Don't Pay CJ!

The word is that Chris Johnson won't be satisfied with being the highest paid RB in the league. He wants to be among the highest paid players at any position. With Larry Fitzgerald's recent 8-year deal guaranteeing nearly $50 million, Johnson must be thinking he's worth something similar. Peyton Manning's recent contract will pay him nearly $20 million per year, and it might make sense to pay one of the greatest QBs of all time 1/6 of your entire payroll. But paying that kind of money to any RB, even the very best, makes no sense.

Adrian Peterson is currently the highest paid RB who is due to make $11 million this season, which is probably two to three times too much. RBs, and the running game in general, do not have the impact on wins and losses the same way the passing game does. But more importantly, the spread between the best and the mediocre RBs is much smaller than the the spread among quarterbacks.

We can't quantify the value of a player the way the MLB analysts can, with Wins Above Replacement and other stats. But we can at least make some back-of-the-envelope, order-of-magnitude estimates.

Comparing the standard deviation of 2010's top 40 QBs' stats to the top 40 RBs' stats shows that the difference between winning and losing rests far more in the QBs' hands. The SD of QB WPA is 1.75 wins, and the SD of RB WPA is 0.47 wins, a factor of over 3 in favor of QBs. On a per game basis, it's a factor of over 4.2.

Comparing EPA, the SD for QBs is 58 points, and the SD for RBs is 18 points. That's a factor of over 3 in favor of QBs. On a per play basis, it's a factor of 1.4 in favor of QBs, but keep in mind that top QBs are involved in twice as many plays as the top RBs.

But when it comes to running stats, it's Success Rate and not EPA or yards that matter in winning games. Chris Johnson's career SR is only 38.2%, compared to the average SR of 38.6% for all RBs. Johnson is a homerun hitter, and although that's what makes it on Red Zone Channel, it's not necessarily what wins games. It's the steady threat of 4- or 5-yard gains that opens up the passing game, not the occasional break-away TD run that makes a winner.  It doesn't matter that you hit 50 HRs when you strike out on 62% of your at-bats.

Johnson does have impressive Yards Per Carry numbers for his career. 5.0 YPC is solidly above average. But much of that is thanks to his one amazing season in 2009. Last year, his average was down to 4.3, nearly average. In 2010, his WPA was -1.0 wins and his EPA was -33 points (yes, solidly negative). For his 3-year career his total WPA is only 0.40 wins, or 0.13 wins per year. In simple terms, that means adding the Chris Johnson running game to an otherwise completely average team would make it...a completely average team.

RBs in general are known as injury risks. Plus, Johnson isn't exactly known for his pass blocking. Both factors weigh against giving him a giant contract.

This isn't a knock on Johnson as much as it is on how RBs are overvalued in the NFL. I think there are a several reasons why this continues. A few decades ago, the NFL was truly a running league, and great RBs could make a much bigger impact on the game. There's also the illusion from the fact that RBs happen to be the guys with the ball in their hands and the stats on the back of their trading cards. We all remember the days on the sandlot or on the varsity team where RB is almost always the best athlete on the field. Lastly, there's the age-old causation/correlation fallacy in which RBs on winning teams pad their yardage totals sealing wins in the 4th quarter, making RBs appear to be causing his team's success.

An accountant might be able to make the case that Johnson is worth more than the wins he can generate because he puts butts in seats and jerseys on backs. But you know what else sells tickets and jerseys? Winning. Take the $12 million per year, put it into your offensive line, and draft a couple 4th round RBs. You'll thank me later.

  • 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

57 Responses to “Don't Pay CJ!”

  1. Anonymous says:

    Did the Titans pay you to write this article? He is the face of that organization and if he did anything negative for the pathetic Titans last season, it was giving them wins that they would not have had otherwise which ultimately caused them not to get the number one pick for this year's draft where they could have brought in a decent QB that could learn under Matt. They should've renegotiated his contract long ago, but they took advantage of paying peanuts to one of the most sought after RB's in the league. You don't treat a star player like that... he could've went Diva long time ago, but chose not to and just played... played hard. I'm not saying he should be the highest paid player in the league, but they need to do something. Just like every other owner who has no business owning a franchise and trying to nickel and dime the operation... Pay your players. Cheap bast**ds. Pay that man so we can move on.

