may be enough to pull him around MVP-average. But as it stands, 2015 is set to be the latest in a long line of “best passing seasons ever” in recent history.

  1. Including 2014 Rodgers, who wasn’t on the initial table. []
  • Nit: Randall Cunningham 1998 was an MVP season. It just wasn’t an AP MVP season.

  • Also, any particular reason you opted for percentage of league average instead of standard deviations over the league? That would account for the “harder to get a big percentage of a higher average” issue. I know we only have ANY/A back to 1969, but we have AY/A back a good deal further. Just a “good enougher” choice?

    • Pretty much. When I crunced the original numbers, I think that was the reason. I’m not sure standard deviation is all that useful when talking about one season, anyway, as the distributions tend to be pretty random. I remember spending some time with Drinen looking at 10 years worth of data, and the STDEVs were all over the map, but if your null hypothesis was one standard deviation for all 10 years, it couldn’t be rejected.

      But frankly, I don’t even know if Palmer would stand out that much by using STDEV. RANY/A is probably the tool you want to use here to promote Palmer, but it’s not like his RANY/A is extraordinary. It’s great, but not historically great. I have an old file from Neil that calculated RANY/A. From 1970 to 2011, there were 39 QBs with at least 300 attempts and a RANY/A of at least +2.3. That’s basically 1 per year. Bump it to 400 attempts, and you get 23. At 500 (which is where Palmer will finish), it’s 8, including all the seasons you would think and Culpepper 2004. But throwing out sub-400 attempt seasons from the ’70s probably isn’t fair.

      Anway, I stand by my initial, unresearched view: Palmer’s season is awesome but it’s not historically incredible. Including first downs might help, tho.

      • I’d be curious to see Carson vs. History according to Bryan’s numbers. I know he’s the top QB by a good bit this season, but no clue where he ranks against years gone by. Bryan?

        • I plan on doing a guest post here after all 16 regular season games are played. Any number of things could completely change the voting between now and then. I only have Air/YAC splits going back to 1992, and I still haven’t parsed spikes and kneels out past 2002, so it could get a little sloppy.

        • I will say that Palmer currently has 3.01 marginal TAY/P, which would tie Steve Young’s 1993 for 17th since 1992. However, if we only give half credit for YAC, Palmer’s +2.99 score would tie Steve Young’s 1994 for 8th over that span. The only better seasons, per play, were:
          Peyton 04: 4.47
          Peyton 06: 4.01
          Brady 07: 3.54
          Young 92: 3.54
          Peyton 05: 3.35
          Cunningham 98: 3.34
          Rodgers 11: 3.31

          Unless you don’t care one bit about minimum play thresholds, because Derek Anderson 2012 was a MONSTER at +14.41.

        • Also, with a hot finish, Carson could get to 5K/40, which I imagine would sway some voters.

          Also, Palmer would be the first QB since Manning ’04 to lead the league in Y/C and win the MVP.

  • FYI, here are Bovada’s MVP odds:

    Cam Newton 1/5

    Tom Brady 7/2

    Carson Palmer 12/1

    Russell Wilson 12/1

    I agree that Palmer’s season is being overlooked. I don’t really know why that is.

    • Richie

      All 4 guys seem like legitimate MVP candidates right now. Of course, that could change over the next 3 weeks. But, usually it seems like there are only 1 or 2 guys battling for MVP. Is it unusual to have 4 guys neck and neck? I really wouldn’t be disappointed to see any of them win the award.

      Though, I have fully capitulated on Tom Brady, and think he probably deserves the award. You just keep pulling guys off his team and it doesn’t matter – he keeps winning. But with 3 other valid candidates, I suspect voter fatigue will prevent him from winning.

      • My problem with Brady is that you have to look at it as two stretches. From weeks 1-10, he played 9 games. In them, he averaged 8.22 YPA, had 24 touchdowns vs. 3 interceptions, and had a passer rating of 111. But this stretch was with his offensive supporting cast at full strength: Dion Lewis played 7 of those 9 games and Julian Edelman played all 9, (plus full use of the Gronk, of course). New England went 9-0 during this span.

