Tuesday, November 11, 2008

I Stand Corrected

Next time you see me talking about numbers, just tell me to shut up.

Yesterday, I let loose some easy CW about the Matt Holliday trade, noting that over the course of his career, he has been helped greatly by Coors Field, and that the A's are probably making a mistake picking him up.

In the comments, reader J. Kuhner noted that I was being too hard on Holliday, in that his road OPS over the past three seasons compared pretty nicely with some good left fielder's overall OPS. That's fair, but even so, trading for what amounts to be the sixth or seventh best leftfielder in the game and paying him that salary isn't exactly a masterstroke. Though I didn't write about it and ultimately didn't change my mind, I took J. Kuhner's point on that, mentally eased up on my criticism of the A's a tad.

Today, however, reader scatterbrain points me to Dave Cameron's analysis at FanGraphs, which pretty much puts me in my place:

His detractors will instantly point to his enormous career splits between his performance at home and on the road - .357/.423/.645 in Coors, .280/.348/.455 everywhere else. He has almost twice as many home runs in Colorado as he does away from the thin air, and given what we know about park effects and the offensive environment a mile high, we shouldn’t be surprised that Holliday has benefited significantly from his home environment.

However, when you see people pointing to his road numbers as a proxy for his true talent level, you should immediately reject the rest of their conclusions, because despite the ease of that kind of analysis, it simply isn’t accurate. You cannot just throw out Holliday’s performance in Colorado and pretend that it didn’t happen simply because the park is hitter friendly. Instead, the correct way to project his future performance is to adjust his past results to account for the park effects, and use the entire sample of data that we have . . .

[mathy stuff deleted]

. . . Even with the move out of Coors Field and into pitcher friendly Oakland Coliseum, Holiday should be expected to be something like a .300/.380/.500 hitter. Considering he’s been both durable and a quality defensive outfielder, that makes him something like a +4 win player for 2009.
Damn fine point you have there, David. Fine enough to where I hereby retract my lazy park-effects-driven criticism of the deal.

Does it still make a good deal for the A's? Maybe, maybe not. But if it is to be criticized, park effects don't seem like a valid point of attack.

14 comments:

Jason @ IIATMS said...

FWIW, says DePodesta regarding the analysis of park effects:

"I think it might be a little overdone. Park effects are definitely important when evaluating an individual player's performance, but on any given night our collective mission is to beat the other team and we're both playing in the same park. In short, I think there's a fine line. You can't ignore the nuances of your own park since you'll play there 81 times a year, but you can't be myopic either."

I would still be concerned about a decline in Holliday regardless.

tadthebad said...

Take heart, Craig. Keith Law does not endorse Holliday's defensive value as "quality", saying Holliday is an adventure in left field. However, Law does not characterize the deal as foolhardy, either.

Nice clandestine plug, Jason...I seem to remember reading that very same DePo quote not too long ago.

Jason @ IIATMS said...

Tad,

I shared that with CC before posting (link-free, mind you and your elephant-like memory).

Craig Calcaterra said...

For the record:

Jason (and anyone else with a blog or vested Internet interest) is allowed to pimp thier own work around here as long as it's arguably on-topic.

The web is a big freakin' conversation. Let's talk.

tadthebad said...

Dude, I know from wence it came! Just ribbing you a bit, no worries. I think most of us understood Shyster's tacit approval of on-topic pimping. Do I need to start using that cute little winking smiley face that CC favors when making little jokes?

And Jason, could you perhaps talk to my wife about my memory? She does not share your keen, technically justified opinion about my, um...my, uh...what was I saying? ;)

Justin Zeth said...

Like I said on BTF yesterday and then posted on my own blog, the big loser in this trade is: Matt Holliday. Because of what the effects of changing parks, going to the tougher league and going to a probably inferior lineup are likely to do to his basic stats, it's likely Holliday's going to lose tens of millions of dollars off his free agent contract a year from now.

tadthebad said...

Justin may be right...and I'm fighting off the tears just thinking about poor Matt...such a sweet kid.

tadthebad said...

;)

Anonymous said...

APBA Guy-

Of course, all of you miss the central point: how does this trade contribute to the Legend of Billy Beane?

Carlos Gonzalez, Huston Street, and Greg Smith for Matt Holliday for a year. Basically the A's trade out about $ 4M in salary to take on $ 9.5M, roughly a net -$ 5.5M. They'll get about an .880 OPS OF for at least half a season.

In case nobody noticed, last year no A's outfielder exceeded an OPS of .800 (lets call Cust a DH).

Smith is replaceable, the A's at one point started 4 lefty rookies last year.

Gonzalez is fast, light hitting, should improve, but was less coachable than the A's had hoped. Again, completely replaceable.

Street was effective, but less so than 4 other A's relievers over the year. Again, replaceable.

Billy has rented 1 year outfielders before (Jonny Damon, etc) If he likes them enough he will make an offer, or he'll flip them for picks.

While it's unlikely that the A's will retain Holliday for the entire year (the rest of the team is truly challenged offensively) and the pitchers are still very young, he has got to stop the slide in attendance.

This move will help do that. And he did it without distorting the payroll structure. Legendary.

Jason @ IIATMS said...

tad,

Wear the hat of idiocy proud at home. It will get you out of doing lotsa stuff if she thinks you're an idiot with no memory.

Trust me here.

And the ribbing is good. I'm not too smart, either, but a smiley thingie is well, um, you know. Not that there's anything wrong with it!

Anonymous said...

Billy Beane has done some good things with a limited salary. However, wouldn't you think a guy who is supposed to value OBP could do a little better than this, even without a fistful of payroll dollars?

Oakland As team OBP:
2004-.343
2005-.330
2006-.340
2007-.338
2008-.318 narrowly eclipsing the mighty San Diego Padres, who got on base at a .317 clip.

drunyon said...

Billy Beane does not "value" OBP. He values whatever the market has currently undervalued. As Michael Lewis specifically says in the epilogue, OBP is a lot more expensive now than it used to be. OBP is not the point of Moneyball.

Anonymous said...

Don't most baseball people, including Billy Beane, believe getting on base is a prerequisite to scoring runs? Are you saying if acquiring players with high OBPs becomes too expensive, then shrewd Billy Beane will get himself some of those cheap players with low OBPs, therefore outsmarting all those dumb GMs who are overpaying?

drunyon said...

Yes, Beane realizes that OBP is useful, but if it costs too much for his limited payroll, then it will be harder for him to line up. As you said yourself, more GMs are aware that OBP is useful, which has raised the price significantly. Getting a .400 OBPer isn't nearly as cheap as it used to be, so the A's will have lower OBP even with the same payroll.

Even though Juan Pierre isn't a good player, if Beane could get him for the minimum salary, I'm sure he would do it. It's all about what's undervalued relative to its real value-- and OBP is no longer undervalued with smart GMs like Epstein around.