Again With the Not Understanding Baseball

Look: If people want to challenge the premise of comparing baseball to teachers in New York, fine. But if you're going to make the argument on baseball terms, then have your facts in order. First there was this, then Leo Casey says:

It now appears that “Billyball,” as its advocates called Beane’s statistical approach, doesn’t have quite the track record of success Carey reported. The most famous account of Beane’s method was Michael Lewis 2004 book Moneyball, which looked closely at Beane’s 2002 draft picks, since the Athletics had accumulated an unusually large number of such picks that season. As New York Times sports columnist Murray Chass recently recounted, the Beane’s 2002 choices chronicled in the Lewis book have proven less than felicitous. Statastically speaking, the other teams which picked based on scouting reports did better than the Athletics. You can count us among the skeptics that evaluating teachers is a process akin to judging baseball talent. But it is interesting to know that the baseball model being proffered as a basis for judging teaching performance was not even successful on its own terms.

As people who have actually read the book know, the premise of Moneyball was that Billy Beane was able to exploit inefficiencies in the baseball labor market created by the difference between what people believed was important in a baseball player and what, statistically speaking, was actually important. Some of those biases were between different kinds of statistics--people overvalued stolen bases and undervalued on-base percentage. Other biases were between human observations and statistics--people believed they could identify talent by looking at it, and were often wrong. People also tended to gamble high draft picks on high school pitchers, even though such choices rarely pan out. Beane's approach was particularly valuable to the A's because they were, and are, a small market team with relatively low revenues. Teams like the Yankees could overvalue players and still win a lot of baseball games because they had money to waste. The A's didn't, and so it was critically important to maximize the number of wins per dollar spent. Beane's success in this respect is completely irrefutable; under his management the A's have consistently won more games than most teams while spending less money.

The Murray Chass column Leo references is about one small slice of that overall strategy, the 2002 amateur draft, which forms a lot of the narrative of Lewis' book. Chass' suggestion that the long-term result of that draft doesn't support Beane's strategy is based primarily on the following:

Four of the seven players picked by Oakland (57 percent) among the first 39 picks in that draft have played in the majors, including [Jeremy] Brown. Of the other 32 picks, 20 have played in the majors (62.5 percent).

The A's first pick in that draft wasn't until 16th. So their average draft position among the first 39 picks was lower than the average position of the other teams teams. More importantly, that draft was a success. Three of the four players--Nick Swisher, Joe Blanton, and Mark Teahen--have become successful major leaguers, which is a fine rate given the small percentage of draftees who succeed. Chass notes that only one is still with the A's as if that's a negative, which is silly--I don't think any believes the Red Sox were wrong to sign Babe Ruth just because they eventually sold him to the Yankees.

Chass also notes that Prince Fielder, one of the players chosen before the A's first pick, turned out to be quite good. Sure--but most of the other players drafted before the A's first pick, Swisher, didn't do nearly as well. In fact, the first round of the 2002 draft is littered with high school pitchers who went bust because of injury problems--precisely the kind of player Beane rightfully avoided, one of the many reasons the strategy Leo derides has in fact worked so well.

Update: A reader points out that "Billyball" is the term for the way Billy Martin managed the Yankees, not the way Billy Beane runs the A's. Another strike against Leo.

Update 2: Matt Tabor has a good post in response to Leo here.

College Transfer Blues

Laura Dempsey, a civil rights lawyer and Army wife, writes in the Washington Post about the many reasons it's hard for her to maintain a career. Among them:

Wives attending college when their service members transfer must choose between paying exorbitant out-of-state tuition if they stay behind or losing a substantial number of credits if they move. Although many smaller and online universities admirably volunteer to accept transferred credits for military wives, most of the country's larger public universities and almost none of the top-tier private schools do.

This is a good opportunity to point out that the "system" of tranferring college credit in this country is a mess, much more so than most people realize. Colleges start with the baseline presumption that credits earned at other colleges are no good. Then they intermittently create "articulation" arrangements with other institutions, often on department-by-department ore even course-by-course basis, most commonly within state university systems or between systems within states. Absent those arrangements, they just decide which credits they'll accept however they like. It's a completely non-transparent and idiosyncratic process, and the worst thing about it is that students don't find out how many of their credits will be accepted by the college they're transferring to until after they apply and decide to go.