        From weeks 11-14, Brady played four games. In them, he averaged 6.56 YPA, had 9 TDs vs. 3 INTs, and had a passer rating of 86.8. New England was just 2-2 over this stretch. All of these represent dramatic drops from his early-season pace, which is to be expected because he lost Lewis and Edelman, (and played one game without Gronk and Amendola).

        The difference between the two stretches is the difference between a 56.3%, 4380-yard, 6.6 YPA, 87 rating, 36:12 TD:INT, 8-8 season and a 68%, 5410-yard, 8.2 YPA, 111 rating, 43:5 TD:INT, 16-0 season. One is good. The other is historic. So saying it “doesn’t matter” that he’s losing his weapons is just wrong. It matters a huge, huge, huge amount. From a YPA perspective, it’s the difference between Andy Dalton (8.4 YPA) and Colin Kaepernick (6.6 YPA). From a passer rating perspective, the first-nine-weeks value would lead the NFL, while the last-four-weeks value would rank 26th, (tied with Matt Ryan and just behind Matt Hasselbeck, Ryan Tannehill, Blake Bortles, Johnny Manziel, and Ryan Fitzpatrick).

        The sleight of hand with Tom Brady is that he accumulated his season-long numbers largely with the help of that supporting cast, but then they got injured late, and now we get to attribute the sum total of his production, (most of which, again, came with his supporting cast in place), to this new, decimated offense. But that’s not how it works. When Tom Brady’s supporting cast all got hurt, his production took a huge nosedive. He’s not doing more with less. He is, in fact, doing a lot less with less.

        Which is why I think he’s not as deserving as Carson, (who has better numbers, and whose production has remain unchanged as his RBs have gotten hurt and his non-Fitzgerald receivers have cycled in and out), or Cam (who actually did accumulate the entirety of his season’s statistics with little offensive support).

        • Richie

          Fair point. However, I wonder if Dion Lewis and Julian Edelman truly count as “weapons”. Nevertheless, as you point out, the numbers dropped without them.

          • I think they absolutely do. I see a similar pattern of dismissing Edelman that we saw with Welker when he first broke out. Because he doesn’t look like a traditional receiver or win in a traditional manner, he’s deemed gimmicky or a product of the offensive system. The reality is that he’s a phenomenal athlete with elite short-area quickness (ranked in the 98th percentile among WRs in the short shuttle and 95th percentile in the 3-cone drill), and he makes Tom Brady’s life a lot easier by winning where Brady excels.

            Interesting aside: Welker ranked in the bottom quartile of all athletic drills except for the Short Shuttle, where he was 91st percentile. Brandon Stokley, another famous slot receiver, was bottom half in all drills except for the Short Shuttle, where he was 91st percentile. It definitely seems like there’s a “prototype” for those undersized slot receivers.

            As for Dion Lewis… my God, he was a revelation. Led the league in missed tackles forced by a scale-breaking margin and routinely showed an ability to adjust on-the-fly to poorly thrown passes. He was an accuracy-erasing, YAC-gaining monster. Brady’s 8.6 AY/A on throws to Dion were the highest to any non-Gronk Patriot since 2012. (Edelman’s 7.8 AY/A ranks 3rd.) He was by far my favorite player to watch play this year, and I desperately hope he comes back at full strength next year.

          • Adam

            According to ESPN, Brady’s receivers have generated 468 more YAC than expected, given the location and length of his throws. That’s #1 in the league by a wide margin. So yes, Edelman, Amendola, and Lewis most definitely count as weapons.

        • Adam

          You’re exactly right. The narrative throughout Brady’s career has been, “He does more with less.” From 2001-2006, this was a fair assessment. But since 2007, Brady has played with all-timers in Moss and Gronkowski, and his production has dropped significantly when neither is on the field. He looks like the GOAT with Moss/Gronk, and reverts to `01-`06 performance without them. When Brady has less, he does less. How can someone who is so dependent on teammates be considered the MVP? I would actually be fine with Gronk winning the MVP this year, but not Brady.

        • SteveH

          I was going to make a similar point. Chris Johnson is out. Andre Ellington has been often injured and ineffective for the most part. David Johnson looks good right now, but he is a rookie. So Palmer will be discounted further because David Johnson looks so good.

