Writers of Pro Football Prospectus 2008

printable version

Extra Points

Articles from around the Web



13 May 2009

Clayton Projects Seahawks To Improve

I always love John Clayton's yearly column on teams likely to improve in 2009, since he actually busts out the numbers.

Posted by: Bill Barnwell on 13 May 2009

22 comments, Last at 15 May 2009, 1:19pm by Chocolate City

Comments

1
by Harris :: Wed, 05/13/2009 - 2:56pm

Projecting the Panthers to improve was hardly a stretch since they stank in '07 largely because Delhomme got hurt. Similarly, it doesn't take a bunch of fancy book-learnin' to predict the 4-12 Seahawks will get better if only because 1) It's hard to get much worse than 4-12 and 2) they won't have to sign QBs and WRs out of the stands during pregame warm-ups. When Billy McMullen is getting quality reps, your team has no reps of any quality.

Hail Hydra!

2
by Keith (not verified) :: Wed, 05/13/2009 - 3:07pm

There are a few issues I have with this analysis.

It is a decent mathematical look, but he even ignores his own numbers.

He says that for every .2 points difference in SOS, it can mean one more loss or win, depending on if the schedule is easier or harder. He also states that if the offense scores 1 PPG more on average than they did the year before, that average can mean one more game win.

Taking that, he says the Seahawks will improve by possibly 6 PPG, because that is their average in previous seasons, so that should equal six wins. And their schedule is .41 easier, so that should be two wins. That is 8 wins, and he only projects five. Either he is using his fancy numbers as a front to his obviously subjective win projections, or he does not even trust his own numbers, and should therefore modify them.

I know that it is difficult to pinpoint projections like this, but three games off, and in other cases on this page, two or three games off, it is time to redo your formula if you do not trust it enough to differ so heavily.

3
by matthewglidden :: Wed, 05/13/2009 - 3:35pm

Reading that as a +5 game change, though, puts the Seahawks at 9 wins. That matches better with your excerpted numbers and is probably what Clayton means. (Seahawks +5, from 4-12 to 9-7.)

--

Better to have watched the game and lost than never watched at all.

6
by Keith (not verified) :: Wed, 05/13/2009 - 3:51pm

Except his idea is to add these wins to their win total from the year before, not from zero total wins.

7
by Keith (not verified) :: Wed, 05/13/2009 - 3:52pm

(For reference)

He clearly states "adding" a win, not simply "getting" a win. By his verbage, he means to build on the year before, not from the ground up. Still, it goes to show that he is just being subjective in the end.

16
by Jimmy Oz (not verified) :: Wed, 05/13/2009 - 11:11pm

Seahawks don't get two wins from beating an easy team and scoring more than 27 points. Clearly a team cannot get 2 wins from a single game in which both of the factors Clayton lists come into play.

8
by The Guy You Don't Want to Hear (not verified) :: Wed, 05/13/2009 - 4:05pm

"When Hasselbeck is on, he usually leads a team that averages 24.0 ppg. That should project to around a five-game improvement from the 2008 season's miserable 4-12 record."

I took that to mean that the increase to 24.0 ppg by itself projects to a five-game improvement. Notice that the sentences preceding that one are "Playing the NFC North and the AFC South won't be easy, but the NFC West's poor records take the Seahawks' projected strength of schedule from a .498 to a .457. The Seahawks could improve by two games." Also, since the change is 5.8 ppg, he may feel more comfortable with projecting 5 than 6, which seems reasonable. I think he's really projecting a 7-win improvement rather than 5.

And I think he's forgetting to downgrade the offense to account for Jim Mora Jr., but that's another issue . . .

9
by Keith (not verified) :: Wed, 05/13/2009 - 4:39pm

Ah. I can see that as plausible. Rounding down is probably safer. Over-reaching would be silly.

10
by Spoon :: Wed, 05/13/2009 - 5:07pm

Giving Clayton the benefit of the doubt, I don't think it's wise to add the projected wins together. Increased production and a weaker opponent can happen simultaneously. Let's say that your favorite team lost last week to the Steelers while your star quarterback sat out with an injury. Well if your QB is coming back healthy for next weeks game against the Bengals, then "increased production" and "a weaker opponent" are going to be two excellent reasons to expect a win - but it's still just one win, not two.

12
by NotSoAnonymous (not verified) :: Wed, 05/13/2009 - 6:23pm

Try rereading the article, Keith, because he's predicting 9 total wins for Seattle. Quoting the article: "That should project to around a five-game improvement from the 2008 season's miserable 4-12 record."

