Old analytics bring a new urgency to MLB

Baseball is doing what Earnshaw Cook insisted upon, sixty years ago.

Roger Maris, Mickey Mantle

Long before there was Bill James and his vast array of baseball codifiers, there was Earnshaw Cook.

He was a metallurgist and, eventually, a professor of industrial engineering at Johns Hopkins in Baltimore, and in 1964 he was the subject of a Sports Illustrated piece called “Baseball Is Played All Wrong,” based on his computerized analyses. In those days, computers themselves were about as big as Boog Powell, the Orioles’ lefthanded slugger. Yet Cook’s theories has somehow trickled down to the major league dugouts of today.

Cook found that starting pitchers should work a couple of innings, be removed for a pinch-hitter, and be relieved by a better pitcher who would work five or so innings. Then the guy whom we now know as the “closer” would work the eighth and the ninth. The theory was that pitchers should not hit, and pinch-hitters would increase the run totals by a dramatic margin. That, of course, is very close to what we have now, except the “opener” only works an inning in most cases, and six or seven pitchers are needed to handle the nine innings. Walter Alston, then manager of the Dodgers, worried that you’d need 15 pitchers to handle such an unconventional workload, a theory at which Cook scoffed, but there are usually at least 12 and more often 13 pitchers on any major league staff, and injuries continue their long, destructive march. Note that Cook, in his book Percentage Baseball, wrote this stuff nine years before the Designated Hitter.

Cook also said that batting orders also made little sense. Using what he called the Player Index, which is somewhere between OPS and the various iterations of WAR, he said that runs would increase if the best hitters hit first. Obviously they would get more at bats. Managers and general managers who never heard of Cook have begun subscribing to his 1-2 punch.

When Alston managed the Dodgers, Maury Wills led off and Jim Gilliam hit second. The theory was that Wills would walk or single, steal second, get moved to third by Gilliam via bunt or right-side grounder, and score when one of the popguns behind Gillilam managed a sacrifice fly. After Tommy Davis was hurt and Frank Howard was traded, the Dodgers did not have sluggers on the level of Mookie Betts and Freddie Freeman today.

But Betts and Freeman hit 1-2 for the Dodgers. Last year they were the only hitters on the club who had on-base percentage of .400 or higher. The Dodgers again led the National League in runs, too.

The Brewers made a leadoff man out of Christian Yelich. He had an .818 OPS and led the club in steals with 28. But in 2018, when Yelich had a 1.000 OPS and was the MVP of the National League, he batted second. The leadoff man was Lorenzo Cain, who had 71 walks and 30 steals. His on-base percentage was .395, Yelich’s was .402.

The next year, Yelich led the league in batting average, OBP and slugging and was the MVP runnerup to Cody Bellinger of the Dodgers. Bellinger usually hit fourth that season, despite his 1.035 OPS. Ahead of him were Joc Pederson (.876), Max Muncy (.889) and Justin Turner 8.881).

The Atlanta Braves in 2023 had a guy who stole 73 bases with a .416 OBP, so of course he led off. Except that he also had 41 home runs. That was Ronald Acuna Jr., who was the unanimous MVP because of it. It worked because the Braves had three other guys who hit at least 37 home runs hitting behind Acuna, but who knows where Acuna would have hit when Cook was cranking up his Univac? Back then, in 1964, the Braves were managed by Bobby Bragan, who experimented with Cook-like lineups from time to time. He sometimes used Eddie Mathews as a leadoff hitter even though Mathews was slow and had an OBP of .344. But Mathews would reach the Hall of Fame with 512 home runs. Bragan also had Hank Aaron, who hit .328 that year with an OPS of .907. Aaron hit third, almost without exception.

Last year’s Baltimore Orioles had two dynamic young hitters, rookie Gunnar Henderson and Adley Rutschman. Henderson stole 10 bases, Rutschman one. But they had the two best OPSs on the team. Cedric Mullins had been the leadoff man when the Orioles were cellar-dwelling, because he stole bases, but no more.

Last year’s Seattle Mariners led with J.P. Crawford and followed with Julio Rodriguez more often than not. They both had an .818 OPS, best on the team. Crawford, in fact, had a .380 OBP, but in olden times he likely would not have led off. Rodriguez, with his 37 steals, might have, but he also had 32 home runs and drove in 103. The third and fourth hitters were Eugenio Suarez and Teoscar Hernandez, who might have hit lower in the lineup way back when. They combined for 48 home runs — and 425 strikeouts.

Willie Mays was one of the few power/speed masters in what we remember as the glory days. Mickey Mantle was another. When Clyde King was the Giants’ manager he fiddled with the idea that Mays should lead off. In doing so he alienated Mays, who would end up hitting 660 home runs. Today he would bat no lower than second.

Mantle’s Yankee teams usually featured less talented “table setters.” In 1956 Mantle won the Triple Crown (52 home runs, 130 RBI, 1.169 OPS, .705 slugging percentage, .464 on-base percentage). Mantle was the No. 3 hitter in a lineup card that could have been chiseled into the dugout wall. Hank Bauer was the most frequent leadoff hitter, with a .316 OBP, and Billy Martin hit second, with a .310 OBP. In 1961, Mantle hit 54 home runs and Roger Maris 61, and Cook would tell you that both could have hit more if he, Cook, had been writing the batting order. Mantle did hit second 18 times and Maris 10, but mostly they were 3-4, and Tony Kubek (.312 OPS) and Hector Lopez (.362) were the normal 1-2.

Another illustration from a championship team: Willie Stargell is a Hall of Famer and Dave Parker should have been one long ago. For the 1979 Pirates, Stargell had a .904 OPS and Parker .906. But Parker generally hit third and Stargell fourth, and Omar Moreno, who struck out 104 times, was the leadoff man, primarily because he had 77 steals. His OPS was .718.

Cook also launched a crusade against the sacrifice bunt and that, too, was successful. There were 1,462 successful sacrifices in 1964, the year Cook’s book came out. Last year there were 429. The Braves had….two. But Cook didn’t get to witness the extent of his vindications. He was 87 when he passed away in 1987.

The Moneyball book, by Michael Lewis, was not the cradle of applied baseball analytics. Neither was the Moneybook movie. Allan Roth had his own charts and spreadsheets for the Brooklyn Dodgers, whose general manager Branch Rickey was preaching the gospel of on-base percentage for years before that. When Dick Williams managed in the minor leagues, he went to art supply stores and bought large white sheets, and superimposed the patterns of the balls that the opponents hit. Now such things are known as “spray charts.”

Hugh Fullerton was the first to win the Spink Award, which honors baseball writers with an exhibit at the Hall of Fame (it has been renamed). Fullerton was played by Studs Terkel in the classic ball movie Eight Men Out, about the 1919 Black Sox. In the film, Fullerton and Ring Lardner had the suspicions that launched the investigation, but in real life it was Fullerton and Christy Mathewson, the great pitcher and moral conscience of the game. In any event, Fullerton devised his own analytical system, breaking down players at each position, and used it to predict World Series winners. He went with the good-field, no-hit White Sox to knock off the Cubs in 1906, which they did to widespread surprise.

With Cook’s ideas baked in, there are more than a few who think baseball is played wronger than ever. But the least you can say is that Cook dealt with real performers and real, tangible data, not projections and abstractions and “replacement players.” You also recognize that the man who’s ahead of his time doesn’t always see it catch up.