March 11, 2008
I received a lot of good reader feedback on last week's column about Moneyball, so we'll stay with the topic for another week. Reader Mark Anderson starts us off on a factual note:
One note....Ben Fritz is no longer with the Oakland organization. The Detroit Tigers selected him in the minor league phase of the Rule 5 draft in December, and as of last check with the organization (yesterday) he's still in camp with the club.
Noted. That's a definite E-10 (Jacques, missed catch), the type of mistake I make at least twice a year because it's easy for minor league transactions to sneak by you.
Reader E.R. is somewhat more emphatic with his thoughts:
I have to say it's refreshing to read some level-headed writing on this subject. I've been annoyed with the BP writers in recent months: they seem to hate the book as much as the people who are insulted by it or still don't buy any of the concepts.
Looking back at last Tuesday's column, there's one thing I feel that I should have emphasized a bit more: I love this book. It took me a long time to write the column, largely because I'd pick Moneyball up to check a fact-what exactly did Beane say about Jeremy Bonderman again?-and the next thing I knew it would be fifty pages later, and my wife's threatening to go to the opera without me.
That's the sign of a powerful piece of literature. The aspect of it that appeals to me most-and the part, I think, that's aged best-is the biographical portrait of Billy Beane. The biggest contrast with Beane's "calculating objectivity" isn't the players he acquires, it's himself-the chair-throwing, volcanic, charming, too-tense-to-watch-the-games guy who is the antithesis of everything people think about statheads. The way that Lewis entwines Beane's personal story with the story of the Moneyball philosophy, the workings of the A's front office, and profiles of certain players is simply masterful.
That said, we're not a monolithic group at BP. I'd consider it pretty unlikely that all the authors and contributors here would share the same opinion of any book (at least, any book we didn't collaborate in writing). As it happens, Moneyball was more polarizing than the average baseball book, and that shows in the array of opinions that you'll find here about it.
The same way that some commentators act as if Beane wrote Moneyball, some people believed that what Lewis wrote in Moneyball represented the thinking of all statheads, everywhere-a point of view that reasonably rubbed many the wrong way. Although you're right that Moneyball did a great deal to promote baseball and sabermetrics, I think the issue for some in the stathead community was the same as it was for the Oakland front office-you'd rather define your own goals and opinions than have someone else define them for you. Moneyball's a great book, but it's not immune from criticism, and I can see why many people-even people who like the book-wouldn't want to be bound by every word of its text.
Finally, we can't produce a graph of BP membership before and after the book, because BP didn't have memberships until just a few weeks before Moneyball's release-if I'm not mistaken, last week marked BP Premium's fifth anniversary.
Reader J.O. spotted a flaw in my definition of the Moneyball ethos:
My take on the book is not just performance-based valuation; it was that Beane and the A's sought to exploit whatever inefficiencies existed in their market. Accordingly, if other clubs suddenly began to value college pitchers and OBP, Beane would seek out other underappreciated ways to value players. They settled on performance valuation not because it was better as much as it was under-utilized.
I think J.O.'s partly right here. I may have conflated two related concepts in the sentence where I defined the Moneyball philosophy-performance analysis on one hand and data-based market valuation on the other. The market valuations described in Moneyball were based in large part on performance analysis-you can't determine if there's a market inefficiency unless you can compare what a player did to what they cost-but that doesn't mean that the markets themselves need to be performance-based or statistical. While market valuation can tell you that walks are undervalued and stolen bases are overvalued, it can also tell you the correlation between price, performance, and non-performance data, such as radar gun readings, 40-yard dash times, personality traits, scouting observations, demographics, or even waist size. So, as J.O. indicates, the market valuation could just as easily tell you that college left-handers whose fastballs top out at 85 mph are undervalued, or that high school outfielders whose power is above 55 on the scouting 80 scale are undervalued, and to exploit those inefficiencies you'd have to use non-performance-based tools to identify those players.
However, the interesting thing about Moneyball wasn't that the A's were hunting for bargains-teams have been doing that since before the mound was moved back to 60 feet, six inches. It was the systematic, data-driven way in which they would try to identify those bargains. It's tough to "find what works" without using some metrics to measure how well things actually work.
