Writers of Pro Football Prospectus 2008

29 Aug 2007

FO Mailbag: Shaun (Alexander) of the Dead

Tom Holley: In my KUBIAK projection - Shaun Alexander is the 95th player overall. Is this right? ESPN has him seventh and FOX has him second. I was just wondering if there was a problem. If not, why the low ranking? Thanks.

No, there's no problem. That's his ranking. He's probably the player where there is the biggest disagreement between KUBIAK and conventional wisdom -- other guys may have a larger difference between KUBIAK rank and average draft position, but we're talking here about a guy going in the top few picks.

I don't think people realize how rare it is for a running back to bounce back from the kind of decline Alexander had last year. In fact, I don't think people realize how rare it is for a running back to decline as much as Alexander did last year, period.

Here's the list of running backs since 1978, 29 or older, who declined by more than 1.5 yards per carry with a minimum 200 carries each season:

Shaun Alexander (2006)
Barry Sanders (1998)

That's it. Two guys. Of course, we have no idea what this drop meant for Sanders, because he retired after that season. Let's loosen the restrictions a bit. Here's the list of running backs since 1978, 28 or older, who declined by more than 1.25 yards per carry with a minimum 150 carries each season:

Alexander and Sanders.
James Brooks (1991): Brooks was done at that point, but he was also 33, so he's not really a good comparison for Alexander. He gained 44 yards on 18 carries in 1992 and then retired.
Randy McMillan (1986): McMillan averaged 4.42 yards per carry for the Colts in 1984-1985, then dropped to 3.22 yards per carry at the age of 28. He never played again.

We need a comparison list with more than three guys on it, so let's open it up some more. Here's the list of running backs since 1978, 28 or older, who declined by more than 1.0 yards per carry with a minimum 150 carries each season:

Alexander, Sanders, Brooks, and McMillan.
Mike Anderson (2001): Had a total of 154 carries the next three seasons, then had a reasonable 2005 season at the age of 32.
Jerome Bettis (2002): Bettis is a special player, the only guy in history to decline like this twice and come back from it twice. He was terrible in 2002 and 2003, then useful again in 2004 and 2005.
Marion Butts (1994): 71 carries and just 185 yards for the 1995 Oilers, then retired.
Corey Dillon (2005): Rebounded slightly in 2006, to 4.08 yards per carry, and was a useful touchdown guy, then retired.
Warrick Dunn (2006): Breaking down before our eyes.
Curtis Martin (2005): Never played again.

What about the touchdowns? Alexander dropped by 20 touchdowns last year, and obviously, that's not something that happens often, because players don't often have 20 rushing touchdowns to begin with. We'll loosen up our restrictions on this question too. Here's a list of running backs since 1978, 28 or older, who dropped by eight or more touchdowns compared to their average from the previous two seasons:

Alexander
Terry Allen (1997)
Priest Holmes (2004)
Adrian Murrell (1999)
John Riggins (1985)
Emmitt Smith (1997)
Wendell Tyler (1983)
Herschel Walker (1993)
Curt Warner (1989)
James Wilder (1986)

Not counting Alexander, these running backs averaged 14.0 rushing touchdowns the year before, 4.4 touchdowns in the year listed, and 3.8 touchdowns the year after. Only Emmitt Smith had more than seven touchdowns the next year, and only Smith and Tyler had more than 700 rushing yards. Most of these guys never played another full season. This isn't looking so good for Alexander.

Let's hit one more list while we're at it. We've mentioned this a few times, but a drop in receptions is often a good indicator that a running back only has one or two good years left. Over the past five years, Alexander's reception totals go: 59, 42, 23, 15, 12.

