THE BOOK--Playing The Percentages In Baseball

So, four choices:

small budget + poor player analysis/development.
large budget + poor player analysis/development. small budget + good player analysis/development. large budget + good player analysis/development.

I think I have them in order from worst to best.

Buzz makes a lot of mistakes here, not going to waste my breath enumerating them.  I do like his suggestion that the A’s scoring 947 runs in 2000 because of Hudson, Mulder and Zito. 

But he is right that Beane was wrong about plate discipline being a good indicator of future success in the absence of power.  There were a lot of players in the A’s system like that.

"The explanation was dazzling, although Lewis barely mentioned the three reasons the A’s had been so successful--pitchers Barry Zito, Mark Mulder, and Tim Hudson...The odds of three young pitchers coming together like that on one team was basically a matter of baseball luck”

“His theory that only college pitchers should be drafted over high school ones because of their experience sounded plausible. But it flew in the face of the Atlanta Braves, who won their division 14 years in a row from 1991 to 2005, and relied on pitchers drafted straight out of high school all the while.”

Huh.

"The explanation was dazzling, although Lewis barely mentioned the three reasons the A’s had been so successful--pitchers Barry Zito, Mark Mulder, and Tim Hudson...The odds of three young pitchers coming together like that on one team was basically a matter of baseball luck”

Aye, it’s a good thing that all baseball teams only have 3-man rosters.

The other funny thing is that Buzz indicts the A’s for not making the WS...But how about those Braves, who lost 4 out of 5 WS?  Doesn’t that mean those teams were badly-constructed?

Wow...that article was exactly the article I was expecting when I clicked the link. A couple of times Buzz doubled back on himself there, but yeah...pretty much exactly what I suspected.

One semi-related point. Guys like Buzz probably put a lot of stock into RBI when evaluating major league players. Why then have I never heard a baseball talent evaluator talk about how many RBI a player had in high school or college? “Well Bob, we realize that our top draft pick doesn’t has little bat speed, can’t catch a cold, and hasn’t faced elite competition...but he was 5th in the nation in RBI this year.”

Bissinger still misses the point of the book.  The point of the book was that to succeed with fewer financial resources, a team must identify and exploit inefficiencies in the market for baseball talent. If it’s OBP and college talent that are undervalued, then exploit that. If it’s defense, exploit that. The point was not that you just get players to take more walks and you ignore HS pitchers.

I do think Bissinger is basically right that the rich teams have become smarter and more efficient in their use of vast financial resources, and because of that it is extremely difficult for low revenue teams to find inefficiencies to exploit.
The financial playing field is still very unbalanced, but the intellectual playing field has become a lot more level since 2001.

I have never really understood the disdain for Moneyball. It’s a simple concept: General manager sees something that others have overlooked and uses that knowledge to help his team win games. Of course the other teams notice what’s going on and respond accordingly, with the extra advantage of money.

On-base percentage only matters in the book because it’s the thing that the others overlooked. (For decades, I might add.) At another time in baseball history, the thing would have been different.

"It’s a simple concept: General manager sees something that others have overlooked and uses that knowledge to help his team win games.”

Sure, but the question is (or a question is): did Beane actually do that?  It’s not clear to me he did.  From 1999 to 2003, the SLG:OBP ratio for the As is virtually identical to the league average.  If they were succeeding by buying “undervalued” OBP, the team should look different:  high OBP relative to SLG. 

Unfortunately, views of Moneyball are so polarized that it’s hard to have a productive discussion.  People in the sabermetric community often defend it, in part, because its critics seem to be bashing all statistical analysis (and some of them are).  Personally, I’m a stathead who’s unconvinced that the Moneyball story is right. I don’t think OBP was ever that undervalued.  But I could be wrong. Has anyone done a good, dispassionate analysis comparing Beane’s actual teams to the story told in Moneyball? If not, would be a good piece for Studes or someone to commission....

I actually thought the real theme of Moneyball was not about OBP or college pitchers or any other particular observation about how to identify baseball value.  It was more generally Beane’s tendency to de-value (relatively speaking) pure, physical “five-tool” potential and look more at actual productivity in performance, or rather the probability of actual productivity in performance.

For Lewis, this makes a dramatic psychogical story because it ties back to Beane’s personal history—the five-tool guy who failed spectacularly in the big leagues.  The emphasis on things like OBP and college pitching experience are subsets of that overall emphasis on performance over appearance: walks aren’t sexy in a five-tool way but contribute to wins; high school pitchers can have glamorously flashy potential but college performance gives a better real track record, etc.  Bradford was another example: sidearmers aren’t usually toolsy power guys, but they can win.

