Writers of Pro Football Prospectus 2008

12 Jun 2008

FO on ESPN: Losing Ogden Shouldn't Hurt Ravens

With the book behind us, weekly FO columns on ESPN.com return. This week, we look at teams that lose a star offensive lineman. The numbers were pretty surprising -- as a group, teams that lose a star lineman don't tend to decline. For every team like the Chiefs of the last couple years, there's a team like the 2000 Vikings or 2006 Saints. Adding a top lineman? That's a different story, one that will give Jets fans a reason for optimism.

Posted by: Aaron Schatz on 12 Jun 2008

24 comments, Last at 17 Jun 2008, 10:32pm by TonyAngelo

Comments

1
by MJK (not verified) :: Thu, 06/12/2008 - 5:41pm

I still wonder about the contention that O-lines play better when they have continuity. It's a chicken-or-the-egg thing. I still haven't seen anything to convince me that, instead, good players tend to be retained and hence there is continuity.

As far as the general conclusion, that losing an elite O-lineman doesn't hurt as much as adding one helps, I think I can see a reason for that.

Usually, when a team is going to lose an O-lineman, it's to free agency or retirement, both of which the team can usually see coming, and they put a contingency plan in place to mitigate the loss. E.g., the Colts drafting Ugoh. On the other hand, when a team lays out big money to acquire an elite O-lineman in free agency, it's usually because they need one, and so naturally getting one and filling a need helps them improve.

I.e. the team losing the player tries to mitigate the loss, and since they have foreknowledge, they often succeed.

The team gaining the player obviously doesn't try to mitigate the improvment--improving the team is the whole point of trying to gain the player in the first place.

2
by Yaguar (not verified) :: Thu, 06/12/2008 - 6:09pm

The concept of "mitigating an improvement" makes my head hurt.

3
by Brian (not verified) :: Thu, 06/12/2008 - 6:36pm

"Football Outsiders research has shown that offensive lines play better when the five linemen have continuity and experience playing together."

I don't think the data shows anything more than the fact that 1st string linemen are usually better than 2nd string lineman.

The continuity effect may or may not be real, but until you can pull apart the tangle of individual talent levels on an offensive line, no quantitative research could ever hope to prove it.

4
by TrollinForJoes (not verified) :: Thu, 06/12/2008 - 6:43pm

Mitigating improvement is easy -- just take #1 and #4 draft picks in consecutive years (improvement) and select a "rocket-armed" ("bad") quarterback, and a "running back" ("fungible player") with them (mitigation).

5
by PaulH (not verified) :: Thu, 06/12/2008 - 8:01pm

I've long since wondered the same thing about offensive line continuity.

I'm sure continuity helps to a degree, but it's pretty obvious that lines only stay together for a long period of team when they play at a high level. If you don't do that, you see a lot of shuffling.

I really don't think there is anything that special about continuity. I highly doubt you can really turn a bad-to-serviceable offensive line into something good-to-really-good just by keeping them all together on the field.

6
by Becephalus (not verified) :: Fri, 06/13/2008 - 12:02am

I would think it is obvious that continuity is a good thing, but that it is incredibly unclear how important t is and if it is basically meaningless noise compared to other factors.

I know it is a completely different sport, but I know in hockey I had some of the best chemistry and got the most out of my ability playing with some people over 5 or 10 games, while people I had played hundreds of games with lack this chemistry.

I think there is a lot more to the subtle interactions of player style than is first obvious in hockey, and I would assume line play to be identical in this respect.

It would be at all surprising to me to see continuity be dwarfed by other considerations.

7
by Becephalus (not verified) :: Fri, 06/13/2008 - 12:04am

"would not" in that last sentence, god I am a poor typist.

8
by langsty (not verified) :: Fri, 06/13/2008 - 12:20am

I think the whole continuity thing is a little bit of a red herring. One of the most important elements of good o-line play is communication, because a line that communicates well will rarely have breakdowns in its protection schemes. You can't measure 'communication' in any meaningful, quantifiable way, but you CAN track continuity, and I suppose it can be assumed that continuity is what allows effective communication to develop.

What you consistently see with the league's best lines is that they're better than the sum of their parts. Each of New England's five starting offensive linemen can be beaten individually, none of them are dominant* players. But they're technically sound, have good study skills and work well as a team, which is the most crucial part. It also helps a lot that they were scouted for that system and developed in it (this is where continuity may come in) because they're in a good situation where their skill-sets are best put to use. And of course Scarneccia deserves a lot of credit for his coaching and the diversity of his blocking schemes.

