LiveBall Sports

Value over replacement is a concept that has long been a staple of baseball sabermetrics, even as the effectiveness of it’s measures have vastly improved in recent seasons.  In football, however, measures of value against replacement are far more confusing and open to interpretation: subjective, if you will.  One thing we all can agree on is that the concept is no less real or important to football analysis than it is to baseball analysis.  The only differences are in the effectiveness of common representations and calculations of a replacement level figure.  Football people have to be able to apply concrete evidence of performance to an abstract interpretation of what is replacement level.  If mistakes are made, there will be arbitrage opportunities for shrewd teams.

The main problem in football is that replacement level varies by position, by year.  In baseball, it can be stated that a replacement level offensive player is one who produces at 80% of the league average.  This is probably the most common representation and will never be a downright improper assumption.  There’s no such shorthand method in football.  With no minor league system in football (yet), a freely available talent is no different from one who is out of work.  From a practical standpoint, it’s really difficult to be out of football one week, and starting for an NFL team the next week, contributing at a level that is anywhere near replacement.

A better operational definition of a replacement level player is a backup at any position who offers no additional value to a team past the fact that he has some experience being that team’s backup.  Teams who are successful in the NFL year after year often employ a “next man up” type of attitude to injuries.  These teams can take replacement level players, apply them to a defensive or offensive system, and get performance out of them that is inconsistent with conventional (even sabermetric) methods of replacement estimates.  It’s my opinion that we have to accept this as a necessary evil: replacement level players aren’t always going to perform at replacement.  This is true of baseball as well.

NFL free agency is really our one opportunity during the football calendar year to get a good wide-view picture of how teams value replacement level talent.  In particular, what I’ve been concerned with is looking at which positions get most drained by teams who need to stack their roster with players who have experience at the position.

What I’ve found by studying these trends has been pretty eye opening.  My findings support the previous assertion that the league-wide idea of replacement level is much, much lower compared to the average for some positions than others.  Take cornerback, for example.  Corner is a position that I would describe as being “highly skilled”.  It takes corners longer to develop as players than defensive lineman or linebackers.  But when we apply the concept of replacement to the position, we find that, despite the refined skill needed to excel, players who are very close to the league average (but on the wrong side of it) can bounce around the league just like a replacement level player would.  Look no further than Titans/Browns/Bears/Cardinals (just in 2009) corner Rod Hood, formerly a very successful 2nd or 3rd corner for the Eagles, couldn’t hold a job for more than two weeks.  It’s not just guys who peaked as no. 2 types.  Look at Carlos Rogers, or Bryant McFadden, or Dre Bly, or Shawn Springs, or Brian Williams.  At some point, all of these guys were bona fide no. 1 corners on their teams.  But in 2009, every one of them posted a replacement level type season.  It’s really just the nature of the position: players are quick to rise, and quick to fall.  And once a player drops to consistently below average at the position, he’s easily replaceable.  The freely available talent that can provide teams with a close to average level of play is everywhere.

On the other end of the spectrum, I have found, lie wide receivers and linebackers (yes, linebackers).  Wide receivers with any history of recent success get sucked up off of the open market onto teams within the first few days of free agency every year, and despite this, you still see incredibly high variances in the quality of backup receivers around the league.  After Terrell Owens, Antonio Bryant, and Josh Reed sign (probably later this week) for good money, any team in need of a receiver will have their choice between 37 year old Mushin Muhammad, Mike Furrey, or Marty Booker.  The quality of receiver that comes at a minimum priced contract is so far beneath the level of an acceptable NFL starter that teams almost have to look to the draft to add players who can play special teams and catch passes.

At linebacker, there’s a reason that Jeremiah Trotter gets dragged out of football purgatory every other season.  Replacement level linebackers are plentiful, but those ‘backers who can both lead and go make tackles never are allowed to hit the open market (and when they do, they get Bart Scott/Karlos Dansby type contracts).  The OLBs have, admittedly, been a lot slower to sign than in recent years, but Keith Bulluck and Joey Porter aren’t exactly going to have to sign for one year “prove it” contracts.  Scott Fujita, Tully Banta-Cain, and Mike Vrabel have been well taken care of.  Vrabel, in particular, isn’t anywhere near an average player.  If he played corner, he would have had to retire three years ago.  Heck, Ty Law has been bouncing around unable to hold a job with a single team since 2006, and he was once a much better player at his position than Vrabel was at his.