Don Coffin had some similar thoughts:
I've always thought that the key issue was the focus on identifying skills and talents that were underpriced. So I agree with your analysis, but I would probably have pushed that part of it further up. And it raises a question-what skills/talents are underpriced now? If it was OBA and power in 2002, is it speed and fielding now? Power pitchers? Control pitchers? What, exactly, represents the 'Moneyball' approach now? That's a really important issue.
Offhand, that level of analysis isn't my bailiwick, so I'd likely punt this question to someone like Nate Silver, who has better statistical chops. However, if you go by what the A's have done in their recent drafts, you'd have to think that they've found some sort of bargain in hard-throwing high school righties. In the immediate post-Moneyball period, the A's and some like-minded clubs-particularly the Red Sox-seemed to emphasize defense, particularly in center field.
Reader M.A. thought that perhaps I overlooked the "money" in Moneyball:
Maybe I missed it, but I think you forgot to mention that except for Swisher, the A's paid less than slot money to all their 2002 first rounders. Money was a big part of the draft decisions recounted in 'Moneyball,' and it's fair to assume the A's would have made one or two picks differently if they were able to offer competitive signing bonuses to all their draft picks.
I thought that the A's paid less than slot to their first rounders initially, and my first draft of the column revolved around that fact. Unfortunately, it turned out to not be true. Here are the slots where the A's picked, and what players signed for in those slots for the two years prior and the three years after the Moneyball draft (hat tip to Baseball America for this data):
Pick 2000 2001 2002 2003 2004 2005 16 400,000 1,500,000 1,780,000 1,850,000 1,600,000 1,600,000 24 1,375,000 1,340,000 1,400,000 1,375,000 1,000,000 1,375,000 26 1,375,000 1,350,000 1,375,000 1,325,000 1,325,000 1,325,000 30 1,225,000 1,175,000 1,200,000 900,000 1,050,000 1,100,000 35 DNS 975,000 350,000 950,000 950,000 950,000 37 850,000 600,000 750,000 925,000 925,000 925,000 39 625,000 872,500 725,000 875,000 875,000 900,000 Totals 5,850,000 7,812,500 7,580,000 8,200,000 7,725,000 8,175,000
As I mentioned last week, the top four Moneyball picks were signed at or above what you'd expect in their slots. After the selection of Steve Stanley in the second round-Stanley got a $200,000 bonus where other players around that slot were getting between $550,000 and $600,000-the next pick the A's made that was really out of step with its surroundings was the signing of Shane Komine for only $17,500 in the ninth round. Aside from Stanley and Brown-who, combined, saved Oakland about $1 million over the expected bonuses in those slots-most of the A's signings were made at competitive prices.
Reader D.B. wanted to get even more specific:
Great article on the 'Moneyball' draft and I pretty much agree with all of your conclusions. The one thing I will never understand though is how they missed Curtis Granderson? Obviously players will be missed but he seems like someone who should have been at least on the radar, no? As far as I can remember he is never mentioned in the book. From doing a bit of research it seems like he intended to go back to school for his senior season but a 469K bonus convinced him to leave. I would love to ask Beane or DePo about Granderson.
Reader Matthew Knight had some similar thoughts, so I'll address both his and D.B.'s questions in one fell swoop:
I am an A's fan, so the unfair criticism of the Moneyball draft is near and dear to my heart. I did a quick study of the 2002 draft by simply looking up the WARP3 of all the players drafted in the first two rounds (72 picks). To date, the most productive player in the draft is... Nick Swisher (#16 by the A's). From picks 17-72, the two most productive players in terms of WARP3 so far are Mark Teahen (#39 by the A's), followed by Joe Blanton (#24 by the A's). Overall, the A's have three of the five most productive players in the draft-Khalil Greene and Scott Kazmir rank ahead of Teahen, but both were drafted before the A's first pick, so you can't fault the A's for not drafting them). And sure, Jeremy Brown (#35) only played four games in the majors, but 12 guys picked ahead of him still haven't made the majors, and three more have but have negative WARP3s.