Here's a list of running backs, 27 or older, who had a season with fewer than 25 receptions two years after a season with more than 40 receptions, and a minimum of 150 carries each year (we're looking for halfbacks here, not pass-catching fullbacks in decline):

Alexander (2004 and 2005)
Roger Craig (1991): This was the season he spent with the Raiders; he was pretty much cooked at this point but he went on a couple more years.
Corey Dillon (2004)
Pete Johnson (1983): We talk about him in the book; the next year, the Bengals sent him to the Chargers for James Brooks, and he gained 2.36 yards per carry in 87 carries, and his career ended.
Kevin Mack (1992): 10 carries in 1993, then his career ended.
Curtis Martin (2005)
Barry Sanders (1996)
Duce Staley (2004): 38 carries in 2005, then his career ended.
Fred Taylor (2005): Well, this isn't a good indicator for 2007 and 2008, is it?
Anthony Toney (1989): 132 carries in 1990, then his career ended.
Curt Warner (1988): 3.25 yards per carry in 1989, 2.84 yards per carry in 1990, then his career ended.
Lorenzo White (1994): 62 carries in 1995, then his career ended.
Ricky Williams (2005): Got high, then his career ended.

Again, not a good group to be in.

Alexander was a great player through 2005, and I'm sure he has worked hard to get back to that level, but the odds are stacked against him. We're probably underestimating how many touchdowns he'll score, but you can double his touchdown total and he still isn't a first-round fantasy pick. (Seriously; try it if you have the KUBIAK sheet.) The point of KUBIAK is to get a general idea of who is overrated and underrated by conventional wisdom, not to quibble with small statistical differences. Stay away from Shaun Alexander.

Posted by: Aaron Schatz on 29 Aug 2007

36 comments, Last at 30 Aug 2007, 2:21pm by Ashley Tate

Comments

1
by B (not verified) :: Wed, 08/29/2007 - 12:22pm

The thing about Alexander is it doesn't matter if he's ranked 25th or 95th. If his perceived value is a top 10 back, you don't want to draft him at all. Although, in an auction league, he could have some value if you use him to start a bidding war. (This will work out best if you have at least one Seahawks fan in your league).

2
by Richie (not verified) :: Wed, 08/29/2007 - 12:37pm

I didn't remember Butts ever playing for the Oilers. Doug Drinen did a fun blog post awhile back about players who played a very small portion of their careers for a particular team.

3
by Richie (not verified) :: Wed, 08/29/2007 - 12:42pm

B, are you in my league? lol

Last year we got a Seahawks fan to pay $107 for Alexander out of a $250 budget.

4
by Ben Riley :: Wed, 08/29/2007 - 1:21pm

For what it's worth, Shaun Alexander's preseason performance suggests his KUBIAK ranking is overly optimistic.

5
by shmup-o (not verified) :: Wed, 08/29/2007 - 1:45pm

Are you seriously using 13 attempts in 3 meaningless pre-season games as a justification of Kubiak? Can you say small sample size theatre?

6
by lionsbob (not verified) :: Wed, 08/29/2007 - 1:56pm

Could it just be that Seattle had bad luck last season hurting Shaun Alexander. I mean they had what 16 different OL line-ups last season (it felt like that) and Maurice Morris who had a similar season to Shaun Alexander in YPC and receptions.

I don't think we will see the Shaun Alexander of 2005, but I don't think he is going to fall off the face of the world either...

7
by Rich Conley (not verified) :: Wed, 08/29/2007 - 2:03pm

I also like the "decline in touchdowns" being used for a player who set the NFL record, and then got hurt and missed significant time the next year.

8
by Ben Riley :: Wed, 08/29/2007 - 2:26pm

#5

KUBIAK predictions are based on a variety of (objective) variables. I haven't (subjectively) seen anything from Shaun Alexander in the preseason that would contradict what KUBIAK is predicting: he's looked slow, he's dropped multiple passes, and he's whiffed on numerous blocks. You are right that the preseason is a small sample size -- but it's also all we have to evaluate right now.

#6

Seattle actually had nine different combinations on the line last year (in contrast to all of two in 2005). Continuity this year will certainly help Alexander and Maurice Morris and Leonard Weaver -- but, in my opinion, not nearly enough to make Shaun Alexander a top 10 fantasy back.