This is partly because we live in a big country with a lot of colleges that are governed in a very decentralized way, so there's no clean public policy solution from a national perspective. But there's also an element of institutional hubris -- departments tend to think that the standards for introductory Econ or what have you at that other institution just aren't up to snuff, when in fact most of the courses students take as undergraduates aren't all that different regardless of where they're taught.

There's also a financial element -- it's not in a college's best interests to make it really easy for students to transfer out, and when they accept students who transfer in, every credit they reject is a credit students will have to pay to re-take. Market pressures don't fix the problem because students don't usually transfer many times, so it's a case where they don't know they're going to get the shaft until they've gotten it, and once they've gotten it they're never in a position to avoid getting it again.

Only students in special circumstances that result in serial transfers--like Army wives--come to realize just how absurd and inefficient the system is, likely resulting in billions of dollars per year of wasted time and money that could be better spent elsewhere.

Other Than That, Spot-On

Mike Klonsky takes issue with my recent baseball / education comparison:

But Carey equates Bloomberg’s N.Y. testing mania with 1990’s Oakland A’s manager Billy Beane’s supposed reliance on statistics (“crunching numbers without prejudice”) to pilot his team to the World Series. What a stretch! No team has better individual player stats than the current Yankee team which can’t quite make it to the top. If statistics decided everything, there would be no need to play the game. Good managers crunch numbers, but often make their most important decisions based on intangibles, gut feelings, and connoisseurship. One more thing—A’s star power hitter Jose Conseco, the first major leaguer ever to hit at least 40 homers and steal at least 40 bases in a season-- was also one of the first admitted steroid users. Steroids may have ultimately done great damage to Conseco’s body as well as to the integrity of MLB. But it sure was good for his stats. Maybe Beane should have been looking past the numbers and Carey looking for a better metaphor.

A few observations:

1) The Oakland A's have not gone to the World Series under Billy Beane, which is the main reason I didn't say that they did.

2) Billy Beane is the general manager of the A's, not the manager, which is not at all the same thing.

3) Yankees players indeed have excellent statistics, which likely one of the reasons that the Yankees have, over the last 12 years, gone to the playoffs 12 times, won the American League East 10 times, won the American League pennant 6 times, and won the World Series 4 times. They haven't won the Series in the last few year because--unlike basketball and football--baseball is a sport where even the greatest teams only win 60-some percent of their games, so the odds of winning three consecutive short series against good opponents are always against you. The Yankees didn't need Moneyball techniques to win all those games, because they have more money than Brunei.

4) Nobody named Jose Conseco has every played major league baseball. Conseco is a large insurance company based in Indianapolis, where I used to live.

5) Jose Canseco did indeed go 40-40, in 1988, nine years before Billy Beane became the general manager of the A's. He left the A's in 1992, returned briefly in '97, and never played for them again, no doubt in part because his on-base percentage for the A's that year was .325, making him exactly the kind of player Billy Beane did not want to hire.

6) There are no equivalents of steroids in teaching--no dangerous illegal substances that boost your classroom performance at the expense of your fellow teachers. So I have no idea what the steroid scandal is supposed to demonstrate here, other than when you give people strong incentives to boost their performance, they try very hard to boost their performance, which is more or less my point.

8 1/2 Million Dollars an Hour

Philanthropist Donald Bren just gave $8.5 million to THINK Together, an L.A.-based nonprofit group that runs afterschool programs, to extend the school day by an hour in 30 or so Santa Ana and East LA schools. That is one huge private donation to expand THINK's afterschool programming, which stands to really help these kids--largely low-income and English-learning--gain the extra learning they probably need. Still that's one seriously expensive hour so I hope a chunk of that goes to evaluate the program's impact . Anecdotes and testimonials are nice but they won't be enough to sustain that kind of expansion beyond Bren.

Please Stop

Acknowledging that few observations are more banal than the soul-deadening nature of airline travel, for me the single worst thing is the televisions BLARING cable news around departure gates. Yesterday -- I'm not making this up -- I sat in the Detroit airport across from a nun while CNN ran an extensive piece investigating the critical issue of why celebrities cheat on their hot wives, complete with references to Divine Brown, Jude Law's nanny, and constant use of words like "hot," "erotic," etc.At the moment it's something about whether Britney Spears is somewhat as opposed to completely insane.

Back to education tomorrow, promise

The Ivy League Just Keeps Getting Greener

But not the trendy, Al Gore kind of green—the old-fashioned, John D. Rockefeller kind of green.

