Wednesday, September 10, 2008

"The Blind Side:" I still can't figure out why left tackles are so valuable

Yesterday, Steven Levitt recapped "The Blind Side" over at his Freakonomics Blog. Levitt points out that while quarterbacks are the highest paid NFL players, at about $5 million per year, left tackles are second, at $4 million. He writes,


"The book highlights how N.F.L. teams only slowly became aware of the immensely important role that offensive left tackles play in protecting the quarterback from blind-side hits."


But I still have a problem with this idea of Lewis's, that left tackles are so valuable because they save the quarterback from injury. How many quarterbacks get injured? Say, 25% a season? (That seems high to me, but let's give the left tackles the benefit of the doubt.)

And of those 25%, how many were injured because of a blind-side hit that the left tackle could have prevented? Again, let's say 25%.

25% of 25% is about 6%, so about one quarterback in 16 can be saved by the left tackle.

But, now, what's the difference between a typical left tackle and the replacement-level left-tackle? Let's suppose the average left tackle is twice as good as preventing QB injury as the backup. That means that the good LT could save only 3% of the quarterbacks under his care.

Now, 3% of a $5 million quarterback is only $150,000. (Of course, some injured QBs are out more than one season, so you'd have to multiply the $150K by the number of seasons. But many injuries are for only a few games, so it seems reasonable to figure one year as the average.)

So how can you argue that the LT is worth $4 million because of only $150,000 in quarterbacks saved? Obviously, if left tackles deserve to be the second-best paid position in the NFL, they must be doing something else right.

Maybe it's sacks. But as I wrote two years ago,


"The Hidden Game of Football" (p. 104) estimates that the difference between a good quarterback and an average one is four completions per game. An eyeballing of 2005 NFL sack statistics shows that the difference between a good individual sack total and an average one is only about four sacks per *season*. That’s a difference of 1500 percent in favor of the quarterback."


So if a good QB is worth $5 million a year for those 60 extra completions, then a good LT should be worth only about $350,000 more for those 4 saved sacks.

So $150,000 in injuries saved plus $350,000 in sacks prevented adds up to only $500,000. Where's the other $3.5 million?

Well, there's also the unmeasured effect of giving the QB more time in the pocket. But that would have to be huge, wouldn't it, to add up to $3.5 million? You'd need the LT to be responsible for 2 or 3 extra completions per game (or the equivalent in yards). That seems unlikely, considering that the rest of the offensive line has a big stake in the outcome as well.

The bottom line is: I can't really figure out why left tackles are considered as valuable as they are.

One possibility: the famous career-ending injury to Joe Theismann came from the blind side. Could that
gruesome example of what can happen to a blind-sided quarterback have caused GMs to overestimate the value of the left tackle's job?

(UPDATE: "Doc" points out in the comments that it's not the full $4MM we have to explain, just the premium above other offensive linemen. He's right. That premium appears to be only $1MM or so. So it's $1 million we have to explain, not $4 million. But, still, it's hard to see how that million dollars is just for the value of protecting the QB from injury. It seems more likely that only a small portion of that has to do with injury, and the rest has to do with LT being the most important blocking position, and teams putting their best lineman there.)

Labels: ,

13 Comments:

At Wednesday, September 10, 2008 11:44:00 PM, Anonymous Anonymous said...

Have you given any thought to the running game? Those tackles do a lot more than help keep the pass rush off the QB. :)

 
At Wednesday, September 10, 2008 11:46:00 PM, Blogger Phil Birnbaum said...

Very true. But wouldn't the entire offensive line have equal responsibilities in regard to the running game? Would the left tackle be especially important there?

Serious question, I don't know football that well.

 
At Wednesday, September 10, 2008 11:51:00 PM, Anonymous Anonymous said...

I meant to add more: these guys are typically considered the best, so in addition to their ability to keep the rush off of the QB, they are likely the primary path to run behind.

See Jonathan Ogden as one example.

So, yeah, looking at it from simply the pass rush doesn't give us the full picture. But if you have a guy that is good with the pass rush and a great bulldozer to run behind then maybe that is a small market and you pay those guys well.

Something to consider at least.

-------------

Just saw your response on preview, and I'd add that yes the play calling will typically go in that direction. Not sure of the numbers (I thought I saw them some where but can't seem to find them now), but there are 11 guys on the field so everyone has some sort of responsibility. These guys can help lead the way.

 
At Thursday, September 11, 2008 12:45:00 AM, Blogger Phil Birnbaum said...

OK, so let's suppose the LT can help gain rushing yards. How many yards would that have to be for the LT to be worth the money?

Let's say the rushing help from the LT is worth 70% of the quarterback's salary. That means it's the equivalent of 3 completions. How much is the average completion? Suppose it's 8 yards. Then the difference between the average LT and the bench-warming LT has to be 24 yards rushing per game.

Seems like a lot for just the LT. Especially considering that you don't always use the LT as the primary bulldozer, right?

Maybe someone can help quantify the rushing factor.