  2. Jules says:

    I wish I could understand the statistical analysis. However, I understand what you're saying. From a non-statistical point of view, I look at recent running backs who held out of camp demanding a new contract who, at the time, were at the top of their game. Names like Shaun Alexander and Larry Johnson come to mind. They totally fell apart once they got that new contract.

    However, I say that CJ should continue to hold out, and perhaps eventually the Titans will trade him (especially if Ringer takes off this year). He has three years left on his current contract, and he should go after that next big contract now. After all, odds are if you go ahead and report and play without a new contract, you will tear your ACL or something equally bad, and then they'll never pay you.

    But if I'm the Titans, do I pay him? If they were closer to contention for a division title, then yes. However, they're not and they are working towards rebuilding their team with a new coaching staff. If Ringer plays well enough, they will let Johnson stew at home.

  3. jditoro says:

    despite all of the very sound statistical reasons why they shouldn't, how much do you want to bet they cave and pay him more than Peterson?

    here is something i've been a bit curious about though: if a team becomes rational in an irrational league, will the fact that other teams are paying too much at many positions cause a cultural issue on the rational team? will players feel that the organization doesn't take care of them, and will those players fail to come together as a cohesive unit to win? just a thought.

  4. Ogro says:

    Good Analysis, but I have an extremely minor issue with this quote:

    "It doesn't matter that you hit 50 HRs when you strike out on 62% of your at-bats."

    This is nitpicky, I know, but an out in baseball is generally the same however it came about (double plays and sacrifices notwithstanding), so the example to use isn't strikeouts, it's batting average or on-base percentage, since if you either get an out or a homerun, and you only homer 10% of the time, you're not particularly useful compared to a more average .270/.330/.500-type player.

  5. Brian Burke says:

    Chase's take.

  6. Brian Burke says:

    Jason's take.

  7. Joseph says:

    Joseph's take: Pay the man, just not exorbitantly.
    My idea: [If this wouldn't fit under the cap & 30% rules, just make some of the $ roster bonuses or LTBE incentives.] Give him a 4 yr, $40M total deal. $8M=signing bonus; "salaries" of $5M, $7M, $9M, then $11M (these #'s would include roster bonuses or LTBEI's) Guarantees would be $2M in the first 2 yrs, $3M in the last two ($18M total guaranteed). That would give him 7 yrs in the league (=29 or 30 yrs old). If he's still playing at a high level, give him a 3 yr extension after yr 6 (3rd year of my proposed deal)--cause that would be a 13M cap charge in the last year. If he's not, cut him, or let him play it out and be a FA RB with a lot of tread on the tires. Maybe he ages like Payton or Emmitt--extend him. If he ages like Alexander or LJ, bye-bye.

  8. Fitz says:

    So I have one methodological questions that's been bugging me for awhile. It popped into my head a couple seasons ago when A. Peterson was particularly hot shit, but he had a lower-than-expected EPA and/or WPA here.

    I was wondering if that's because the better a running back is, the more likely it leads a team to run instead of pass, even though passing is more efficient. Is this a possibility with the stats?

  9. Brian Burke says:

    Fitz-Very perceptive. Yes. CJ actually had significantly fewer rushing atts in 2010 than 2009--mostly due to the fact TEN was behind in more games.

    It's all about the game theory aspect of run/pass. TEN's passing game collapsed in 2010, making running relatively more attractive. But defenses are just as aware of this, and can load the box.

    The important thing to realize is that this cuts both ways, equally. The running is supposed to open up the passing, and the way to do this is with high SR.

    For a runner like CJ (low/avg SR, home-run threat) to be truly effective, TEN needs a passing game. Think Marshall Faulk.

    In AP's case, it's mostly about early-career fumbles.

  10. Bob says:

    Why are you looking at the SD of WPA - is it the SD across the ensemble of QBs? The WPA is what matters.

    That would seem to be completely irrelevant, other than demonstrating that the RBs have a much more consistent WPA than QBs have. Which makes sense, since throwing a TD, or throwing an INT have large effects, whereas running for 2 yds or for 10 yards has a smaller effect.

