When baseball's elders swap stories about fireballers, the name that ends the conversation isn't Nolan Ryan or Sandy Koufax. It's one that never appeared on the back of a major-league uniform: Steve Dalkowski. Legend has it that the 5-foot-11-inch, 170-pound lefty threw his fastball well in excess of 100 mph. We don't have an exact number for the same reason Dalkowski, who toiled in the minors in the late 1950s and early 1960s, never made the big leagues: He was too wild to time. When a scout tried to gauge Dalkowski's fastball with a primitive radar gun—a beam of light the width of home plate—the pitcher couldn't hit the target until after his arm got tired.
Steve Dalkowski sounds like a genetic freak, but so is anyone who can throw a baseball 90 mph. What he really represents is a blow to the basic notion of human progress. In almost every measurable physical activity, athletes show improvement over time. Jumpers jump higher and farther, and runners and swimmers go faster. Since the late 1950s, the high-jump world record has improved by more than 10 percent, the 100-meter-dash mark has improved by 5 percent, and swimming's best 100-meter freestyle has dipped 12 percent.
Pitchers, though, don't seem to be getting any faster. Pretty much every generation since the early 1900s has boasted a supposed 100-mph pitcher, from Walter Johnson and Smoky Joe Wood to Bob Feller to Dalkowski to Nolan Ryan. If we stick with speeds registered since modern radar guns became ubiquitous in the 1970s, peak velocity seems to be a shade north of 100. Major League Baseball doesn't keep official records on pitch speeds, but the Guinness Book of World Records credits Ryan with the fastest pitch ever, a 100.9-mph heater from 1974. This article disagrees, crowning Mark Wohlers the radar-gun champ with a 103-mph pitch. (For an explanation of why radar gun readings can be inconsistent, click here.)
Maybe it only looks like the outer limit for pitchers is stable at around 100 mph because we can't consistently and accurately measure minute improvements in speed. When it comes to flamethrowers, after all, it's hard to figure out what's the truth and what's a tall tale. Feller once sent a fastball zooming by a speeding motorcycle. Maybe the ball really was traveling at 104 mph, as the organizers of the stunt claimed. Or maybe not.
Still, according to experts in biomechanics, that 100-mph ceiling isn't an illusion—it's a basic property of human physiology. A pitcher generates momentum by rocking onto his back leg and thrusting forward. After that he rotates his pelvis and upper trunk, then his elbow, shoulder, and wrist. Intuitively, it seems like building up the muscles in the legs, upper body, arm, and shoulder would generate more force and make his arm move faster. The reality: There's a point when more torque doesn't yield a faster pitch. It simply causes tendons and ligaments to snap, detaching muscles from bones and bones from one another. (Tendons connect muscles to bones; ligaments connect bones to each other.)
Glenn Fleisig, a biomechanical engineerwho studies pitching at the American Sports Medicine Institute in Birmingham, Ala., has calculated that about 80 Newton-meters of torque act on an elite pitcher's elbow when he throws a fastball. The ulnar collateral ligament connects the humerus and ulna—two of the bones that come together in the elbow. To test the outer limits of the ligament's strength, Fleisig subjected cadaver elbows to increasing amounts of rotational force. These experiments showed that an average person's UCL snaps at about 80 Newton-meters. Smoky Joe Wood said that he threw so fast he thought his arm was going to fly off. It turns out he wasn't far from the truth.
Another way to test the proposition that ligament fragility limits velocity is to see what happens when pitchers strengthen their muscles. Mike Axe, an orthopedic surgeon and protégé of Fleisig's partner James Andrews, advises pitchers to build up their shoulder muscles by practicing with a weighted glove on their throwing hand. According to Axe, a pitcher can add up 2 to 5 mph to his fastball with this regimen. The potential gains are lower for those who throw fast to begin with, though. Axe has seen pitchers increase their velocity from 84 to 88 mph and from 88 to 91 mph. He's never seen anyone improve from 98 to 100. The chief benefit for these hurlers is that they suffer fewer muscle tears.
Why do sprinters keep getting faster while baseball pitchers seem to have maxed out? Because track athletes don't approach the limits of what human tendons and ligaments can handle. When you run the 100-meter dash, no single stride represents as violent a motion as the arm makes during a single overhand pitch. Sprinters can build up their muscles without worrying that the extra force will rip their ligaments apart—that's why steroid use seems to make sprinters faster but won't help pitchers generate velocity beyond a certain point. (A better reason for a pitcher to take steroids would be to decrease the time it takes to recover between games.)
Ligaments and tendons can get stronger, but at a much slower rate than the muscles that surround them. There are rumors that pitchers who've undergone Tommy John surgery—that is, a replacement of the UCL with a tendon from the hamstring or wrist—can throw harder than they did before having surgery. But any increase in velocity probably has less to do with getting a new superligament than with the strict rehabilitation program Tommy John patients are supposed to follow. The reason pitchers get injured in the first place is that their muscles, tendons, and ligaments weren't as strong as they should have been.
What about growing taller, more massive pitchers? That doesn't necessarily make a difference, either. Small, slightly built pitchers like Dalkowski, the 5-foot-11 Pedro Martinez, and the 5-foot-10 Billy Wagner throw just as hard as giants such as Randy Johnson. The physical principle here is fairly simple. If two levers move at the same speed, the ball released from the longer lever will have more velocity. But as a lever becomes larger, it requires more torque to move. Randy's lever is larger; Wagner's moves more quickly. The trade-off makes their velocity roughly equal.