Whenever skeptics told pitcher Dizzy Dean that his curveball was just an optical illusion, he shot back the same snarky reply:
"Stand behind a tree 60 feet away, and I'll whomp you with an optical illusion."
If the ballplayers of the 1930s had any sense, they took the Hall of Famer at his word. But it wasn't until 1959 that someone proved that curve balls really do curve — not by beaning a batter but by using physics.
The baseball breakthrough came, thanks to Lyman Briggs, a former director of the National Institute for Standards and Technology who had helped with research that led to the atomic bomb. A onetime outfielder on the Michigan State baseball team, he was deeply interested in the science of the sport, so he talked Washington Senators manager Cookie Lavagetto into letting him study the team's pitchers.
Next, Briggs set up his own pitching machine in an unused industrial building at the NIST campus. Every so often, visitors would be startled by the loud "bang" of a baseball being fired at its target, a small piece of paper posted 60 feet away. At least once, Briggs's experiments ended in a shattered window — even machines throw the occasional wild pitch.
On March 28, 1959, Briggs published the results of two years' hard work.
"U.S. Upholds Curve" read the headline in the next day's New York Times. "The United States Government, taking a breather from such weighty matters as Berlin, Iraq, atomic energy and taxes, announced yesterday that a baseball really curves."
Curveballs curve — or break downward — because of the spin imparted by the pitcher as he flings it toward home plate. The way Briggs explained it, the rotation of the seams creates a "whirlpool" of air around the ball and causes the pressure to be lower on one side.
"The difference in pressure tends to push the ball sideways or to make it curve," he told the Times.
In the 60-foot, 6-inch journey from mound to plate, a curveball can break up to 17.5 inches, Briggs concluded. But it has to be pitched slowly — optimum velocity was just 68 mph.
All that physics didn't do much good for Lavagetto and the Washington Senators. The team finished at the bottom of the American League that year, with an undistinguished 63-91 record. But it did seem to put an end to curiosity about the curveball's curve — for a while.
Half a century later, neuroscientist Arthur Shapiro took another crack at the question. Shapiro specializes in the science of visual perception — particularly optical illusions — but he's also a longtime baseball fan. (When we asked for his team, he demurred: "That's a dangerous one to put out there," he joked.)
"There's no question that the curveball is curving, but there's also a question about 'what do we perceive?' " Shapiro told The Post.
Viewed from the batter's box, the wickedest curveballs seem to drop suddenly just before they arrive at the plate. Yet physics tells us that's impossible — an object in motion might descend gradually but definitely not all at once.
"There’s the physics and then there's the perception, and in order to see something the physics has to be translated to the eye and the brain," Shapiro said. " So the brain has a really hard problem to solve, because it has to figure out where the ball is going to be from the very first 0.2 seconds of the pitch."
In a 2010 study in the journal PLOS One, Shapiro summarized how optical illusions might explain that deceptive drop. As the ball moves through our field of vision, its position and rotation appear to change because they're being processed by different parts of the visual system, much the same way that viewing an optical illusion in your peripheral vision will change the way your perceive it.
(To test this for yourself, take a look at this interactive Shapiro developed for his study. If you focus on the purple fixation point on the right hand side, the spinning disc will appear to curve away as it falls.)
It's just a hypothesis, Shapiro said. And whether the break of a curveball is pure physics or partly an illusion, we can all agree on one thing:
"It's still hard to hit," he said.