Bill James talked about his own "Favorite Toy" in the 1990's … a little system for predicting an active player's % chance at 3,000 hits, 250 wins, 400 HR, etc.  He amused himself by using it to estimate, say, Ichiro's chances at getting 3,000 MLB hits.

Dr. D has a "Favorite Toy" — noodling around with b-ref.com's OPS+ and ERA+ stats. 

. 

. 

View article …

Every week or so we amble by USSM to see if they've got some inside info that the other sites don't have … Seems every time we go over, they're grousing about Carl Everett.  No worries there – D-O-V has its rubber popup punching clowns too.  Heh.

But let's not frame the argument in irrational terms, if we're going to FKey it and punch it up every other day …

A while ago, I was [grousing] about the M’s roster makeup and how their bench choices as discussed then didn’t really help them solve their problems. Look! Right here! You can have him be Ibanez’s defensive caddy and general all-around 4th OFer! He hits for some power! He’s free!

Ryan Church’s weighted mean PECOTA forecast: .272/.337/.468 (!)
Carl Everett, just to bring this up again, weighted mean .268/.334/.449
(Hee Choi: .270/.371/.511)

View article …

Chuck Knox used to laugh about pre-season predictions … "Nobody ever goes back and looks at those after the season's over."

Implying, of course, how worthless our pre-season rants are :-)  The bell starts and everybody's yap closes real quick-like, especially mine!

MarinerGeek's combo Chat Board / Pinball Arcade / Family Fun Center has a Beltre HR Call Game … go register, get your name down, get on the record … C'mon if y'don't even have the guts to get in a little prediction game, what are you doing second-guessing Mike Hargrove?!

Weeks ago, D-O-V went on the record with the firstest and the mostest on Beltre's tee-off.  Now, we goin' with 40 bombs.  Yep, that would be a very high prediction for ANYBODY in the league — it would even be stretchin' it to actually *predict* 40 for Manny Ramirez.

But Adrian Beltre is the closest thing to a minotaur I have seen on two legs.  Entering his prime, I'm thinking we're going to see some serious carnage wandering the mazes of the AL pennant race.  This guy has shown .700-type slugging percentages for half-years at a time.  Wouldn't shock me if he became (returned to?) something amazing, starting now.

View article …

Blissed axed earlier … hey, the 2006 rotation doesn't look so far off the 2001 rotation.  What would it take to repro the '01 starting five?

.

=== Close enough for Gov't Work, Dept. ===

Bill James invented Win Shares so that he would have a single number that represented a player's contribution to a team.  Why did he want that?  So that he could flip through hundreds of team seasons at a time, and see patterns in them.

For example:  how does a team have a "Cinderella season"?  Go grab 10 Cinderella teams, such as the 1969 Red Sox, and look for what happened between 1968 and 1969.  Maybe your 30,000-foot stat will make certain patterns easy to find.  (In this case, it was:  Cinderella teams tended to have three good players move up to "big star" contributions in the '69 year, and tended to have two or three terrible black holes move up to about average.)

Anyway, I like baseball-reference.com's ERA+ and OPS+ even better than Win Shares for this kind of pattern searching.

We know, we know, there are imprecisions in that.  …Chill, babe — we're not trying to splice DNA; we're trying for basic over-arc'ing patterns.

View article …