  11. Brian Burke says:

    Bob-Yes, WPA tells the story. But scarcity and $ value is about the ability to implement a substitute 'good.' QBs will certainly have higher WPA, even to the point where poor QBs have better WPA than good RBs, but that's not important. It's the marginal value of a player above what you can get with a substitute that matters. That's why the baseball guys use 'above replacement' stats.

    On a 100-point scale, QBs are all over the map. In terms of their impact on the game, QBs range from the 90s to the 10s. RBs, on the other hand, range from the 40s to the 50s in terms of impact.

    You don't want to pay through the roof for a '60' RB when you could easily replace him with a '50' for cheap, and take the savings and put it elsewhere.

  12. zlionsfan says:

    Honestly, I think the argument for not paying Chris Johnson big bucks is ... well, Chris Johnson. It's not as though his 2009 season made the Titans into contenders: their problems aren't going to be solved in the backfield.

    And it was just three years ago that Tennessee went to the draft to shore up their backfield. When LenDale White didn't seem like he'd locked down the job, the Titans picked up Johnson in the first round. They could just as easily have had Forte or Rice or Charles or a handful of other backs who could possibly be productive in that offense.

    To be competitive, the Titans don't need an All-Pro RB, particularly not if that money can be used to fix other problems. They might just as well use Ringer and spend that money on a veteran OL or two, a WR to complement Britt, and maybe CB or LB help. It's not as if Tennessee's ownership is known for handing out big checks, either ... if they do sign Johnson, and they don't spend any more money, I don't think there's any reason to expect them to contend.

  13. Jim Glass says:

    Everyone loves an exciting RB, but the list of league-best RBs whose teams never won squat is impressive: Gayle Sayers, OJ Simpson, Earl Campbell, Eric Dickerson, Barry Sanders, etc.

    OJ Simpson's teams for his career averaged 4-10.

    Walter Payton in a 13-year career was on only two good winning teams -- and was washed up on the last one getting only 500 yards. But the one the year before that had maybe the best defense in NFL history, and *that* was the *one* year his team really won.

    You win in the NFL with passing, O-line and D-line play, the whole team. To the extent that over-paying a star RB weakens the team at other positions, you lose.

    (In fact, I've always suspected that a real *star* RB is an outright liability because the team rebuilds itself around him to make the most of him, weakening areas more important to winning. See: Barry Sanders, Lions; OJ Simspon, Bills, etc.)

    OTOH, lots of teams have won the SB with a crew of good all-round backs but no top one -- and without paying for one.

    CJ should hold out for every penny he can get -- life as a top RB is short, he'll probably never have a chance to get such money again, he should go for all the gold he can get right now. This is a business.

    But it's a business for the team too.

  14. Jeff Clarke says:

    I think the important thing to remember is that when the salary cap basically mandates that all teams spend the same amount of money there is no such thing as a "cheap owner". Its not a question of to spend or not to spend, its a question of where to spend the money.

    Its a simple fact that if you spend more than average in one place you are going to need to spend less than average in another place. When you decide to overpay a couple of "star" players, you're going to end up with a bunch of other scrubs playing. If you believe this will work, just count the number of trophies in Dan Snyder's office.

    Somebody else asked what happens if you have a rational team in an irrational league. You have one: The New England Patriots. For ten years now, the Patriots have had somebody decent starting at practically every position. They have a great quarterback (the one position that really matters). Practically everywhere else, they fielded good but not great starters. When these good but not great players start to demand more money than they are worth, the Patriots willingly let them go. Their new teams often find that they seriously overpaid. Value investing does work in the NFL.

  15. zlionsfan says:

    but that's not what the salary cap does. It prevents teams from spending more than a fixed amount (except for certain circumstances), and now requires that the average of all team's payrolls be at a certain percentage of the cap, but it doesn't require that they all spend the same amount.

    This deal is significantly different from pretty much any other (either in the NFL or in other capped sports) because the floor is so high. Typically the only discussion is about the cap ...and even so, there's quite a bit of leeway for individual teams. The Titans can spend as little as 89% of the cap as long as enough other teams meet or exceed it to bring the league "average" up to 99% of the cap.

    Of course, that presents another bit of strategy to Johnson. If he and his agent have access to updated salary numbers, he can get an idea of whether or not the Titans are close enough to the cap without him. If they're not, they'll have to sign him (although I don't know what the penalty is for being below the floor, nor when it would be enforced). Then again, it's also possible that the Titans could sign other players and move close enough to the cap that they wouldn't have room for the money CJ is requesting ...