(*btw, saying "they're not dominant" doesn't mean you can just field a bunch of scrubs for a long time and wait for them to 'gel' or something - you still need hard-working, skilled players who fit your scheme.)

9
by nugent (not verified) :: Fri, 06/13/2008 - 7:41am

It's footwork.
Cohesion.
Coreography.

Continuity is absolutely critical.

10
by Pete (not verified) :: Fri, 06/13/2008 - 7:48am

I believe that continuity in most of the offensive line can be an issue. Replacing a star OL with a backup who has played with the team and scheme for several years is not that big a shift.

I think a bigger difference is whether a team forces a new blocking scheme (and terminology) on a team. If most of the OL (and team) are using the same (good, or at least well-suited) scheme then changing a couple players (other than bringing in a brand new QB) will bring continuity. Most of the players and coaches (again, continuity of the coaching staff and scheme) will KNOW what is expected and they just need to do what they have practiced.

11
by Parker W. (not verified) :: Fri, 06/13/2008 - 9:29am

Considering the teams that improve after adding a top lineman:

The added player definately brings a higher level of play that attributes to team success (improved levels of success) but I think the high level of improvement can largely be attributed to the teams new "focus" on improving an offense. I realize there is a two game swing here and that an improved offense does not necessarily mean more wins, but this issue was largely addressed by pointing out that teams gravitate towards 8-8 and that fact that teams went from 6.2 wins to 8.4 when adding the linemen only highlights that idea.

So the improved offensive numbers could much more likely be the result of a new shift in philosophy towards improving the offense. Consider, most teams that go out in free agency and add the big name linemen are the ones who had a horrible offense the season before and believe by adding a big-time lineman, they can immediately improve their offensive performance. And maybe they can. But it is probable that they made a number of other moves in order to bolster their O.

12
by Ted Max (not verified) :: Fri, 06/13/2008 - 9:41am

# 1, 5, 6, and others hit the nail right on the head:

The causal flow is probably the other way around. Good offensive line players are kept together for long periods of time, while bad offense lines get mixed up a lot as teams try to fix them.

"Continuity" is one of those terms like "Consistency" (see firejoemorgan.com for some fun with that one) that SEEMS like it should matter, but is incredibly hard to prove, and can often just as easily be simplified to "Having good players that are playing well helps teams win."

Here's a sort-of timely reference: Before this year, the Lakers kind of sucked for a while after they let Shaq go. Is that because losing Shaq hurt their team "continuity"? Or is it just that losing good players hurts teams, and teams without very many good players do a lot more shuffling to try to fix it?

13
by Cathedraticum (not verified) :: Fri, 06/13/2008 - 11:00am

Bravo #4 for the subtle prod to get a tirade from RaiderJoe. Sometimes the Great Ones need that little extra bit of motivation and we are all better off for it

14
by The McNabb Bowl Game Anomaly (aka SJM) (not verified) :: Fri, 06/13/2008 - 11:01am

There are two major problems with this article. One is that the before and after numbers for teams losing or gaining a linemen are not adequately compared to league averages. It looks like the improvement in teams gaining a linemen might be mostly regression to the mean. How do teams who average 6.2 wins and don't add a top lineman fare the next year? These numbers require context.

Secondly, the numbers in the first box are judged to be not significantly different, while the numbers in the second box are. What is this judgment based on? Eyeballing? Is it that hard to open SPSS and run a one-way ANOVA? And please tell us the alpha-level.

15
by Displaced Cane (not verified) :: Fri, 06/13/2008 - 1:27pm

My thoughts about the correlation/causation issue regarding continuity among offensive lines:

FO posits that continuity breeds success among O-lines because O-lines that stay together play well together. People suspicious of this theory point out that O-lines that play well together are more likely to stay together (reverse causation) because coaches are less likely to tinker with a successful lineup and because the injury bug didn't bite.

Here's one way to settle the dispute: Discount the teams with good O-lines. If FO's theory is correct, then teams with sub-par O-lines who keep their lineup intact should see improvement the following year. Does this data exist? Is the sample size large enough for an analysis with sufficient reliability and validity? If so, I'd love to see it. And I suspect most FO readers would as well.