You may think of quarterback as a highly skilled position where all the best talent gets bought up as soon as it hits the market, but this is simply not the case.  Jake Delhomme and Derek Anderson were starters as recently as last season, but the odds on either of them catching on somewhere as a backup next year are pretty much even.  Other replacement type talent like Patrick Ramsey and Jeff Garcia could start the year on teams, even though they did not finish last year on teams.  The backup QB market is pretty interchangeable with the former starters market, performance seemingly has less to do with it than coach’s preference for one type of player over another.  Such is the life of the former pro-bowler who can only play at -15% of the league average.  At that level of performance, your career needs an “in” to continue.

But one position trumps all the others in the market’s recent interpretation of replacement.  And when I say “trumps,” I mean it’s so glaring that its’ clear that this position functions differently than all the other positions in NFL free agency.  Chad Clifton of the Green Bay Packers just got a 3-year $20 million dollar extension to remain the Green Bay Packers’ left tackle for at least one more season (he’ll receive more than 40% of the total contract value just for this season).  Clifton is 34, a one time probowler in 10 NFL seasons.  Upon becoming a free agent, Clifton visited the Redskins, who had just watched their offensive tackle of Clifton’s draft class (Chris Samuels) retire after 6 pro bowls.

The difference in skill between Samuels and Clifton is roughly equivalent to the difference in skill between Eli Manning and Kerry Collins.  But the Redskins, desperate for a tackle, wanted to seriously consider letting Clifton be the franchise’s left tackle for the next two years in the twilight of his career, that is until they couldn’t beat the best offer of the more desperate Packers.  Why was Clifton, a below average player, worth so much to these teams?  The answer is that, despite being decidedly below average, Clifton was the only left tackle who hadn’t been locked up by their team ahead of time who hadn’t fallen into the dark depths of the replacement.  After him: Levi Jones, Damion McIntosh, and Barry Sims sit on the market, unwanted.

Across the line, the right tackle market has gotten absolutely preposterous.  The prized lamb in the free agent class was San Francisco’s Tony Pashos, who also visited Washington, couldn’t agree to terms, and ended up signing a 3-year contract with the Browns.  Just one year ago, Pashos was an incumbent right tackle on a Jacksonville team that felt it prudent to spend it’s first and second round draft choices on offensive tackles, as well as to sign a veteran LT.  Pashos has never played a position on any line besides right tackle.  But seemingly by virtue of being a free agent when every other team has jumped through hoops to seal their top two tackles if they ever had it, Pashos had the NFL world at his feet, just one year after being completely unwanted.

Pashos and Clifton will both be among the vast landscape of the replacement level lineman before these contracts expire, but desperate times call for desperate measures.  There is a clear shortage of capable starters at the offensive tackle position in the NFL.  While Albert Haynesworth and Julius Peppers have hit the market in consecutive years to sign elsewhere for mega-bucks, teams can hardly find anyone capable of protecting the quarterback.  And so they are hiring players who might be able to help by the dozen.  Cornell Green, a former Raiders lineman best known for his inability to stop anyone, signed with Buffalo for 3 years and 8 figures.  Green is the very definition of replacement level tackle, although apparently the Bills feel otherwise.  The Raiders felt the need to send 2009 FA signing Khalif Barnes to the bench after a foot injury and underperformance derailed his 2009 season.  This year, he’ll be back under a one year contract, and if the Raiders can’t address the need in the draft, he’s in the starting lineup.  Khalif Barnes has very nearly underachieved his way into a promotion.