  16. Anonymous says:

    @zlions, there is no per team spending floor until 2013

  17. chris says:

    he does more than run.....he is also much like LT was in his prime where he was a threat for screen passes, in 2009 he was also totaled the most catches for the titans. Whether you want to admit it or not he was and is a huge portion of their offense

  18. Brian Burke says:

    Chris-Very true, but the WPA and EPA numbers include his receptions. Plus, among RBs in 2010, CJ was 22nd in the league in rec yds.

  19. Jeff Clarke says:


    You're right. It doesn't say that teams all need to spend exactly the same amount. It does put them all into the same neighborhood though. There are a lot of rules, exceptions to the rules, etc. I'd like to see more articles on the whole subject.

    The basic point remains though. The more you spend on any one player, the less you have to spend on all the others. When teams don't give in to salary demands for high profile players, its usually because a greater proportion is going to other players.

    The question shouldn't be to spend the money or not to spend the money. The question is where do you spend it?

    I agree with Brian that the Titans would get more bang for the buck if they invested elsewhere.

  20. Brian Burke says:

    On the other hand, the Titans do have cap room under the current year for a pretty big raise. So at least to some degree, the $ they can offer Johnson won't rob from other parts of the team. Beyond the current year, I'm not sure. Still, it seems some teams, like TEN, have a self-imposed cap, $10 or $20 million under the league cap. If we just accept that as "the cap", the same principles apply.

  21. Anonymous says:

    man you titans fans whine when you dont have playmakers and when you do you want to trade them lol bunch of people not knowing exactly what they want!!!!!

  22. Anonymous says:

    Brian don't read too much into the numbers the end of the day any team would love to have cj as there rb....and would probably win a super bowl alot quicker than the titans...the titans problem is they trade away any decent talent that they get and then wonder, "why are we so average" lol

  23. bob says:

    Brian, while I understand that point, that a typical RB can easily be replaced with another typical RB, that does not apply in this case.

    You have chosen one stat and showed that the SD is low, and therefore they are interchangeable. However, one can simply look at a different stat and it puts CJ into rare HoF company indeed. A 2000 yard season.

    Yards are an important stat, and probably a better indicator of an RBs value than WPA is.

    When you calculate these WPA values, they are based on single plays right? Of course QBs will vary a huge amount compared to RBs.

    What you need to do to follow your argument is measure the year to year stability of each qb, and get a measure of how robust this measure is.

    For instance, on your scale of 1 to 100, let's say a qb can vary by 30 points. Now, a qb with an 80 is might also score a 50.

    Now look at rbs, and see how they vary from year to year, and to see what that value is (certainly much lower than the qb variations).

    Then you can compare how much better an RB is than other RBs, as compared to a qb to other qbs, using their respective proper deviations.

    (and even still, i don't think individual wpa for each play is the way to compare the value of qb to rb)

  24. Chris says:

    You think 2,000 yard seasons are the way to compare the value of QB to the value of a RB. How about let's just look @ the last 2,000 yard season by a RB who was that, J. did he do after that 2,000 yard season?

    I'm sorry, but your logic is ridiculous. I mean what's so special about a 2,000 yard season. Why don't we look @ 1,955 yards, or maybe we should make our arbitrary cut off line 2,012?

    (Oh and a 2,000 yard season isn't going to get J. Lewis into the HoF).

  25. Ian Simcox says:

    Bob - what does a 2,000 yard season have to do with it? The Titans finished 8-8 the year he had 2,000 yards. That year Tennessee racked up twice as many rushing yards as Indianapolis - but would you rather have the 8-8 or the 14-2?

    Truth is, that behind a good O-line nearly any college running back would be able to grind out effective enough rushes. Teams don't need a 2,000 yard rusher to get to the playoffs, they need a passing game that can get 7 yards per attempt.

  26. bob says:

    Chris, Ian, the 2000 yd season is a statistical outlier, and directly addresses the original argument, that CJ is easily replaced by another RB. Obviously, another RB would have been highly unlikely to hit 2000 yards.

    That is what this article is about - it directly uses the SD of WPA as it's criteria, and choses the lower SD of RBs to conclude he can be replaced and thus to indicate a lack of value.