16
by gerry (not verified) :: Fri, 06/13/2008 - 3:25pm

#3/#14. Amen. I'm a fan of FO as most of their work is innovative and insightful (eg. the FO Basics page). But this article is fraught with poor logic. They're arguing both sides of the same coin - reversion to the mean on top two charts, but not reversion to the mean on the bottom chart? Huh? And doesn't this contradict an artcile in PFP 2007 or PFP 2006 which argued losing a top defensive player was more important than gaining one?

17
by NewsToTom (not verified) :: Fri, 06/13/2008 - 3:54pm

Re #15
Excellent question. Take a team with a below-average ALY, find one that returns 4 or 5 starters, and see how they do. I'll see if I can't take a crack at it this weekend. Query: something you see a little bit, particularly from mediocre lines, is position switching. If you have the 4 of the same 5 guys, but one of them switched positions, say from RT to LT, is that 4 returning starters or just 3?

18
by Tom D (not verified) :: Fri, 06/13/2008 - 4:22pm

You also have to make sure the o-line stays healthy.

19
by Goo (not verified) :: Sat, 06/14/2008 - 11:47pm

Seems like selection bias is playing a role here. In other words, the sample of linemen leaving is somewhat tainted by the fact that some teams knew they couldn't afford to let their best lineman go, so they did what it took to bring him back, and he never made the list. Meanwhile, some teams probably knew they could replace their top lineman relatively well with a younger/cheaper option. So those the types of linemen who make it onto the list -- ones for whom management had at least some idea of how to replace. That's why it looks like losing a lineman makes no difference, because the linemen who didn't make a difference were let go, and those who would have made a huge difference never had a chance to make the list.

20
by Roscoe (not verified) :: Sun, 06/15/2008 - 12:19pm

When looking at team performance after losing a top offensive lineman, I am surprised Luke Petigout didn't come up. He was the Giants' best offensive lineman, who left last year for Tampa Bay. The Giants management drove its fans crazy by not picking up a replacement left tackle in the draft, choosing instead to try to convert guard David Diehl to the position.

The rest, as they say, is history.

21
by Yakuza Rich (not verified) :: Sun, 06/15/2008 - 1:17pm

The 2005 Cowboys were an exception to the rule because they were absolutely brutal to watch once Flozell went out for the season. That being said, losing Jacob Rogers at right tackle before the season started probably hurt just as much.

22
by Bobman (not verified) :: Mon, 06/16/2008 - 11:44am

#10, You're treading on thin ice there for Colts fans, who saw a starting OT go out in the SD game last year to be replaced by a guy who last played 3 quarters in the SB for them.

You'd THINK he was competent, but you'd be wrong. One might PRAY that he'd be competent, but one would then learn that God doesn't always answer prayers within the confines of a 60-minute football game (or even the 6 games remaining the season). In the end, you just buy a Charlie Johnson voodoo doll and a gross of rusty hat pins. Just for sadistic fun.

Last season for Indy, more than anything else for me, underlines ahat a few posters above have said: If you plan for it (Tarik Glenn's demise, Tony Ugoh's draft selection) it works well, but if it's a sudden replacement even to a guy who knows the system, the dropoff from "the annointed one" to "his replacement" can be huge even if that replacement aquitted himself well in the SB.

23
by Herm? (not verified) :: Mon, 06/16/2008 - 1:14pm

I subscribe to the continuity idea, but it conjures the chicken egg question.
You're not going to have a good line if you have bad linemen. You will have to break continuity by replacing bad linemen. If you replace them with good players, these players should be long term players, creating continuity (with exceptions).

Common sense would tell us that if you play a Eugene Chung clone line together for 10 years straight, they're never going to live up to this metric of success.

My belief (complely lacking the ability to measure it or data to back it up)would be simply that having good players will put you in the top 20% of the league at that position, while continuity and cohesion, coupled with individual experience and improvement will bump you into the top 10%. I'd call continuity one of the facets of maximizing a line's potential, and a strong indicator of success, but not the sole reason for it.

24
by TonyAngelo (not verified) :: Tue, 06/17/2008 - 10:32pm

RE: 4 & 13

That also describes the Colts.

1998 #1 pick Payton Manning.

1999 #4 pick Edgerrin James.

Seems to have worked out okay for them.