          On the other hand Jonathon Stewart has had a hot start to his career, but then injuries have lead to an underwhelming career and lowered expectations. He is doing well this year, therefore it must be because of Cam. Stewart has had a hard time staying healthy (kind of like Ellington) but if I remember correctly was always considered to have a lot of potential.


          Michael Floyd (AZ ‘s supposedly #1 receiver) missed the beginning of the season with dislocated fingers, and was largely ineffective until week 6 or 7. The 1st and 2nd year receivers Nelson and Brown have both missed time and even if they were active were playing injured. Fitz was considered washed up. So Palmer doesn’t get the credit for making these young receivers look great, and of course Fitz has just performed up to his expected high level. Fitzgerald is even blocking so that makes Palmer look bad (sarcasm).

          On the other hand Cam gets full credit for making Ted Ginn Jr. look like a world beater, because Ginn was washed up. Cam also gets credit for making his 1st and 2nd year receivers Funchess and Brown not look terrible.


          Then AZ has below average tights ends. Darren Fells only caught 5 passes last year. Gresham is barely used. He has name recognition, but with his back surgery he joined the team late and is still barely utilized in the passing game.

          On the other hand Greg Olsen has been a great tight end for a number of years. Yet he doesn’t seem to be considered part of the horrible supporting cast that Cam Newton has around him.

          • Adam

            Agreed. The supporting cast argument can be twisted in favor of or against just about any QB. I find it laughable that Newton backers gloss over Greg Olsen, as he has been arguably the second best TE in the NFL this year. If we’re going to look at supporting casts objectively, we must consider every teammate who affects the quarterback – WR, TE, RB, OL, HC, OC, and even defense and special teams for their impact on field position.

            Then again, when have QB narratives ever been rational?

  • sacramento gold miners

    Palmer has a strong case for MVP, but I wish the voting was done after the postseason. It would be great to see Palmer vs. Newton in the playoffs, and that would play a role for me in making the final decision.

    • James

      That would be more relevant if it were actually Palmer vs Newton in the playoffs, and not Palmer vs the Panthers D and Newton vs the Cardinals D.

      • sacramento gold miners

        Point taken, but it’s still relevant. If we went back and checked playoff results, the winning team usually has the better QB performance. It’s Palmer’s team versus Newton’s team, and how each offense performs affects everyone. Can’t dodge the fact QB is the most important position. If Palmer struggles versus the Panthers D, that automatically helps the Carolina offense, in terms of the scoreboard and field position.

  • Trepur

    I think Palmer is MVP for this season.

    Though Cam will win it because voters place way too much emphasis on a QBs WL record.

    • James

      Really interested to see how voter opinion shifts if the Panthers don’t finish undefeated. Will Palmer’s stats carry the day if it’s 15-1 vs 14-2?

      • Trepur

        Hopefully it slides Palmer into the win, he’s truly been a phenomenal quarterback this season, better then Rodgers from last year in my opinion.

        But the Cam Newton hype train has already left the station, and if Newton gets benched for the second half of week 17, even if they’re losing the game at that stage, and they go 15-1 as a result, I don’t see the voters voting against a 16-0 QB, which is unfortunate.

        A QB is by far the most important player on a team, but on a 53-man roster, especially one who plays less then half the game (Special teams and defense making just over half the plays) can only do so much, and it’s a team, not a quarterback, that wins games.

        • sacramento gold miners

          Under the format of awarding the honor before the postseason, Palmer likely clinched the MVP yesterday, easily outplaying Aaron Rodgers. I think Palmer himself would admit he’s a much better leader than those years with the Bengals, and that’s been a real factor in his personal and team success in 2015. And as the NFL becomes even more passing-oriented, winning and leadership will always be a strong indicator of the QB position.

          • Trepur

            I think he did. but some people are more impressed with the 15-1 Panthers.

  • Josh Sanford

    This would be the perfect set-up for JJ Watt to win the MVP (no glaringly transcendent QB) but he has only turned in a classic, offense-wrecking All-Pro season (shame on him), and he hasn’t done anything to galvanize shallow media observation in his own support. Come on, JJ–what are you thinking?