    Also, you are both assuming that it had no effect in the wins that year. perhaps they were a 4-12 team or a 2-14 team without CJ, and 8-8 with him.

    The fact that the 53 man roster of the Colts had more wins than the 53 man roster of the Titans is not relevant.

  27. bob says:

    This leads possibly an interesting calculation that should be possible, just run the WP for all the games in that season as they were (i.e. like the calculations showing the "luck" of the teams) and then replace CJ with his backup for that year (or for whatever RB is currently available right now, tiki barber? ha) and see those results. That would give an idea of how valuable his 2000 yd season was.

  28. bob says:


    "You think 2,000 yard seasons are the way to compare the value of QB to the value of a RB. "

    I do not think that. In my post I am stating that I think comparing the variance in qb WPA to rb WPA is not a good way to compare the two.

    In 40 years, this has occurred 5 times. CJ accomplished this, and he is only 25.

  29. zlionsfan says:

    I don't think it's obvious that another RB wouldn't have hit 2000 yards with the 2009 Titans OL blocking for him. Johnson himself hasn't come anywhere near that number in either of his other seasons. Furthermore, it's not something that can be supported either way: you can say it couldn't happen and someone else could say it could, and neither side would convince the other.

    Your idea that perhaps the Titans would have been much worse without Johnson is a similar contention ... and unfortunately there is some evidence to the contrary. That's the idea behind mentioning WPA for various positions and for Johnson in 2009. With Johnson posting an WPA of 1.33 that season, it's unlikely that the Titans would have been a 2-win team without him.

    If you really do believe that yards are a better indicator of a RB's value than WPA, then I'm not sure any of this will matter to you ... but at the very least you should keep in mind that only one 2000-yard rusher has been on a team that won even one playoff game that season, and that offense wasn't relying solely on the running back in question; the QB was pretty good himself.

    People may believe that RBs in today's game are the lions ... er, giants ... er, monuments that they were in a previous era, but there really aren't numbers that support that feeling, and if the Titans do spend a lot of money on Johnson and not on other areas of the team, they're not likely to see any more success than they've had to date with him.

  30. Andrew Foland says:

    There are two relevant questions. The first is, what quality of a running back most contributes to team victories? The second is, how much more does CJ have of this quality than an average back? Those two answers, together with a cap number to set the scale, tell you how much more than an average back CJ should be paid.

    Brian has done a lot of data-driven analysis to answer the first question; the work on SR, EPA, and WPA comes out of that. The choice of these metrics is not an "arbitrary selection of just any stat"; it is a well-motivated selection of a meaningful stat. You are not free to "look at a different stat" unless you first demonstrate the other stat has a higher prospective correlation to team success.

    The issues about standard deviations of QBs versus RBs get at the second question. The point being that a 2-SD-above-average RB (i.e. "the best back in the league") does not add nearly as much to a team as a 2-SD-above-average QB does.

  31. bob says:

    Andrew, good points thanks.

    The 'arbitrary selection of a stat' is directly referring to the indeed arbitrary selection of the SD of ALL RBs, and comparing it to the SD of ALL QBs. Yes, that is an arbitrary selection of a highly refined stat, and I don't know if that has ever been used before. In fact, it obfuscates the question.

    The question, 'is do you pay CJ a lot of money'. The question is not "is he worth more than peyton manning'.

    So, you can plainly see the specious argument. QBs are more variable than RBs in WPA as a general rule, therefore do not pay one particular running back who had the 5th best alltime performance.

    Additionally, no one has addressed the main point. WPA is calculated as a change per play. RBs do not have a large WPA per play. A QB can have huge swings (both positive and negative).

    the WPA for CJ does not accurately reflect his value when compared to a vastly different WPA distribution, though it can help determine his value among running backs.

    Also note: CJ was not merely 'above average', his 2009 season was 41% higher than 2nd place. Not than average, than second place!

    CJ gets 1 million for this season i believe. It is perfectly reasonable in my opinion that he gets paid much more than that. Rumours float around about 30 mill per year, or highest paid player in nfl, that is obviously insane. highest paid RB is a reasonable demand.