19
by Spenczar (not verified) :: Thu, 05/14/2009 - 1:37am

Not necessarily. The sentence right before is that Hasselbeck should bring the Seahawks' PPG up to 24. He might mean the better offense will add five wins.

Also, arguing that "a one-point increase in scoring can add a win for a team" is not the same as saying "a six-point increase in scoring can add six wins for a team."

20
by Spenczar (not verified) :: Thu, 05/14/2009 - 1:38am

Not necessarily. The sentence right before is that Hasselbeck should bring the Seahawks' PPG up to 24. He might mean the better offense will add five wins.

Also, arguing that "a one-point increase in scoring can add a win for a team" is not the same as saying "a six-point increase in scoring can add six wins for a team."

4
by TXNiner :: Wed, 05/13/2009 - 3:39pm

I can't figure out how the Texans made the surprise list by going from 8-8 to a projected 9-7. Hoo boy, a 1 game increase, I'll be shocked.

21
by iapetus (not verified) :: Thu, 05/14/2009 - 5:53am

Hey, a winning season for the Texans should be enough of a surprise for anyone.

5
by Temo :: Wed, 05/13/2009 - 3:47pm

I'm not forecasting the Browns to be a surprise team. I'm just putting them on the radar in case the numbers work out.

Well, at least he's honestly covering his ass.

11
by chappy (not verified) :: Wed, 05/13/2009 - 5:20pm

Best method for picking a surprise team. Pick 8 teams. If we assume that only teams that were losers or generally mediocre last year, this means half the league is likely to surprise!? My prediction, at least half his predicted teams are still bad. Of course, my prefered prediction method, which is nearly as accurate, is reading a daily horoscope...but then again, I'm a Detroit Lions fan.

13
by NotSoAnonymous (not verified) :: Wed, 05/13/2009 - 6:25pm

"I'm adding the Browns solely because of the numbers."
That has to be the most chicken-sh!tted comment I've ever seen. IT'S YOUR FREAKIN' SYSTEM, DUDE! If even you don't believe in it, why should anybody else?

15
by Joseph :: Wed, 05/13/2009 - 7:09pm

It's because he doesn't think the Browns will improve after getting rid of Winslow Jr., losing Stallworth (legal issues) and whatever other moves the Browns made this offseason.
Let's say the Jets were "projected" to improve by his system. Does ANYBODY think that going from "he who shall not be named" at QB to Sanchez/Clemens is going to help the Jets THIS year?
Let's face it--he should have put the LIONS on there--there's no way their W/L record will be worse. Good grief, a +5 improvement would make a lot of teams playoff/SB contenders--for the Lions, it means still being in last, just a lot closer to 3rd place than last year.

22
by Chocolate City (not verified) :: Fri, 05/15/2009 - 1:19pm

Yeah, but Nate Silver does this when PECOTA predicts some heinously awful record for an MLB team (think White Sox, year after year).

It might be chicken shit, but sometimes you just can't reconcile what the numbers say will happen with what your eyes tell you.

14
by JMM+ (not verified) :: Wed, 05/13/2009 - 6:54pm

Using '08 w/l records to predict '09 results is a waste of time. I have looked at correlation coefficients for strength of schedule for three seasons predicting the next season record. Each R2 was below 0.1. Useless.

18
by Never Surrender (not verified) :: Wed, 05/13/2009 - 11:13pm

Whoops, should have read to the bottom of the comments. Thanks for the info. I'd be interested to see that published somewhere.

17
by Never Surrender (not verified) :: Wed, 05/13/2009 - 11:12pm

Have there been many / any studies done on SOS and how much it actually predicts? I always get the feeling that it's not worth all that much at the beginning of a season, since what appears to be a cake-walk can become a nightmarish gauntlet depending on how the supposedly weak teams have improved.

Anyway, Clayton mentions that the Redskins have a particularly easy schedule. And while I'd like to believe that, I'm not quite getting my hopes up, yet.

Post new comment

The content of this field is kept private and will not be shown publicly.
  • Web page addresses and e-mail addresses turn into links automatically.
  • Allowed HTML tags: <a> <em> <strong> <cite> <code> <ul> <ol> <li> <dl> <dt> <dd>
  • Lines and paragraphs break automatically.

More information about formatting options

CAPTCHA
To skip this, please log in.