This tools vs. performance contrast also produces the scouting issue that comes up in Moneyball: traditional scouting is portrayed as being about looking for the physical tools, while Beane wanted to add to that (not replace it, but add to it) more emphasis on mental makeup and statistically idenitifiable performance.  Beane, as portrayed in Moneyball, also saw his performance over tools emphasis as being a workable niche for his small market club on the theory that the market was overvaluing tools, so he could grab some advantage by relying on tools evalautions to a lesser degree.

Ultimately, Moneyball seems to me less about sabermetric issues, which is the way it has been caricatured (both by sabermetrics proponents and opponents), and more about Beane’s own psychodrama, in which he projects his own experience as a young man on to his judgments about how to run a baseball club, with some success.

Birtelcom, your fair reading and clear understanding of this book have no place in the modern world. None.

If you read a book, or watch a movie, you can expect, even need, literary licences, so that the writer/director (artist) produce quality and makes a point.

Introduce a baseball-theme into that same piece of work, and all of a sudden, you get “Shoeless Joe doesn’t bat righthanded!  The movie is crap!”

Imagine if the Da Vinci Code was the Cap Anson Code.  The baseball world would never recover.

Billy Beane should never have written that book.  He’s become a laughingstock!

I think “Moneyball” was actually just a story about how Oakland managed to be successful without spending a lot of money. It’s journalism, not argument. That’s my opinion, anyway. But it has textual problems that invite a reading of it as argument. All the fighting over it is basically because people insist that it’s an argument, when in fact it’s just bad reporting (e.g., only telling one side of the story, assuming causal relationships between events, etc.). I loved the book as a story, but it has serious flaws as journalism or argument, and since no one does or cares about textual criticism, people just argue about its ideas like it’s an op-ed piece in the Times.

They’ve had almost seven years, but I’ve yet to see anyone who feels compelled to “explain” Moneyball consider this question:

What if Billy Beane was not being a straight shooter if a writer asked for your secret formula (e.g. “Look for OBP.")?

That is, if you were Beane, what would you tell (directly or indirectly) Brian Cashman about constructing a baseball team?

Michael Lewis has a great eye for a catchy quasi-idea, and a demonstrated ability to pound that hook into the ground, regardless of the facts.

But it might be good to remember that he, not Beane, is the author of Moneyball.

Guy @ 9 - Moneyball was written in 2002, so forget 2003.  Billy Beane was willing to give up his secret ("OBP") in 2002.  So it wasn’t a secret anymore, nor was it particularly undervalued.  The A’s ranked higher in OBP than SLG every year from 1999-2002, so it seems like that was the skill they chased.

But remember the storyline is that Beane sought out slow but high-OBP sluggers.  One year, he started Matt Stairs, John Jaha, Ben Grieve and Jason Giambi.  Those guys aren’t there for their fielding.  Any claims that the 1999-2001/2 A’s weren’t about OBP are crazy.  And any claims that the 2003-09 A’s were about OBP ignore the fact that Billy Beane found other talents that were undervalued and moved on.

I think the problem with “Moneyball” is it attempts to sell OBP (or has that practical effect) as the “secret to success”.  Beane’s success wasn’t based upon, fundamentally, the OBP secret.  It was based on exploiting market inefficiencies.  In the late 90’s and early 00’s, teams underpaid for OBP.  Beane saw this as a market inefficiency and built a contender by exploiting it.

To quote JC Bradbury of Sabernomics (The Baseball Economist):

The diffusion of statistical knowledge across a handful of decision making units in baseball was apparently sufficient to correct the mispricing of skill. The underpayment of the ability to get on base was substantially if not completely eroded within a year of Moneyball’s publication.

PW/17: that conclusion is nothing short of b.s.  How can the market correct itself so quickly, when you have substantial long-term contracts.

More regression hocus-pocus.

I understand the point.  But what is the evidence that OBP was undervalued prior to 2003?  How do we know there was ever a market inefficiency?

Hawerchuk:  OK, let’s exclude 2003.  From 1999-2002, Oak was .350 OBP and .444 SLG, for an OBP/SLG ratio of .785.  The AL overall was .340/.433, for a ratio of .788.  The Oakland stats are exactly the proportion we expect a team to have (but obviously above average).  If OBP was undervalued in comparison to SLG—and what else could it be undervalued in comparison to?—and Beane exploited an inefficiency, then this ratio has to be different.  Otherwise, he’s getting OBP at a bargain (or so we are told), but he’s overpaying for the SLG at the same time and it’s a wash. 

I haven’t read the book in a while, but slow/fast seems like a small part of the story—it was the alleged disparity in valuing OBP vs. SLG that was the core story.