Players at the offensive tackle position in the NFL don’t have to be anywhere near league-average to command a multi-year contract on the open market.  In fact, those who are league-average or close to it do not get anywhere near unrestricted free agent status.  It seems like it’s only a matter of time until NFL teams start kicking underutilized backup interior lineman out to the tackle position and letting them compete for playing time as starters.

Rod Hood (age 28), who in the most conservative estimates, offers play at -10% of the league average cornerback, has been on four teams since playing in Super Bowl 43 with Arizona.  Chad Clifton (age 34), who in the most optimistic of estimates, offers play at -10% of the league average LT, is offered a 3-year contract by two different teams and gets $8 million to play this season.  When you consider that corners take longer to develop than offensive tackles, we’re left with only two reasonable explanations: either NFL decision makers act completely irrationally, or a league average offensive tackle offers ten to twenty times more value over replacement than a league average cornerback.

And that’s what’s astounding in the NFL labor markets today.

There’s a pretty good article in yesterday’s San Francisco Chronicle, guest written by Houston Chronicle columnist Richard Justice.  The subject: David Carr’s career, in wake of his signing with the 49ers to “compete” for the starting job.

If there is a fault in the analysis performed here by Justice, it’s that he basically exonerates the Texans from overdrafting Carr at first overall in the 2002 draft (my analysis on the class can be found here).  Justice’s sharp criticism of the work done by Charley Casserly and staff between the Carr selection and the disaster of a season in 2005 is pretty much on point.  Of course, the fact that the Texans bucked conventional wisdom and made the best draft pick in their short history just a year after Carr was drafted, giving him a superstar wideout in Andre Johnson via the third overall pick, is ignored.

Carr’s developmental career path was completely normal through the first three years of his career: he was never going to be a franchise passer in the truest sense of the term, but at the conclusion of his best season–to that point– in 2004, Carr was already an above average passer.  Above average is very valuable in NFL dollars, and Carr could have passably fulfilled his billing as a first overall draft pick if he had stayed at that level.  But by the end of the 2-14 2005 season, the Texans were just looking for some way they could salvage the David Carr investment,  and at that point, they probably wasted an 8 million dollar roster bonus (in Carr’s rookie contract) trying to salvage what value had already been lost.

As far as I can tell, the only thing that actually changed in 2005–from 2004–is that the Texans ended up succumbing to their offensive line weakness.  Sure, Carr’s play was a limiting factor on Andre Johnson that year, but no more a limiting factor on Johnson than in both 2004 and 2006 (this is evidenced by the emergence of Johnson into a top level WR as soon as Matt Schaub took over at QB).  In 2005, the Texans could run the ball just as well as they could before, probably even better.  Dominick Davis (Williams), who would not play another snap after 2005, was every bit as productive that year as he was in 2004.

More to the point, Justice is criticizing the leadership of the Houston franchise for allowing the offensive line to reach offensive levels.  In 2004, David Carr was sacked on more than 10% of his dropbacks, and no observer batted an eye at it.  The line was fine in 2004, and that number kind of typifies David Carr as a passer.  In 2005, the two guards from the 2004 team ended up combining for 28 starts at the tackle positions.  The 2005 Texans, despite having a QB and a WR drafted in the top three picks of their respective NFL drafts, were a team that were not built to throw the football.

The fact that they had to throw the football was more due to a complete and utter defensive collapse than anything wrong with Carr, Casserly, or the offensive coaching staff headed by Chris Palmer, but the suggestion by Justice is valid: why invest two top picks exclusively in a passing game, and then not build a team to throw the football?  But a more intense look at the draft patterns of the Texans suggests that the Carr project might have been doomed from the start.

Consider: the chances of the Texans landing a player as dynamic as Andre Johnson at any point in their first four drafts was not very good.  Maybe as high as 30%.  Well, at the point that the Texans landed Johnson, they were one pick into their second draft, and had already landed 4 above average NFL players (including Carr, at the time): Johnson, guard Chester Pitts, and receiver Jabar Gaffney.  Later that draft, they landed tackle Seth Wand and Davis/Williams.  At this point, the Texans turned their draft attention strictly to the defensive side of the ball for the duration of the next two years.