I focused on the first round of the 2002 draft last time out, so I'll present a quick list of some of the prominent names the A's missed out on in the second round and beyond. As a Toolbox-y nitpick, I'll note that I'm using WARP1 rather than WARP3 for the career totals-the era adjustments contained in WARP3 don't add much to the conversation since we're looking at an entirely contemporary cadre of players. Also, I've thrown the 2008 PECOTA projections into the mix as a proxy for potential, since some longer-developing prospects shouldn't be judged on just their major league numbers:
Career PECOTA Hitter Pick PA WARP WARP Curtis Granderson 80 1557 19.5 6.8 Russ Martin 511 1088 13.7 5.8 Brian McCann 64 1248 13.1 5.2 Joey Votto 44 89 0.8 4.8 Howie Kendrick 294 636 4.3 4.3 Elijah Dukes 74 220 0.5 3.5 Chris Snyder 68 1076 9.2 3.1 Career PECOTA Pitcher Pick IP WARP WARP Matt Capps 193 163.2 9.5 4.9 Rich Hill 112 318.0 7.1 4.0 David Bush 55 630.1 15.2 3.7 John Maine 166 324.2 7.7 3.1 Jonathan Broxton 60 172.0 7.4 2.9 Jon Lester 57 144.1 3.5 2.5 Josh Johnson 113 185.0 5.5 2.5 Pat Neshek 182 107.1 6.3 2.2 Scott Olsen 173 377.2 5.3 2.1 Joel Zumaya 320 117.0 6.0 2.1 Jesse Crain 61 199.2 9.5 1.5
And here are the totals for the players from the A's 2002 draft class who made it to the majors:
Player Pick IP/PA Career PECOTA Nick Swisher 16 1924 15.1 6.0 Joe Blanton 24 633.2 16.2 3.9 Mark Teahen 39 1538 12.2 4.6 Jeremy Brown 35 11 0.1 1.8 Bill Murphy 98 6.2 0.0 1.4 Jared Burton 248 43 2.4 1.2 Shane Komine 278 20.2 0.2 0.6
Now, this is an unfair comparison, because after the first round, the field's picks outnumbered the A's by a factor of 29 to one. Granderson stands out as the most successful-both historically and prospectively-of the group. Four years after the draft, he and Joel Zumaya-whom the Tigers snapped up in the 11th round of the draft-would help Detroit reach the World Series, sweeping the A's in the ALCS along their way. Performance-wise, Russ Martin looks a lot like the player Oakland hoped Brown would develop into. Since the Dodgers took Martin in the 17th round, it's not like the A's didn't have the opportunity to select him. Another player who seemed like he'd be right in the A's wheelhouse was Pat Neshek, a college right-hander whose unorthodox delivery would have fit in with Moneyball misfits like Chad Bradford and David Beck; the Twins picked him up in the sixth round. Joey Votto's another one that'd fit into the Moneyball mold of ideal A's hitters, although I think Oakland can be excused for not using one of their first-round picks on a Canadian high schooler who went to the Reds early in the second. Maybe they'll be able to trade for him after Dusty Baker breaks his spirit…assuming the A's still like guys who can take and rake, that is.
Doug Pappas, The Numbers (Part Four)-Player Compensation: E.R.'s letter, which mentioned how much sabermetrics and Baseball Prospectus owes Michael Lewis and Moneyball, brought to light my greatest oversight from last week's column-writing a "Further Reading" section about Moneyball without mentioning the man who inspired Lewis to write his bestseller. This article by the late, great Doug Pappas introduced Marginal Dollars per Marginal Win metric, a way of evaluating which teams were getting the most bang for their payroll buck. The A's standing in this metric, as further detailed in Pappas's 2003 follow-up article was the spark that lit the Moneyball fire in Lewis's head. The original article was part of the eight-part Numbers series, a comprehensive look at the game's finances, which may be the finest thing ever published on this site. This link takes you to the end of the series, which contains links to Parts One through Seven. These articles, like all of Doug's work on this site, are available to non-subscribers, so no one has any excuse not to read them.