  32. Chris says:

    41% higher than 2nd place? By what measure? Yards? Running yards have been proven time and time again to not be a useful indicator of just about anything other than who won in the past. Its not predictive. So, fine if CJ is really that much better than the 2nd best RB, pay the man, maybe. First though, prove to me that CJ is the best RB, because I don't believe it one bit. I bet the best RB is languishing somewhere in the league behind a crappy O-line and none of us know it.

  33. bob says:

    An interesting note on OL.

    in 2009,
    TEN OL had a -0.80 WPA for running (3rd worst in the league) and a -21.9 EPA, also third worst in the nfl.

    also, as Chris noted above, they were only 8-8.

    That is not at all consistent with the alleged items that CJ's yards are only an indicator of who won, or that the best RB is behind a crappy OL (there are only 2 other teams)

  34. zlionsfan says:

    Highest-paid RB is absolutely a reasonable demand for Johnson to make; RB's careers are notoriously short, so he should make use of what leverage he has to get as much money as he possibly can.

    It's also a silly demand for Tennessee to concede. It's one thing to pay someone based on his being the highest-valued RB in football, if indeed he were ... but that is not the same as being the highest-paid RB. You also have to take into account the foolish owners who overvalue their own RBs and the ones who have enough money under the cap to spend more on RBs who maybe don't reflect that in their own value to the teams.

    You should probably keep in mind that of the six players who broke 2000 yards, only O.J. came close to reproducing a season like his best, within 1 YPC of his 2000-yard season and close to his production in yards ... and that was on an 8-6 team that missed the playoffs.

    Dickerson broke 1800 once more and 1600 once after that, both with more carries than his 2105. One team lost their only playoff game and the other missed the playoffs. Lewis never broke 1400 yards again. Sanders was driven insane by Mr. Ford's incompetence, and Davis barely broke 1100 yards in his career after his season. O.J. had 1800+ and 1500+ (better than Dickerson, given the 14-game schedule), but at 5.5 and 5.2 rather than 6.0 YPC, and his last good season was on a 2-12 Bills team.

    If the Titans want to pay Johnson for being the type of back that records seasons like he did in '08 and '10, they're taking a gamble, but it's reasonable. If they want to pay him for being the back he was in '09, it's completely indefensible from the team's perspective: there is zero evidence that he will ever do that again and plenty of evidence that that was not only the peak of his career, but that even the teams that had similarly-productive backs had nothing to show for it from a team perspective.

  35. Anonymous says:

    The Titans don't have anyone else. They don't have a QB. There are a limited # of good QBs. Even if you believe the running game is worthless sometimes you can't do anything about it because you don't have any reasonable way to acquire a QB. In that case you've got to run and play defense.

  36. Anonymous says:

    Blah, Blah, Blah... forget the stats and history lessons... the question is whether or not they should pay CJ and whether or not he is entitled to a contract that pays more than 800k/yr or whatever they've been paying him. Its not just running backs that are overpaid, you can find overpaid players everywhere on both sides of the ball. The Eagles spent plenty of money on one single cornerback... how does he affect this winning success rate crap? The point is the Titans are cheap, they've been cheap and they simply don't want to pay. Everybody here is all concerned about all the money going to CJ and not being able to be spent in other areas. Well where in the hell has the money been going all these past years? IN THEIR POCKETS! They are on the bottom rung of the salary cap and I am quite certain there is enough room to accommodate him... they just choose not to. So once again I will say this... THEY ARE CHEAP BAS***DS and they need to just pay the man...

    By the way, Troy Palamalu (Not a QB) or Big Ben (QB)... when Ben was out, they won... when Ben was in, they won... When Palamalu went out, what happened?.... So who's more important now?

  37. Anonymous says:

    Why has no one suggested that they trade Chris Johnson? They have a rookie quarterback, rookie coach, and the league has shifted away from the running back position being a gamechanger. THIS IS A PASSING LEAGUE. They should trade him to a team dumb enough to waste their money on the position. They can trade for picks and draft some more receivers, linemen, and pass rushers.

  38. Ian Simcox says:

    Bob, let's assume you're right. if the Titans were a 4-12 team that cj made into an 8-8 team, then where should they spend their money? I think I'd beef up the rest of the team, than put all my eggs in one rb basket

  39. Chris says:

    I love the point by anonymous there. They should trade him away for whatever they can get. It will almost certainly have more value than Johnson has to their team.

