Three True OutCusts Dept.
Q. SSI's basic reaction to this move ... being?
A. SSI's basic reaction to the move being this:
- 6.7 -- Jack Cust, runs created per 27 outs, 2010
- 5.7 -- Ichiro
- 3.9 -- Franklin "MVP" Gutierrez
- 3.9 -- Michael Saunders
- 3.6 -- Milton Bradley
- 3.6 -- Chone Figgins
- 2.9 -- Casey Kotchman
- 2.9 -- Jack Wilson
- 2.8 -- Rob Johnson
- 2.8 -- Jose Lopez
- 2.0 -- Adam Moore
- 3.5 -- Mariners' team average
So: Jack Cust produces twice as many runs as your basic Mariner. As baseball percentages go, 90%, 100% is a good percentage.
Hey, Russell Branyan was all we needed as a cleanup hitter in 2009, when we won 85-90. So I'm psyched.
.
Q. By how many runs would a 2010 version of Cust have improved the 2010 Mariners?
A. About 25-30 runs, which is 2.5 to 3.0 wins.
Fangraphs doesn't give splits by team position. But, here are the M's DH's and their runs "above" replacement:
- Griffey: -8 runs below replacement
- Carp: -3 runs
- Bradley: -2 runs (of course all these guys played some games outside of DH)
- Sweeney: +5 runs
- Branyan: +8 runs
- Total of the above: About =0 runs vs. replacement
- Cust: +23 runs
The M's DH's as a whole, "buoyed" by Russell Branyan's late run, hit .194/.269/.340 in 153 AL games. A .269 OBP has got to put you below replacement, even in Safeco in 2010.
So, a 2010 Cust would have been worth what, +25 or +30 runs in his 112 games.
More than that, if you give Cust more than 112 games, and more than that, if you don't have Russell Branyan in the equation. So +25 runs is pretty blinkin' conservative.
.
Q. How good a fit is the park?
A. As we'll discuss later ... SSI sees Oakland-Alameda as being more-or-less a push for Safeco, if you are Jack Cust, specifically.
One thing is for sure: Safeco can't do a thing about walks, so a base on balls skill here is worth 1.1, maybe 1.2 times as much here as it is anyplace else. Any player who draws lots of walks? That guy is worth more to a Safeco team, just like a RH pull hitter is worth more to a Boston team.
I don't believe that Safeco can devalue a TTO player, period. We had a test case on that, Russell Branyan, and the TTO bad boy worked out beautifully in Safeco.
Cust has played 29 games at Safeco, one month's worth of games, which ...
.
Q. Yeah! Are 29 games too small a sample? How can you tell what is a big enough sample?
A. ... one month's worth of games, wouldn't be a "sample" even if we took splits like that as "completely representative of Cust's entire universe of results," which we don't. And can't.
Very few baseball so-called "samples" are representing that player --- > even in terms of that player's true ability the following season, much less in terms of his entire career's (universe's) total results.
A "sample" from 2009 (or from the road, or from the NL, or without that hammy pull, or vs. an 80/20 platoon split, or etc) would be perfectly representing a player's true ability in all circumstances, kiddies, if you wanted to call it a "sample." Otherwise, call the FAA, this plane's grounded.
A "sample" is a handful of ball bearings that perfectly represent the other ball bearings in the can. You don't have that in baseball, guys.
Jim Thome's July 2008 at-bats are a handful of ball bearings, but they are not like the other ball bearings in the can. His July 2008 at-bats were different, for a dozen different reasons.
.................
Anyway. In that month's worth of games, Cust has hit .240/.360/.425, a line that could easily be a Cust month under any other splits or conditions.
.................
James rule-of-thumb, this one ca. 2010 not 1985: for almost all baseball splits, you need many more than 162 games' worth to --- > consider them important evidence.
.
.