    • Richie

      It also would’ve helped if his team was on pace to win 10 or 11 games.

    • Adam

      I agree that JJ Watt has played at an MVP level the last two years. The problem is that his amazing play hasn’t translated to wins, so voters / fans think “How valuable can he be if the Texans aren’t winning? I don’t like this rationale because the MVP is an individual award, not a team award. Honestly if JJ can’t win an MVP, then no defensive player ever will. And that is a shame.

  • Adam

    My MVP preferences, in order:

    1) Carson Palmer
    2) Rob Gronkowksi
    3) Ben Roethlisberger
    4) Russell Wilson

    These four men have been the drivers of successful teams, which is why I give them the nod over Newton or Brady. While Newton has certainly elevated his mediocre offensive teammates, the Panthers’ success is ultimately driven by their defense and an easy schedule. The Pats’ offense collapses without Gronk, so that makes him more valuable than Brady in my mind. The Steelers offense is a juggernaut with Ben and barely average without him – the very definition of valuable. As Seattle’s defense has faltered, Wilson has picked up the slack and quarterbacked his team to victories on his own.

    • Andy

      As of last week, the Panthers had beaten more teams with winning records and more projected playoff teams than any other team in the league. Not quite the easy schedule that you seem so convinced the Panthers have.

      Also, the Panthers rank 1st in points per game while sitting at 4th in points allowed per game. By this metric, it’s pretty clear that the team’s success is not solely due to having a good defense.

      Cam Newton is the league’s MVP. He carried this team offensively with a brick-handed receiving corps and an offensive line made up of castoffs and nobodies. How anyone can say otherwise at this point is beyond me, because the Panthers aren’t anywhere near 13-0 without Cam Newton.

      • Adam

        The Panthers will play half of their schedule against the two worst divsions in the league (AFCS, NFCE), plus six more games against the mediocre NFC South. Carolina’s DVOA is 7% lower than its VOA, indicating a very easy schedule. I agree that Cam has played great this year, and the Panthers wouldn’t be 13-0 without him, but that doesn’t make him the MVP. It’s an individual award, not a team award, so the Panthers’ record is irrelevant to me in this discussion. Cam is one of my favorite players, and I’m rooting for Carolina to go undefeated, but I believe Carson Palmer has added more to his team’s offense than Cam has to his.

  • Clint

    Palmer will – and should – win if you the Panthers lose next week. If the Panthers win, Cam wins.

  • WR

    I know I’m late to this discussion, but I just feel the supporting cast argument that’s now being made against Brady is inconsistent. Sure, he puts up better numbers when Gronk is healthy, and his best season happened when he had Moss. But so what? Every QB who has put up all-time numbers has had a great supporting cast, whether it was Brady in 07, Manning in 04 and 2013, Rodgers in 2011, Young in 94, Montana in 89, Warner 99-01, Cunningham 98, etc. One of the reasons why Manning has such great numbers is that he’s never been without at least one star pass target. Every year of his career, Manning has had at least one of Faulk, Harrison, Wayne or Demaryius Thomas.

    Since the start of 2005, Brady has been just as good as Manning, and just below the level of Rodgers. The periods when he’s been weak statistically were 2006, when his best receivers were Gaffney and Caldwell, and 2013, when the Pats offense, number one in the NFL in 2012, was decimated by injuries, defections, and unexpected prison sentences. Do you really think it’s fair to rip Brady for putting up weaker numbers when his best receivers are Edelman, Aaron Dobson, and Austin Collie?

    I understand that a lot of people don’t like Brady, whether it’s the unrelenting success, the good looks and supermodel wife, or Spygate and Deflategate. But far too often, I see people applying standards to him that never get applied to a lot of other players. With 2 weeks to go, I think I would vote Newton for MVP, and there’s also a case for Palmer and Wilson. But what Brady has done with all the injuries this season is remarkable. The Pats offense is still number one in points per drive, which is crazy when Gronkowski, Edelman, Amendola, Lewis, Blount, and LaFell have all spent time on the sidelines.