    Also, its a fair point that it doesn't do them any good to not sign Johnson and also not spend the money elsewhere. However, it does not logically follow that they should then sign Johnson. That's very poor reasoning. What follows is, they should spend the money that they save by not signing him on other positions. As Chase points out in his article, you can't buy help @ QB, but paying to keep your own best OL, DL over time is a much preferable solution.

  40. bob says:

    Ian, I am certainly not advocating paying CJ a large amount (not 30 mill/year, not manning $, that is insane).

    The terms in this article are vague though i.e. 'don't give him a huge contract'.

    I say CJ is vastly underpaid right now (in a 5 yr 12 mill contract). He should get a new contract, therefore i disagree with the title of this article. He should at least be in the range of the top RBs. And that still leaves 110 or 115 mill for the rest of the team.

    My point was mostly about the analysis, and that the WPA SD did not lead to the conclusion reached.

  41. Anonymous says:

    Brian, a minor quibble with your last point. Gregg Easterbrook noted in the last Tuesday Morning Quarterback article that although there is a disparity between the revenues of winning and losing teams, it's often better to be a losing team that doesn't have a large payroll than a winning team that does, from a business perspective. I personally think that CJ doesn't deserve the huge contract he's demanding, but I don't necessarily think that it would make a huge difference business-wise even if they did invest all that money into players that increased wins.

    In all reality it's probably a push between having a star player that sells jerseys, and a winning team that sells tickets.

  42. Ian Simcox says:

    Bob, fair enough. You've given me an idea, to look at the salary data that's available to see whether you get more wins for your dollar with a rb or a qb. The top rb might be good value if he's cheaper than a manning.

  43. Ian Simcox says:

    Well, first signs aren't good for RBs. I've looked at the top 20 (by attempts) RBs and QBs, and compared EPA to Base Salary. On the rank correlation co-efficient, QBs came out at +0.16 (the higher paid QBs got more EPA) - the RBs came out at -0.16 (the higher paid, the less EPA).

    There also seems to be a trend towards younger (i.e. cheaper) RBs in general. 10 years ago, the average age of a 200+ attempt RB was 26.3, and only 6 of 24 of those were 24 or under. Last year, those same numbers were 25.7 and 10 of 23.

    Basically, if I were Chris Johnson and I wanted a job this coming season, I'd be careful not to annoy the owners too much. There's little proof that veteran RBs add any more than rookie ones, and owners/coaches aren't afraid to stick a youngster back there any more.

  44. Anonymous says:

    Jeez... Chris Johnson haters... there would be a ton of teams to jump all over Chris Johnson... just like they would AP and Arian Foster... come on guys, this is silly... my beef is with the Titans organization... they don't want to pay him and they don't want to trade him, but yet still... they are making no efforts to try and come to some kind of agreement... who in the world would want to play for such a vindictive team... look how they've treated their players in the past... absolutely no loyalty and no gratitude towards their players... I hope they lose every single game this year and I hope the fans boycott leading to blacked out games.

  45. Ian Simcox says:

    We're not Chris Johnson haters. Anyone can see watching him that he is a good running back. Our problem comes from the fact that in general, all RBs are overvalued. The megabucks all star RBs don't perform any better than the rookies these days and for all the millions spent on them, they don't contribute any more to winning than a no name RB does (case in point, had anyone heard of Arian Foster before last season?).

  46. Brian Burke says:

    Well put. I LOVE CJ. He won me big $ carrying my fantasy team to a 2009 title, with a little help from Aaron Rodgers. (Go Pitbulls.)

    I remember the BAL-TEN playoff game in 2008 very well. During the first half, CJ gashed my Ravens' D mercilessly. He made one of the top two defenses of the year look like they were in slow motion. I made a mental note to draft him in the first round next season. Somehow he got hurt and didn't play after halftime, and BAL barely squeaked out the W.

  47. bob says:

    Ian, one possible conclusion of your analysis "the RBs came out at -0.16 (the higher paid, the less EPA)." would be that RBs are used to maximize win probability, rather that to maximize expected points per play. i.e. when a team is ahead on the score board, they run the ball to kill the clock, rather than attempting to score more points.

    Perhaps the better runners are used in this manner more often, icing the game rather than scoring points.

  48. Brian Burke says:

    That's a good point, Bob.