In any case, it still seems to me like the "protect the quarterback from injury" story can't account for that much of the LT salary premium.

 
At Thursday, September 11, 2008 5:13:00 AM, Anonymous Anonymous said...

How is it that the 'good' QB is getting those extra 4 completions per game? Part of the reason is that they have time to find open receivers. The o-line is obviously critical in this regard, and the LT is going to be the best pass blocker on the team.

So part of the credit that HGoF gives the QB might belong elsewhere, such as with the LT.

I agree LTs are probably overpaid in proportion to their actual contribution. But keep in mind that's not how markets often work.

How much would you offer to re-sign your team's starting shortstop if you've got a prospect coming up through the minors who might be just as good? Probably not much. You'd offer a lot more if there is no one waiting in the wings and not many FA SSs available. So a player's value over replacement is just as important as production share in determining pay.

Every team needs a good LT, but they're really scarce compared to other positions like LB, RB, or WR. There is a quick drop off in ability after the top few guys.

High demand and low supply makes for an overpaid LT.

One other quick thought. The LT is usually simply the better of the 2 starting tackles on a team. So comparing LEFT tackles with other positions might be (really) unfair. For example, compare LT salaries with those of just the best cornerback on each team, or the best WR etc., whether left or right. I'd bet the average salaries among positions would be much closer.

 
At Thursday, September 11, 2008 8:39:00 AM, Blogger Phil Birnbaum said...

Hi, Brian,

1. Good point about the LT just being the better T. I bet you're right.

2. You're suggesting that replacement level for the LT is lower than replacement level for QB. That could explain it. I wonder if there's any way to test that?

3. Agreed that the offensive line accounts for a proportation of the QB's stats (much as fielding affects a pitcher's stats in baseball). And you'd think it would be a HIGH proportion of the QB's stats, given that the offensive line (in the aggregate) seems to be paid a lot more than the QB.

4. You write, "But keep in mind that's not how markets often work." Absolutely. I guess that's my point ... the Michael Lewis theory is that LTs are paid so much because they protect the QB from injury. Looking at the numbers, that doesn't seem to be right.

 
At Thursday, September 11, 2008 11:48:00 AM, Blogger Don Coffin said...

It's not the entire $4 million that needs t be explained, it's the *premium* paid to LTs that needs to be explained. Your rough estimate is a premium of $350,000. How's that compare to the LT premium, the difference between LTs and other offensive linemen?

 
At Thursday, September 11, 2008 12:00:00 PM, Blogger Phil Birnbaum said...

Geez, that's right. I should have caught that.

According to the Massey/Thaler regression, defensive linemen make more than offensive linemen, by about $1 million. Since we know LTs make more than the average defensive lineman, the premium must be at least $1 million.

From the same link, but the comments: Offensive tackles make $3.17 million (average of left and right). Offensive guards make $1.94MM. The center makes $1.65MM.

So the premium, again, is at least $1 million.

And, of course, it's possible that the $1 million is the right number for other reasons. Maybe only $150K goes for protecting the QB from injury, $350 goes for sacks, and the other $600K goes for general blocking tasks.

But, in any case, the narrative that the LT position makes the most money because he protects the quarterback from injury is still questionable.

Thanks for the catch, will update the post.

 
At Thursday, September 11, 2008 12:26:00 PM, Anonymous Anonymous said...

Phil, you've inspired me. I did a little research using 2007 salary data from USAToday.com. It turns out that the top 32 WRs and the top 32 CBs actually make more than the top 32 OTs (left or right). I haven't even looked at other positions yet, so there could be more positions higher paid than OT.

Further, only 33 of the 100 highest paid offensive linemen are tackles. There are lots of very highly paid guards and centers. I was really surprised.

(This is based on "cap charges," which combine base salary and amortized signing bonuses.)

I'm inclined to think the whole LT Blindside thing is really overstated.

 
At Thursday, September 11, 2008 5:24:00 PM, Blogger kds said...

If the "blindside theory" is correct, then we should see more highly paid right tackles on teams with left handed QBs. Could someone test this?

 
At Thursday, September 11, 2008 5:34:00 PM, Blogger Phil Birnbaum said...

kds: Steven Levitt says he tested it (in the post linked to above). Still, it could be that there's a left/right salary difference for all offensive linemen based on the quarterback's handedness.

 
At Thursday, September 11, 2008 11:46:00 PM, Anonymous Anonymous said...

How much is each second a quarterback has in the pocket worth in terms of completion percentage and yards?

That would be a better measure of value than simply keeping the quarterback healthy

Also, a team will rotate players at other positions to maximize production based on situation and possibly fatigue: third down running back, goal line defense, two tight ends, five wide recievers, two to four linebackers. But no matter what, they want their five OL to be on the field every offensive play

 
At Sunday, October 05, 2008 2:58:00 AM, Blogger colin said...

Don't they talk a lot about how the best tackles are physical freaks in Blind Side? I think you need to be doing more to look at the actual distribution of talent at the tackle position. Scarcity would certainly drive the price up.

 

Post a Comment

<< Home