The fact that Beane didn’t care much about fielding gets less attention now, I suppose because it’s now fashionable in saber circles to highly value fielding.  But even there, is the story even true?  Oakland’s DER in these years was .691, vs. league avearge of .688.  I think they had some low BABIP pitchers, so maybe their defense was a bit below average—I don’t know—but doesn’t look like it was particularly bad. 
And even if it was bad, do we think Beane made the right call on defense?  Was fielding really overvalued ten years ago?  What’s the evidence for that?

Guy/Hawerchuk: how about looking at the SLG/OBP ratio of all players drafted or signed between 1999 and 2001ish, regardless of what years they played?  That would be the real test.

FWIW, here’s a paper from academia that seems to confirm both the existence of the OBP inefficiency and the correction of it by the time that Moneyball was published.

http://hubcap.clemson.edu/~sauerr/working/moneyball-v2.pdf

I guess the correction would be seen in the signing of free agents after the inefficiency was identified and became widely known.  It’s not clear to me that there’s enough data here to make the conclusions a lock though.

i like how buzz can resist calling jeremy brown fat.  multiple times.  what a hack.

i also like how in this short comment thread, several legitimate beefs regarding moneyball and lewis’ book have been brought up, vs the complete dearth of new information or relevant points brought up by buzz.  just a bunch of partisan (to the extent that that word applies in this case) retread arguments that makes bissinger sound bitter and lame.  and hes a professional writer who presumably got paid for that piece.  whereas a bunch of amateur yet intelligent baseball enthusiasts that read this sight can offer more insight in this 50 word comments (or however long they are) under tom’s simple link. 

buzz can still be an entertaining read, but not here.

Guy,

Beane found John Jaha (101 BB, 970 OPS in 1999) for $525,000.  Matt Stairs (153 OPS+ in 1997) made $165,000.  Dave Magadan, an OBP all-star - $255,000 in 1997.  Olmedo Saenz (401 OBP in 2000) - $290,000.

You see these guys all over the place - guys with power, but not top 10 or 20 in the league, and a huge number of walks.  Available as free agents.  They got paid next to nothing for a year or two with the A’s and moved on to higher salaries as other GMs figured out what was up.

When you look at guys like this, particularly slow ones, you’ll see the OBP discount prior to 2001.

I don’t think slg/obp ratio is the way to do it.  Many of the high obp guys were also high power guys.  obp/war would be better… showing what proportion of their total value came from obp.

OBP/SLG ratio is definitely the right way to do it in terms of evaluating the claims of Moneyball.  See pages 127-129, where it discusses OBP “really” being 3x as valuable as SLG, and how OPB is undervalued by the market.

That said, a more plausible version of the Moneyball theory is that BBs specifically were undervalued, vs. BA.  The A’s pretty consistently had a below-average team BA but above-average OBP.  If BA was overvalued by the market, in comparison to walks, that might give the A’s a competitive edge. 

Hawerchuk:  cherry-picked examples don’t prove much.  And Jaha was a well-paid player before having two low-PA seasons—I assume he was injured those two pre-Oakland years.  Magadan had 1 good year in Oakland, but lousy years both before and after—looks like Oakland got lucky with one for one season.  Stairs was a good pickup, and might be an example of a low BA causing a player to be undervalued.

The “OBP is 3x more valuable than SLG” is DePodesta running a regression of runs scored vs SLG and OBP and finding that the OBP coefficient is 3x higher.

The A’s were able to find players with high OBPs at low prices.  I was merely listing some.  The A’s vastly underpaid for having the league’s second-best offense, and a lot of that was underpaying for high-OBP players.

Tangotiger,
21 (weskelton) said it well.  It’s the contract’s signed AFTER Moneyball’s publication (and not the long-term ones signed before but extended until after) that JC deals with.  Looking at each individual market (each off-season) is a more productive exercise for his purposes.

"The “OBP is 3x more valuable than SLG” is DePodesta running a regression of runs scored vs SLG and OBP and finding that the OBP coefficient is 3x higher.”

If you run a regression comparing the change in SLG and OBP to linear weight values, OBP is roughly 1.7 times more valuable than SLG. 3 to 1 isn’t even close to the proper ratio. It’s nearly double.

Wes/21:

http://www.insidethebook.com/ee/index.php/site/comments/why_i_hate_regression_techniques/

PW/17: that conclusion is nothing short of b.s.  How can the market correct itself so quickly, when you have substantial long-term contracts.

More regression hocus-pocus.

I noticed that the quote that PW attributes to Bradbury, actually originates from p.14 of the paper I linked to in #21 above.

So I guess Tango is saying that the process and thus conclusion of the paper is flawed.  Is this because they are not looking at the individual signings on a Y2Y basis?

Terpsfan @28:

Yeah, the irony is that DePodesta made a mistake in Excel at some point, and told Lewis the wrong thing.  I can’t remember where I saw this, unfortunately.