I received an email this morning from Stanford University announcing that it (like Harvard, Princeton, etc.) will be expanding its financial aid program. Now a family with an income of less than $100,000 will not need to pay tuition, and a family with less than $60,000 will not pay tuition or room and board fees. It’s great that wealthy institutions like Stanford are starting to one-up each other on generous financial aid packages, but it also points to some troubling issues.

Stanford’s aid will provide free tuition to households with incomes in the top 20 percent of incomes in the United States, and the offer of free tuition and free room and board will go to households with incomes in the top 40 percent. Something seems wrong when a college’s tuition—even an elite private school—is beyond the financial capacity of those defined by the New York Times to be “upper class” and “upper middle class”. Suddenly, there is a lot of aid money going to students who, when looking at the income distribution in the United States, shouldn’t need it.

And this aid money is not distributed evenly across colleges—most institutions don’t have the financial resources to offer these generous aid packages. The Council for Aid to Education released today its “Voluntary Support of Education” report, which tracks private donations to colleges. These donations have increased for the 4^th year in a row, totaling $29.8 billion in 2007. But even more interesting is that the top 20 institutions, representing 2 percent of responding institutions, raised over a quarter of all the money going to higher education. Topping this list? Stanford University with $832 million last year.

The Brookings Institution, meanwhile, released a report today indicating that—while a college degree is still a powerful ticket to upward economic mobility—the growing gap in college attainment between the rich and poor in the United States threatens the ability of those from low-income families to climb the income ladder. And our most recent Chart You Can Trust shows that if you’re from a low-income family, it’s still much harder to get into the most elite institutions, even if you have high test scores, than someone from the “upper middle class”.

So while students that attend an elite institution can increasingly be assured that college will be affordable, this assurance does not extend to the vast majority of students who attend less wealthy or open-access institutions. And since low-income students are much less likely to attend an elite institution like Stanford, they might end up paying a higher tuition bill than someone who can attend--even if that person is from the top quintile of household incomes.

Brrrrrr......

I'm in Grinnell, Iowa, at the moment, visiting Grinnell College, giving a presentation / speech about NCLB. There's a lot I could discuss -- the friendly people, pleasant accomodations, lively dialogue before and after the speech, the fact that Grinnell has the 6th largest endowment per-student of any college or university in the nation, larger than Stanford, Amherst, or M.I.T., but all I can really concentrate on at the moment is the fact that it's very, very cold. I mean, REALLY COLD. When I accepted the invite to come here last Fall, I knew it was cold in theory in Iowa in February, but to actually step into this upper midwestern maelstrom of coldness is to be reminded that it's never, ever, actually cold in Washington, DC. Chilly perhaps, nippy from time to time, occcasionally on the cool side, but never cold like it is here, now. The drive from Des Moines takes a little over an hour and on the way I must have seen 100 cars buried in the median or on the side of the road in the ice and snow. It's cold, I say!

Update: Temperature when I walked out the door at 8AM this morning: Minus 12 degrees. Just to put that in perspective, I'd say the typical resident of DC would regard 30 degrees as cold. The difference between here and that is the same as the difference between 30 and warm spring day of 72.

How We Deal

Eduwonkette’s recent post on linkages between leading education think tanks poses a couple of important questions: Do the leading education policy shops amount to “multiple organizational outposts” for a small and insular group of thought leaders? And can think tanks claim to be independent evaluators when they share board members and funders?

Since Eduwonkette included Education Sector in her post and accompanying chart, it’s worth discussing how we deal with the issues that she raises.

To start with, we eschew government money and we don't do fee-for-service work. We are funded not by one or two foundations but by many different ones. We list them on our web site and we name them in individual reports if they fund that work specifically, though a substantial percentage of our funding is general-operating support rather than project-specific grants. Our contractual agreements with foundations give us editorial control over the work we produce. We’ve found that what our funders value most is good work—thoughtful analysis, clear writing, and an ability to advance our ideas effectively.

To further promote transparency, our web site includes biographies of all of our team members, directors, research advisory board members and non-resident senior fellows. In each instance, our goal is to draw on the expertise of smart people with differing backgrounds and perspectives. The fact that some of these people are also affiliated with other organizations means only that others recognize their talents as well.