9
by admin :: Wed, 08/29/2007 - 3:13pm

For a running back in his late twenties, "missed time last season" is an indicator, not an excuse.

10
by Rich Conley (not verified) :: Wed, 08/29/2007 - 4:19pm

"#

For a running back in his late twenties, “missed time last season� is an indicator, not an excuse. "

If thats what you mean Aaron, then say that, not something silly about "decline in touchdowns"

For a guy who set the NFL record, a simple average performance would be seen as a huge decline, when its really just regression to the mean. If LT goes out and scores 18 this year, does he also have nothing left?

(I agree that alexander might be done)

11
by zlionsfan (not verified) :: Wed, 08/29/2007 - 4:29pm

Well, if he loses a yard per carry or so and sees a significant drop-off in receptions, then yes, I'd be reluctant to draft him in following seasons.

12
by Unshakable Optimist (not verified) :: Wed, 08/29/2007 - 4:38pm

For a running back in his late twenties, “missed time last season� is an indicator, not an excuse.

But the injury was a foot fracture. You guys have said yourselves that broken bones aren't chronic, so shouldn't that particular injury be considered more of a fluke rather than a sign that his body is breaking down?

13
by shmup-o (not verified) :: Wed, 08/29/2007 - 4:46pm

Yes, he did get hurt in the first game last year, played hurt the next two games, and then came back to play in 7 regular season games and 2 post-season games last year.

In the 9 games he played last year, after missing half the season, he put up Rudi Johnson type numbers (if one were to take those stats and take them across 16 games).

I don't see why you shouldn't expect the numbers you saw in 2001-2004. 2005 was a fluke year anyway, but the numbers being put up last year correlate with the numbers he put up previously with a slight decline due to AVG per rush (which was more likely due to line play and injury).

Will he return to 2005 numbers? No - but does anyone really expect that?

14
by Unshakable Optimist (not verified) :: Wed, 08/29/2007 - 4:48pm

I know you guys aren't trying to say that the foot fracture will come back, I just don't think it's the type of injury that's necessarily a harbinger of decline.

15
by zip (not verified) :: Wed, 08/29/2007 - 4:48pm

You guys have said yourselves that broken bones aren’t chronic,

Perhaps they aren't chronic, but it seems like they could certainly be more likely as a back gets older. He's not more likely to break his foot again, but he's more likely to break something.

16
by shmup-o (not verified) :: Wed, 08/29/2007 - 4:53pm

RE #13

By 9 games played last season, I meant after the return from injury.

And what's up with the Anti-spam word being "doofus" for me? ;)

17
by Pat (not verified) :: Wed, 08/29/2007 - 5:09pm

#14, #15: Break, tear, pull, etc. Things happen. More importantly, however, age makes you slow to heal, which means minor injuries that might only cost you a week or two cost you months or more.

18
by Will Allen (not verified) :: Wed, 08/29/2007 - 5:09pm

The article reminds me of something regarding running backs in general I've often had thoughts about; does trading for an established star running back, for an established star at another position, make a lot of sense usually? How likely is it that even a star running back who is three years younger than a star cornerback, is going to have more years of well above average productivity for his respective position, compared to the star cornerback?

19
by NewsToTom (not verified) :: Wed, 08/29/2007 - 5:14pm

Re #16
I thought we had a moratorium on pointing out "doofus" was the anti-spam word. That got old almost as fast as "First!"

Aside from perhaps Rod Smith, and excluding one-year flukes (hello, Michael Clayton) and injured players, who was the last non-RB to suddenly change from "one of the top couple players at his position" to "mediocre NFL starter, at best" in a single offseason?

20
by Unshakable Optimist (not verified) :: Wed, 08/29/2007 - 5:20pm

re 18:

I think it depends on where your team is at. A young team would want to keep their star, but a team that's knocking on the door might want to take that star running back and mortgage their future for a chance at the title(the assumption obviously being that a running back is the missing piece to their puzzle).

21
by B (not verified) :: Wed, 08/29/2007 - 5:33pm

19: Daunte Culpepper? Or am I getting the threads mixed up again.

