Moneyball by Lewis, Michael (mobile ebook reader txt) 📕

the morning and to the ballpark was to remind himself that he wasn’t pitching only for the White Sox. “By the middle of the 1999 season I was pitching for every other big league team that might be watching,” he said. “I’m just sitting there hoping someone’s watching.”

Someone was.

The pool of people Chad Bradford didn’t know who had nevertheless found him worthy of their attention had tripled. When he was an amateur, one big league scout had taken an interest in him. As a professional, he had two more distant admirers. One was Paul DePodesta, who couldn’t quite believe that the White Sox were keeping this deadly pitching force in Triple-A, and had mentioned to Billy Beane how nice it would be if he somehow could talk the Chicago White Sox into making Chad Bradford an Oakland A. The other was a bored paralegal in Chicago named Voros McCracken. Looking for a way to ignore whatever he was meant to be doing for the Chicago law firm that he loathed working for, Voros McCracken had taken up fantasy baseball. He didn’t know it, but he was about to explain why the Chicago White Sox had so much trouble grasping the true value of Chad Bradfordand why the Oakland A’s did not.

Voros was thinking of drafting Chad Bradford for his fantasy baseball team. But before he did, he wanted to achieve a better understanding of major league pitching. Specifically, he wanted to know how you could tell if a pitcher was any good.

Voros had played baseball as a kid, and there had been a time when he had been obsessively interested in it. The moment his interest became an intellectual obsession was in 1986 when, at the age of fourteen, he picked up the most recent Bill James Abstract. He was astonished by what he found inside. “Basically, everything you know about baseball when you are fourteen years old, you know from baseball announcers,” said Voros. “Here was this guy who was telling me that at least eighty percent of what baseball announcers told me was complete bullshit, and then explained very convincingly why it was.” Voros’s interest in baseball waned in his late teens and early twenties; but when he rediscovered it, in the form of fantasy leagues on the Internet, it was in the spirit of Bill James.

The Internet of course had consequences for the search for new baseball knowledge. One of the things the Internet was good for was gathering together people in different places who shared a common interest. Internet discussion groups, and Web sites like baseballprimer and baseballprospectus, sprung up, created by young men who, as boys, had been seduced by the writings of Bill James. In one of the discussion groups, where he went to discuss what to do with his fantasy baseball team, Voros saw someone say that no matter how much research was done, no one would be able to distinguish pitching from defense. That is, no one would ever come up with good fielding statistics or, therefore, good pitching statistics. If you don’t know how to credit the fielder for what happens after a ball gets put into play, you also, by definition, don’t know how to debit the pitcher. And, therefore, you would never be able to say with real certainty how good any given pitcher was. Or, for that matter, any given fielder.

When Voros read that, “I thought, ‘That’s a stupid attitude. Can’t you do something?’ It didn’t make any sense to me that the way to approach the problem was to give up.” He tried to think about it logically. He divided the stats a pitcher had that the defense behind him could affect (hits and earned runs) from the stats a pitcher did all by himself (walks, strikeouts, and home runs). He then ranked all the pitchers in the big leagues by this second category. When he ran the stats for the 1999 season, he wound up with a list topped by these five: Randy Johnson, Kevin Brown, Pedro Martinez, Greg Maddux, and Mike Mussina. “I looked at that list,” said Voros, “and said, ‘Damn, that looks like the five best pitchers in baseball.’” He then asked the question: if his reductive approach of looking at just walks, strikeouts, and homers identified the five best pitchers in baseball, how important could all the other stuff be?

As it happened, 1999 was supposedly an “off” year for Greg Maddux. His earned run average had risen from 2.22 in 1998 to 3.57 in 1999, mainly because he gave up fifty-seven more hits in thirty-two fewer innings. Several times during the season Maddux himself mentioned that he was astonished by how many cheap hits he was giving up; but of course no one paid any attention to that. What Voros noticed was that Maddux’s hits allowed per balls put in play was far above what it usually was—in fact, it was among the highest in the big leagues. As it happened, the same year, Maddux’s teammate, pitcher Kevin Millwood, had one of the lowest hits allowed per balls in play. Even stranger, their statistics the following year were reversed. Millwood had one of the highest ratios of hits per balls in play, Maddux one of the lowest. It didn’t make any sense.

Voros asked himself another question: from year to year is there any correlation in a pitcher’s statistics? There was. The number of walks and home runs he gave up, and the number of strikeouts he recorded were, if not predictable, at least understandable. A guy who struck out a lot of hitters one year tended to strike out a lot of hitters the next year. Ditto a guy who gave up a lot of home runs. But when it came to the number of hits per balls in play a pitcher gave up, there was no correlation whatsoever.

It was then that a radical thought struck Voros McCracken:

What if the pitcher has no control of whether a ball falls for