What drives our work at Education Sector is a set of core principles about the purposes of education and the nature of educational reform. We posted them on our web site early in our history under the heading What We Believe. These principles undergird all of our research, analysis and commentary and they make it easy to tell where we’re coming from on policy questions. Agree with them or not, you know what they are.

We also have a Transparency Policy that governs potential conflicts of interest. Every Education Sector employee has sign the document anew every year. And we are developing for our web site documents that describe our “theory of action”—our sense of how to improve American education and how we as an organization propose to promote those improvements happen—in the policy areas where Education Sector does the bulk of its work.

In often makes sense in the course of doing our work to collaborate with other organizations. Such networking sharpens our thinking and helps us have impact. It’s what every good think tank and policy organization does in every field. Conversely, Education Sector does not as an institution sign policy manifestos or other multi-organization documents. Nor do we take institutional positions in policy debates, in contrast to membership and advocacy organizations, which frequently do. Each of us shares Education Sector’s organizing principles, but we often have very different takes on issues. It keeps us sharp.

Ooh, Good News

It seems that a new Neal Stephenson book will be published in September. Apparently, "It's set on another planet and has aliens and so on. It's really about Platonic mathematics, but he needed the aliens and space opera-ish elements to spice it up a little bit." Exciting! I've read objectively better books over the last five years, but I probably haven't enjoyed any books as much as Cryptonomicon and the time it took to plow through the entire 2,500 page Baroque cycle was well spent. I believe this may put me in common company with Instapundit, which is kind of disturbing, but what can I do. Stephenson's Snow Crash remains one of the more prescienct novels in recent memory, he basically described Second Life in 1991.

The Wire, Season Five, Episode 7

In which Clay Davis plays every card in the race deck, except, presumably, the King of Diamonds, because that's McNulty.

Summary: Davis beats the rap with the help of real-life Baltimore criminal defense attorney Billy Murphy, who once got Don King acquitted, so I'm guessing this wasn't much of a stretch. McNulty gets everything he ever asked for, with predictable results. Steintorf tells Carcetti he has no choice but to continue the homeless initiative, recalling the immortal words of Slim Charles at the end of Season Three.

Omar continues killing his way through the Stanfield organization, but is looking less invincible by the hour, with Kenard, as usual, calling out the truth. Saint Gus says to hell with the nut graf, go talk to the people! Kima sings the traditional Baltimore classic "Goodnight Fiends, Goodnight Hoppers" lullabye to Elijah, which the writer of this episode, Richard Price, borrowed from his book Clockers. I'm loading that into the MP3-enabled crib I buy for my kids, if and when I have some. I wonder if they'll sell it at IKEA?

Gratuitous, slightly jarring cameo of the week: Munch from Homicide. Poot, meanwhile, is clearly finalizing plans for his Michael Corleone-style simultaneous assassination of the entire board of the former New Day Co-Op.

Earlier in the season, State's Attorney OBonda turned down the opportunity to take the Clay Davis case federal with the "head shot" real estate charge. Oops! Lest anyone think the acquittal is implausible, I know several people who have worked in the Baltimore federal court system, and they all say this is the main reason cases up there go federal--avoiding jury nullification. While the state jury pool is local, federal cases draw juries from the whole northern half of the state, which means a black man from West Baltimore like Clay Davis who avails himself of his Constitutional right to a jury of his peers ends up being tried by an all-white jury comprised of people from the Eastern Shore and the westernmost rural counties of the state.

Also, there's a been a fair amount of bloggy discussion about whether Prop Joe really would have let himself get played like that. I think the answer is: sure. He wasn't ominiscient, he was thinking two steps ahead and Marlo was thinking three. Every day you play the Game, you put yourself in harm's way; Joe was lucky he lasted that long. Plus, much as I enjoyed Joe as a character, let's not forget that he was just as much of a dirtbag drug-dealing murderer as all the rest. He was smart, his families had roots in the community, he could have done a lot of things with his life and he decided to spend it spreading violence and poison throughout East Baltimore. Think about it: what was the practical result of Joe and Stringer's Co-op innovation? A smoother running heroin and crack distribution network, resulting in a more consistent supply of product, higher profits for the dealers, and lower prices and greater availability for the consumers. In other words, the worst possible thing for the community. Think of people like Bubbles, or Jonny (RIP), or the drug prostitute who kicked off an earlier episode this season. The Co-op meant more addiction, more death, more people becoming those people, and Joe was responsible. He got what he deserved.