This piece on Michael Lewis and Billy Beane is nice to read, even if you haven't read Moneyball. (And if you haven't, consider doing so - it's not perfect, but it's well worth the time). Several thoughts occurred to me while revisiting all this, some of them actually relevant to drug discovery.
First off, a quick peaen to Bill James. I read his Baseball Abstract books every year back in the 1980s, and found them exhilarating. And that's not just because I was following baseball closely. I was in grad school, and was up to my earlobes in day-to-day scientific research for the first time, and here was someone who applied the same worldview to a sport. Baseball had long been full of slogans and sayings, folk wisdom and beliefs, and James was willing to dig through the numbers to see which of these things were true and which weren't. His willingness to point out those latter cases, and the level of evidence he brought to those takedowns, was wonderful to see. I still have a lot of James' thoughts in my head; his books may well have changed my life a bit. I was already inclined that way, but his example of fearlessly questioning Stuff That Everybody Knows really strengthened my resolve to try to do the same.
A lot of people feel that way, I've found - there are James fans all over the place, people were were influenced the same way, at the same time, by the same books. It took a while for that attitude to penetrate the sport that those books were written about, though, as that article linked to above details. And its success once it did was part of a broader trend:
Innovation hurts. After Beane began using numbers to find players, the A’s’ scouts lost their lifelong purpose. In the movie, one of them protests to Pitt: “You are discarding what scouts have done for 150 years.” That was exactly right. Similar fates had been befalling all sorts of lesser-educated American men for years, though the process is more noticeable now than it was in 2003 when Moneyball first appeared. The book, Lewis agrees, is partly “about the intellectualisation of a previously not intellectual job. This has happened in other spheres of American life. I think the reason I saw the story so quickly is, this is exactly what happened on Wall Street while I was there. . .”
(That would be during the time of Liar's Poker, which still a fun and interesting book to read, although it describes a time that's much longer ago than the calendar would indicate). And I think that the point is a good one. I'd add that the process has also been driven by the availability of computing power. When you had to bash the numbers by hand, with a pencil, there was only so much you could do. Spreadsheets and statistical software, graphing programs and databases - these have allowed people to extract meaning from numbers without having to haul up every shovelful by hand. And it's given power to those people who are adept at extracting that meaning (or at least, to the people willing to act on their conclusions).
The article quotes Beane as saying that Lewis understood what he was doing within minutes: "You’re arbitraging the mispricing of baseball players". And I don't think that it can be put in fewer words: that's exactly what someone with a Wall Street background would make of it, and it's exactly right. Now to our own business. Can you think of an industry whose assets are mispriced more grievously, and more routinely, than drug research?
Think about it. All those preclinical programs that never quite work out. All those targets that don't turn out to be the right target when you get to Phase II. All those compounds that blow up in Phase III because of unexpected toxicity. By working on them, by putting time and effort and money into them, we're pricing them. And too much of the time, we're getting that price wrong, terribly wrong.
That's what struck me when I read Moneyball several years ago. The problem is, drug research is not baseball, circa 1985. We're already full of statisticians, computational wizards, and sharp-eyed people who are used to challenging the evidence and weighing the facts. And even with that, this is the state we're in. The history of drug research is one attempt after another to find some edge, some understanding, that can be used to correct that constant mispricing of our assets. What to do? If the salt has lost its savour, wherewith shall it be salted?