This is pretty much Justice’s point.  The Texans ignored the offensive side of the ball in two straight drafts.  This caused Carr to not progress as expected, it caused Johnson’s breakout season in 2004 to be followed by two largely disappointing years before his 2007 14.1 average yard per catch (injury shortened) season.  According to Justice, the Texans’ owed it to Carr and they owed it to themselves to keep putting talent on the offense, and after consecutive years of adding only defensive players, the management of the team got what they deserved when the team went 2-14.

The more important thing to get out of this is just how hard it is to build a team around a quarterback, even if you pick up the “best” quarterback in the draft.  As an expansion team, the Texans could have picked Julius Peppers, the consensus best player in the draft, or they could have gone for the quarterback and start the building process from there.  Carr’s tragic flaw is, and will always be, his desire to want to hold onto the ball for a second too long.  Because of that flaw, he’s a tough choice at No. 1 overall.  But he’s not by any means a horrible pick: Carr gave Houston a bunch of above replacement seasons that they could not have gotten from quarterbacks off the scrap heap.  It just so happens that 2005 was not one of those seasons.

Here’s the moral of the story: teams that are considering picking a quarterback in the upcoming couple of drafts need to understand why the Texans won 5 fewer games in 2005 than in 2004.  Because of the spot where David Carr was drafted (1st overall), his contract, and the Texans’ place in the league after the 2004 season, the team had no choice but to turn Carr loose and to throw him to the wolves, so to speak.  The results were positively disastrous.

Compare for a second the amount of homegrown talent on the 2-14 2005 Texans offense, to that of the 11-5 2005 Bengals offense.  The Bengals spent two first round picks in the top ten on offensive tackles.  The Texans: none.  But the homegrown talent on defense is pretty similar.  The difference is most stark when you look at that Bengals team, and see all the talent that was drafted–by the Bengals–prior to Carson Palmer being the first overall pick in 2003.  This list involves both tackles, the top two receivers, and the running back.  Palmer was the part that brought it all together for Cincinnati, making the Bengals offense of 2005 and 2006 one of the best units in the NFL.

Teams considering Sam Bradford at the top of this draft should best be doing a self analysis to see if they have the parts of their future team already in place.  If you’re the St. Louis Rams, you have two first round tackles on your roster.  Can Alex Barron and Jason Smith be your tackle tandem of the future?  What about Donnie Avery?  Does he have Chad Ochocinco type breakout potential at wide receiver?  Can you land a competent guard, tight end, or second wide receiver in the second round?  If your best self-analysis measures answer yes to all these questions, perhaps Sam Bradford is the piece who will bring all the parts together.  If Barron, and Avery are not parts of the next strong Rams team, then maybe picking for your need at quarterback over value is going to be repeating the same mistake the Texans made.  If Avery and co. aren’t NFL type talent, Bradford will probably suffer the same fate as Carr when thrown to the wolves.

If the Detroit Lions have an offensive breakout this year, it will be because of their decision to make Matt Stafford the part that brings all of their offensive pieces together: first round tackles Jeff Backus and Gosder Cherilus, first round receiver Calvin Johnson, and first round TE Brandon Pettigrew, and homegrown RB Kevin Smith.  If the Lions do not break out this year on offense, it will be due to the fact that they overestimated the usefulness of players like Cherilus and Calvin Johnson and Pettigrew in developing a young quarterback.  The Lions have a high percentage of Detroit’s GDP hinging on the fact that Stafford’s development will not stall.

The 2005 Texans teach us that an expansion team, and teams in similar position, need to be focused on adding talent above need, because even a well-researched, easily-defensible pick at a need position can and will go awry if a team runs out of goodwill and needs to throw it’s investment to the wind in hopes of winning games.  The Texans Proverb:

Lucky is the team that is merely a quarterback away from being a winner, as their problem is easily solved.  Unlucky is the team who has a quarterback and nothing else, for their fans should never know that they had a quarterback at all.

Alex Smith, meet David Carr.  You have much to learn about each other’s plight.