22
by Andrew (not verified) :: Wed, 08/29/2007 - 5:40pm

,p>who was the last non-RB to suddenly change from “one of the top couple players at his position� to “mediocre NFL starter, at best� in a single offseason

Kurt Warner, 2002?

Trent Green, 2006?

Chad Lewis, 2005?

Kennan McCardell, 2006?

Joe Horn, 2005?

23
by Alex (not verified) :: Wed, 08/29/2007 - 5:52pm

If thats what you mean Aaron, then say that, not something silly about “decline in touchdowns�

He did say that. Health is a skill, one that Alexander is losing. That was one reason he had a decline in touchdowns. A decline in TDs as drastic as his, for any reason whatsoever, is a very bad sign for his future.

For a guy who set the NFL record, a simple average performance would be seen as a huge decline, when its really just regression to the mean.

But he didn't deliver a simple average performance. He only had 7 TDs. He didn't regress to the mean, he regressed to replacement level. That's not normal for a star RB, and as this article showed, there are very few examples of players that declined by even half as many TDs as Alexander did and then returned to a high level of play.

If LT goes out and scores 18 this year, does he also have nothing left?

No, but if he goes out and only scores 7 TDs this year, he probably doesn't have much left.

24
by Phill Skelton (not verified) :: Wed, 08/29/2007 - 6:36pm

I know you guys aren’t trying to say that the foot fracture will come back, I just don’t think it’s the type of injury that’s necessarily a harbinger of decline

The issue is that while that might seem like a reasonable position (skipping over the bit where broken bones never heal back quite as strong, particularly as you get older - so a foot fracture for a running back can certainly cause a related chronic problem), you need to find evidence to support it. Life is full of reasonable-seeming ideas that turned out to be wrong. The only actual evidence is that missing time for an older running back is a predictor for trouble in the future. Paying no attention to the details of the reason for missing time. Maybe some reasons are less of a concern than others, but then you need to accumulate the data broken down by acute vs chronic problems (say breaks vs torn tendons / ligaments) and seeing if there is an actual difference in the future stats of the players.

25
by James G (not verified) :: Wed, 08/29/2007 - 6:44pm

If LT only scores 18 this year, I don't think it would mean he was toast, but I do believe that it would mean he might begin to see a decline a la Emmitt Smith in 1996 (who dropped from 25 to 15 that year and then to 4 the next year before putting up 15, 13, and 9, with people not drafting him at the top of the 1st any more) or Marshall Faulk in 2001 (who dropped from 26 to 21 and then down to 10 and 11) or Barry Sanders in 1992 (dropping from 17 to 10, then to see 3, 8, 11, and 12 as his next 4 totals).

Arbitrarily picking out some great backs, it does look like Y7 is a pretty good time for at least somewhat of a decline. Alexander hit Y7 last year. Tomlinson hits Y7 this year. I predict that Tomlinson will not be the consensus #1 pick next year, and that the original #1 back this year according to FO, Frank Gore, will be many people's #1 next year.

26
by lionsbob (not verified) :: Wed, 08/29/2007 - 6:49pm

yes because if there is any RB I want to trust it would Bionic Legs Frank Gore.

I will be taking Steven Jackson instead thank you.

27
by joe (not verified) :: Wed, 08/29/2007 - 11:11pm

You'd almost do better to list 25th-to-75th %ile ranges, it's too easy for (statistically challenged) people to take that chart as predicting that Alexander "will score" 138 fantasy points, and run in here saying "WHEN WILL YOU IDIOTS ADMIT YOU WERE WRONG" if/when he scores point #139.

Showing ranges would get across the point "we're not saying FWP can't possibly be a top 10 pick, we just don't think it's the most probable outcome"....

28
by hooper (not verified) :: Wed, 08/29/2007 - 11:39pm

Re: 18

How likely is it that even a star running back who is three years younger than a star cornerback, is going to have more years of well above average productivity for his respective position, compared to the star cornerback?

