Before there was Moneyball, there was Davey Johnson.
One of the many things that stood out to me when I read Gary Carter’s books over and over and over in the late 1980s and early ’90s was his disdain for Davey Johnson’s computers.
Carter, the catcher for the Mets during the glory years, very much liked Johnson, the manager for those great teams. He just didn’t like Johnson’s computers, his printouts or his penchant for hunkering down in his office and using stats to prepare for games in the hours leading up to first pitch.
“Davey Johnson had come up to manage the Mets from their Triple-A team in Tidewater after the ’83 season,” Carter wrote in his book, ‘The Gamer.’ “Davey managed by the numbers…literally. He was a computer whiz who could pull up more baseball statistics on his screen than most of us knew existed. I liked Davey, but I didn’t care much for his computer.
“Not that there’s anything wrong with having all the latest stats available, but it’s tough to quantify qualities such as ‘heart,’ ‘desire,’ and ‘intestinal fortitude.’ Moreover, Davey’s computer might dutifully report that a player went hitless in four at-bats but totally miss the fact that all four outs were hard-hit balls or long fly balls just inches away from being home runs.”
Carter’s sentiments were in line with most baseball insiders back then. Constructing a lineup was based mostly on a player’s batting average and how he fared in the past against the opponent’s starting pitcher, and not much else. Sure, there were debates about where players should hit in the order and how valuable — or wasteful — a sacrifice bunt was, but for the most part, baseball analysis was not something that necessitated the use of a computer to help crunch the numbers. Or so the “experts” thought.
Times have changed, obviously. Johnson’s insistence on including computers into his daily routine wasn’t kooky, it was innovative. He was ahead of his time not only as a manager, but as a player, too. Apparently, his mathematics-minded approach to baseball dates all the way back to the 1960s, when he played for the Orioles.
Johnson’s motives were simple: he was hitting seventh, and he wanted to be hitting second. He needed to come up with a way to convince manager Earl Weaver to move him up, so he came up with a computer program designed to back up his philosophy. He presented it to Weaver as a way to, as he put it, “optimize the Orioles lineup.”
It was a tough sell. Weaver liked matchups. He liked using his hitters who had the best numbers against that day’s pitcher in the middle of the order, regardless of how they’d performed over the course of the season. If an MVP-caliber player had poor numbers against a specific opponent, chances were, he was going to sit that day. Other than that, Weaver’s favorite thing to do was rely on his pitching and defense, and wait for the long ball.
Johnson, who obtained a mathematics degree from Trinity University, was a formidable challenge for Weaver. Johnson wasn’t the prototypical, cookie-cutter ballplayer — he was smart, and he was curious. While finishing his coursework at Trinity University, he took a computer class at Johns Hopkins University, just for the heck of it. It was during that time he had lunch with an engineer-turned-author named Earnshaw Cook, who left an immeasurable impression on the numbers-minded Johnson.
Cook, also a professor at Johns Hopkins, had written a book called “Percentage Baseball.” In hindsight, it was sabermetrics before sabermetrics. It was also rebuffed by a baseball community unwilling to embrace Cook’s oddball philosophies that suggested the best hitters should bat first, that sacrifice bunts were wasteful and relief pitchers — get this — should start games and then be replaced by a pinch-hitter when it came time for his first at-bat. Starting pitchers would then enter the game and pitch the next four or five innings. This, Cook argued, could tack on another 25 wins for a team.
Cook’s theories piqued the interest of a few people, according to this Alan Schwartz feature — Tal Smith, a young, up-and-coming executive with the Astros and self-proclaimed stat nerd, Ewing Kauffman, owner of the Kansas City Royals, and Johnson.
While Johnson didn’t follow Cook’s philosophy to the letter, Cook’s views gave him some level of validation as he worked on his computer programs to solve baseball riddles.
Now it was just a matter of convincing Weaver. Johnson would take his giant optimization printouts, generated from his oversized computer, and present them to his manager.
“Earl, do you know what the standard deviation chart is?” Johnson recalled asking. Weaver’s response was dismissive, but Johnson plowed on. The standard deviation chart, Johnson explained, requires a larger pool by which to make judgments, to be able to predict with a plus or minus five percent. If you flip a coin five times, you might get five heads. If you flip it 500 times, it’s more likely to be half heads and half tails, or close to it.
“Six at-bats doesn’t really give you a good way to predict,” Johnson told Weaver. “There’s more things involved.”
When Johnson started managing, he developed a program to obtain data from the opposing manager to gauge hitters’ tendencies and to determine what counts were more favorable when a manager decided to hit and run. He pored over stats to determine the best possible batting orders. He viewed his computer as “another coach, with a better memory than me.”
As a player, Johnson and his theories didn’t get very far with his manager, and as a manager, his players disregarded his computer with a flippant wave. He didn’t get far with his own teammates, either. He recalled watching pitcher Dave McNally struggle to hit the inside corners during one particular start. Between innings, Johnson said to him, “You’re in an unfavorable chance deviation.”
In other words, stop aiming at the corners. Aim it down the middle, and you’ll hit the corners.
Predictably, that particular bit of advice didn’t go over very well.
“From that, I was nicknamed ‘Dum-Dum,’” Johnson said.