    I'd bet that the difference has to do with rookie contracts. If we only looked at performance under FA or (non-rookie contracts) and reran the analysis, the result might correlate better with pay.

    I'm sure a lot of really good RB performance is coming in early career years under very inexpensive contracts, just like CJ now.

  49. Anonymous says:

    Isn't age a problem? CJ turns 26 next month, so one would expect maybe 2 good years before the inevitable statistical decline. Why would any team sign him to a long-term contract unless it was very front-loaded. I would suspect Brian is right that a lot of really good RB performance is under the rookie contract. This probably makes the rookie contracts unfair for RBs or other positions with a short longevity in the NFL. RBs may be unique in that technique and experience almost never make up for pure physical ability, which starts declining as the rookie contract ends.

  50. Anonymous says:

    all of you who say 'forget the stats' and then launch into your own, lazy, uninformed opinions should shut up and try and learn something. all pro teams use very intricate statistical analysis to make many of their decisions.just because you can't understand them, don't try to blow past a real argument, keenly made with insightful statistics and on you own intelligence, you stupid,moronic fools. Don't pay CJ! Its and idiotic move to make from an organizational standpoint.he's just not worth it.

  51. Jonathan says:


    The Titans just signed him to something like 13 million per year for four years. Good luck building a championship team with that albatross.

  52. Anonymous says:

    Pay that man hish money.... ;)

  53. Anonymous says:

    Reading the thread of comments since your article was posted, I am amazed, as always at peoples failure to separate reality from emotion. CJ has brought NOTHING to the Titans organization to improve the chances of winning games. It doesn't seem to really be an arguable point either. In the fantasy football world, CJ has been a workhorse. But last time I checked, a coach just got run out of town because he couldn't win enough...with CJ the last few years. You can't have it both ways. Bottom line: an Eddie George-style 85 yard performance is worth more than 150 from the win column, that is.

  54. Andy H says:

    Revisiting this after the end of the season, this article seems prophetic. Outstanding analysis Brian.

    One change the Titans made was a shift to more passing. Using the first four games as a sample, it appeared as though the Titans would be throwing the ball more and rushing less than 2010.

    L vs Jax - 34 Passing attempts, 12 rushing 26%
    W vs Bal - 42 Passing attempts, 29 rushing 40%
    W vs Den - 36 Passing attempts, 14 rushing 28%
    W vs Cle - 20 Passing attempts, 24 rushing 55%

    Average rushing attempts:
    2011 First four games - 37%
    2010 Season - 46%

    When the 2011 season concluded, the Titans attempted to run the ball 39% of the time. We can speculate as to why the Titans made this move. I suspect it likely has to do with Brian's article.

    CJ defenders have blamed his 2011 performance, or lack thereof, on the offensive line. One way we can measure if this is an accurate claim is to look at sacks and QB hits. Combine these and we can call this category "Offensive Line Failures."

    2010 Season - 73 OLF (27 Sacks, 46 QB hits)
    2011 Season - 79 OLF (24 Sacks, 55 QB hits)

    The Titans' offensive line was surprisingly consistent. Sorry CJ defenders, but as Brian assessed in the beginning of the 2011 season, Johnson was overvalued.

  55. Andy H says:

    One footnote...

    2009 Season - 59 OLF (15 Sacks, 44 QB) hits

    This was Kevin Mawae's last year as the Titans' Center AND CJ's record breaking season.

    We can probably say that the Titans' Offensive Line was responsible for making CJ look better than he actually was.

  56. scp1957 says:

    "It's the steady threat of 4- or 5-yard gains that opens up the passing game, not the occasional break-away TD run that makes a winner. It doesn't matter that you hit 50 HRs when you strike out on 62% of your at-bats."

    In this regard, I prefer to consider the effect on opponent's scheme. Barry Sanders routinely faced 8- and 9-in-the-box. When CJ does, then I'll buy his demands for re$pect; not before then.

  57. Anonymous says:

    Here it is 3 years later and what have the Titans done with all that money they gave to CJ? Three years of utterly banal football is where it went. The Titans then asked CJ to kindly leave the building and not make a scene. Now the Jets' OC Marty Mornhingweg calls CJ a 'first ballot HOFer' although they seem to be paying him a more reasonable market value of $2mil.


Leave a Reply

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