And the genius of the Portis-Bailey trade lives on. Hold on for a second while this Broncos fan enjoys the moment...

...

...aaahhhh...

Thanks :)

29
by LOB (not verified) :: Wed, 08/29/2007 - 11:40pm

Alexander always LOOKS slow.

30
by calbuzz (not verified) :: Thu, 08/30/2007 - 2:30am

#8 Alexander's decline in passes caught is caused by his dropped passes and missed blocks. Those things didn't start this pre-season. Holmgren figured out he wasn't going to block or catch, and reduced the number of opportunities to be in on 3rd down pass plays.

If I were looking for an indicator of SA's 2007 performance during the preseason, I would look to the 1 yard TD run against the Vikings starting defense, right up the middle. That was a play they could not pull off last year, and was accomplished against a strong run defense.

31
by JKL (not verified) :: Thu, 08/30/2007 - 8:46am

I don't suspect that the foot injury to Alexander was a truly random event. Higher incidences of leg injuries are correlated with recent workload history. The "mother" of these injuries is certainly the ACL tear, but foot injuries and hamstring and groin pulls are also associated.

As for Alexander's prospects, he has several things going against him:

1) At high serious leg injury risk because of recent workload. He averaged over 26 attempts per game after returning from injury, including the post season (237 in last 9 games).

2) Risk of serious injury increased because of recent injury history and age. Both older (28 and over) and younger (under 24) backs show increased serious injury rates when subjected to higher workloads. Backs with recent significant leg injury history also struggle with staying healthy if subjected to higher workloads.

3) Risk of non-serious injury (those that cause a missed game or two at a time) is increased due to age.

4) Performance decline, as outlined by Aaron, when he did play.

32
by Israel (not verified) :: Thu, 08/30/2007 - 9:08am

Jerome Bettis (2002): Bettis is a special player, the only guy in history to decline like this twice and come back from it twice. He was terrible in 2002 and 2003,

It's not so much that he was terrible, though he had some injuries, I recall. Those were the Maddox years, when the team chose to pass instead of run. Then in 2003 Cowher replaced Jerome with Amos Thataway (or was it Thisaway?) based of what he himself called his "gut feeling."

33
by James G (not verified) :: Thu, 08/30/2007 - 9:31am

30 - Alexander's decline in receiving numbers not starting this season is a bad sign. One of the most interesting studies that FO has done, IMO, was the one where they figured the decline in catches was a couple seasons ahead of a decline in rushing for RBs. Reading their study and talking about Priest Holmes was the reason I was ahead of the curve vs. my fantasy leagues in anticipating his decline.

34
by Doug Farrar :: Thu, 08/30/2007 - 10:12am

One positive indicator for Alexander is that Cardinals cornerback Antrel Rolle has lost his starting job. It was Rolle's blatant (and uncalled) horse-collar tackle on Shaun in Week 2 that really set his 2006 on the road to decline. I'm sure that Larry Johnson's facemask also appreciates that particular example of karma. So that’s two games for Alexander without one of the more renowned “hatchet men� in the league zeroing in on him.

The negative indicators are numerous. He turns 30 today (8/30), the line in front of him will probably be better than it was last year but it’s nothing compared to that 2005 line, his decreasing reception numbers really are worrisome from a production standpoint, and there isn’t a fullback on that team right now who can consistently pick up a blitz.

35
by Rich Conley (not verified) :: Thu, 08/30/2007 - 11:00am

32.

"It’s not so much that he was terrible, though he had some injuries, I recall. Those were the Maddox years, when the team chose to pass instead of run."

No, it was that he was terrible. The team was pass-first, and he STILL couldnt put up 3.5 ypc.

36
by Ashley Tate (not verified) :: Thu, 08/30/2007 - 2:21pm

Another reason Sanders doesn't make a good comparison is that his 1.5 ypc decline from '97 to '98 was from a fantastic 6.1 average! Alexander dropped 1.5 ypc from a much lower 5.1 average.