Third Door on the Left


Seen at Bill James Online:

After most of a big offseason making some good teams great and some great teams better, which team do you think will have the most success/win the most games next year?
Asked by: bill byrd
Answered: 1/28/2012
I have no idea.   I'm in the knowledge business; that would be the [baloney] department.   Third door on the left.  

Bill refuses to guess who will win a division, any division, any year.  He has refused to predict pennant races since, I don't know, would guess about 1985 at least.

From my standpoint, that's kind of unfortunate, because this is the one guy who has put the last 220 pennant races under a microscope, and kept them there for thirty years.  All he'll tell you about this year's pennant race is, he has no idea what will happen.

Most family doctors will eagerly give you advice as to how to raise your child.  The doctor has extensive knowledge of biochemistry, he has memorized 1000's of medical terms, he has been trained (probably) in experimental design and control; he has a prescription pad and he is not afraid to use it.

What any of this has to do with raising well-adjusted children, nobody ever seems to ask ...


Your local sabermetrician (let's say, me) knows a lot of things about the past.  

He knows why Franklin Gutierrez was credited with 6.2 WAR in 2009.  He knows what (homer per fly ball) rate usually comes out to.  He knows that a pitcher who had a 3.0 CTL pitched better than one who had a 1.2 CTL.  He knows that expected ERA predicts ERA better than ERA predicts itself.

What any of this has to do with predicting the future of a team with 25 moving parts, much less a league with 750 moving parts, nobody ever seems to ask ...


The Arizona Diamondbacks went 65-97 two years ago, finishing -15 games behind the #4 team in the NL West and -27.0 games behind the division winner.  The next year, 2011, they went 94-68 and finished +8 games ahead of their closest rival in the division.  A 35-game swing, relative to the division, for Arizona in one season.

After the DBacks' 97-loss season, sabermetricians predicted them to have a terrible team again.  It's a free country; you can say what you want.  But these predictions were the baloney department.  Third door on the left.  

Just because you know how to calculate xFIP does not mean that you are entitled to intone, with an air of authority, what the likely W/L record is for a team with 25 moving parts.  You don't know what a team will do.


Just three years ago, 2009, the Seattle Mariners were predicted to perform terribly.  Baloney department, third door on the left. They improved by +24 games for no particular reason.  Shrug.  Oh well whatever nevermind.  Some things just happen....

In 2000, they improved by +12 games, and in 2001, they improved by +25 more for no particular reason that anybody expected.  They went from sub-.500 to 116 (!) wins in two years, and nobody saw it coming.


Listen to your local sabermetrician when he tells you whether Casper Wells was better than, or worse than, average in 2011.  But you don't have to listen to him when he shares his opinion on what Casper Wells will do in 2012.  When he does that, he's a doctor trying to give you advice on raising children.

Me, I'll be rooting for the 2012 Mariners to make the playoffs until they demonstrate that they won't.  There is nothing irrational about it.


Dr D


Fett42's picture

Just because there's a lot of variance and error and unknowns in statistical predictions doesn't mean they aren't useful or informative or better than nothing. Otherwise sports bookies wouldn't make so much money.


As fetal heart rate is "better than nothing" at predicting baby gender.  As a full moon is "better than nothing" at predicting whether I'm going to be killed tonight.  ;- )
The problem comes when people take "better than nothing" and start depolying it with faux authority.  
"Using 2011 WAR to predict the 2012 AL West race" is better than tea leaves.  It doesn't mean I get to scoff at you if you reject my "better than nothing" paradigm and you are going to root for the A's to win the division.  Right?
Not in your case Fett, but a lot of guys, you hear them scoff and laugh at people who disagree, and when you call them on it, they say "Well my logic is better than nothing."  FAIL.
The bookies' odds are always informative, but the bookies themselves would not deploy them as dogma.  If they've got M's the M's a greater than 100:1 shot to win 90 games, I'll take some of that action ;- )

Add comment

Filtered HTML

  • Web page addresses and e-mail addresses turn into links automatically.
  • Allowed HTML tags: <a> <em> <strong> <cite> <blockquote> <code> <ul> <ol> <li> <dl> <dt> <dd><p><br>
  • Lines and paragraphs break automatically.

Plain text

  • No HTML tags allowed.
  • Web page addresses and e-mail addresses turn into links automatically.
  • Lines and paragraphs break automatically.


  • Allowed HTML tags: <a> <em> <strong> <cite> <blockquote> <code> <ul> <ol> <li> <dl> <dt> <dd>
  • Lines and paragraphs break automatically.
  • Web page addresses and e-mail addresses turn into links automatically.