Using Computer Simulation: In Search of the Perfect Curve Ball
We tend to think of computer simulation being used for scientific and industrial purposes. What if the technology could be used to, say, improve the performance of a Major League Baseball pitcher?
St. Louis Cardinals pitcher Adam Wainwright had a rough first inning in the game 1 of the 2013 World Series against the Red Sox at Fenway Park. Wainwright, typically a strike-throwing machine, had a tough time finding the plate. His normally devastating curve ball lacked its typical bite. What was wrong?
It could have been the weather: It was cold. It could have been World Series jitters: It’s a big stage. But whatever the reason, for the early part of the game Wainwright wasn’t himself.
Imagine there was another option.
Suppose Wainwright had access to a pitching motion tool that captured and analyzed his pitching motion, so he (and the Cardinals’ coaches) could pick up where the mistakes were, and allow him to make adjustments. It turns out such computer simulation software exists.
Convergent Science, a Middleton, Wisconsin company, simulated Wainwright’s curveball using computational fluid dynamics. Using their software, the company claims, Convergent Science could help Wainwright (or another pitcher; I’m a Red Sox fan so I don’t mind if one of my guys listens) understand how to tweak variables such as force, spin, and axis to achieve optimal pitching results. The company achieves this, it says, by graphically depicting the flow of fluid, gas, or liquid and applying what they learn to a pitcher's motion (or to just about anything that moves, but who cheers so loudly for optimizing auto engine performance or perfecting a wind farm?).
Over lunch, several Convergent Science employees – who happen to include fans of the St Louis Cardinals, the National League entry in this year's World Series – came up with the idea of using the software to simulate Wainwright's curve ball, says Rob Kaczmarek, the company’s director of sales and marketing.
“Of course, [the Cardinals fans] went on and on about how Wainwright was going to demolish [the Red Sox] with his curveball. The seeds of simulating just what’s happening in that curveball were planted that day,” he says.
It isn’t as though the pitcher is unaware of his mechanics. In this interview on MLB Network with former Major League pitcher Al Leiter, Wainwright explains what goes into throwing his curve, in great detail.
As Wainwright explains, his curve ball is a combination of grip, motion, and release, all coming together to deliver a pitch that, much of the time, fools the best major league hitters. (Except maybe not for the first couple of innings last Wednesday night.)
It’s the kind of information that the Convergent software might illustrate and document. As Kaczmarek explains, “Wainwright puts a huge amount of spin on the ball in order to achieve his 55 inch drop. What we found was in order to get the amount of curve Wainwright is able to achieve, he has to put 1,800+ RPMS on the ball."
So how does it Convergent measure that?
According to company-provided information: "Traditional [Computational Fluid Dynamics (CFD)] software would rotate the baseball in a fixed position while Convergent’s software moves the software through space and time from the pitcher’s mound to the plate. The simulation starts by subdividing the 60 feet 6 inches of air from the pitcher’s mound to the plate into tiny cells, then simulates the ball cutting through these cells, and calculates the effect of each cell on the ball’s motion."
Wainwright, his pitching coach, or any other pitcher could use this tool – theoretically, at least – to analyze his motion and figure out the ideal release point (to the extent, of course, that any human can repeat a motion and release to the point of perfection every time). If the pitcher were going through a time where he lost his release point, the simulation could help him find their way back to success.
It's not unusual for successful major league baseball pitchers to “lose it” for periods of time and give up lots of hits in an uncharacteristic fashion. Often, the fix is a series of minor adjustments, with the pitcher working with the team’s pitching coach to refine the pitching motion and delivery. But the process can take weeks or months of trial and error, including reviewing video and other techniques.
Using the Convergent tool, a faltering pitcher might be able to find his way back faster, or it could even help successful pitchers refine their motion ever more.
"The ability to simulate 20% more or less spin on a ball would be highly beneficial to coaches and pitchers alike. Imagine if you could predict accurately what 20% more spin would do to a ball," Kaczmarek explained.
At the very least, Kaczmarek says, "The simulation determines what spin speed, spin axis, and force must be applied to the ball by the pitcher to move the ball by a certain amount as it travels to the plate. So the simulation can be used to determine exactly what the pitcher needs to do to hit a specific point inside or outside the strike zone."
The Convergent simulations can help batters and fielders too. "We could use this technology to simulate the effects of wind on baseballs to help batters better place their hits and fielders better anticipate where hits may be headed," Kaczmarek explains.
The same simulation software and techniques can be used across any number of industries. Convergent’s software is used to look at fluid mechanics and flows to help improve all sorts of processes: everything from engine performance, efficiency, and reducing emissions in the automotive market, to optimizing wind farms to achieve the maximum output from a given terrain, says Kaczmarek.
It’s unlikely that Wainwright is working without help. Major league baseball players have lots of analysis tools to work with these days, far beyond the velocity measured by a radar gun. Some teams are using a tool called Trackman to measure a variety of pitching variables, such as extension, spin rate, and speed. The Sports Performance Center at Massachusetts General Hospital has a software package to create an avatar for a pitcher, to analyze his motion and detect flaws in even greater detail than Convergent.
The San Francisco Giants have video cameras built into the light stanchions at AT&T Park, which generate data to help coaches improve players’ pitching, batting, and fielding. In fact, video – and quick access to it – is the “secret sauce” for most baseball teams today.
Baseball is not alone in using technology to drive better performance. RedFir, a German company, has a system for soccer teams to measure performance of individual players and teams. They use sensors on players, a chip inside the ball, and receivers on the field, combined with software to help the team get a better picture of player performance and to aid in training.
Sports is driven by the talent of individual players. No piece of software is going to make a player something he or she's not. However, simulation tools like these can help players refine their playing technique and perhaps get the most out of their talent – and that's an exciting prospect.