The New Math

Some fascinating new work in the most recent issue of SABR's Baseball Research Journal. One story that might have proved useful a few months ago: Victor Wang's investigation of exactly how much a prospect is worth, and when they should or should not be traded. Wang culled Baseball America's Top Prospects over a decade period, and tracked their performance (in WARP) during the first 6 years of their careers (those being the years that they would be controlled by their original teams). These players were then ranked into four relative categories, from star to bust, given their overall WARP scores, and the results were stark. For hitters, the likelihood of a top prospect developing into a star was 17 percent. Bust was 21 percent. The middle categories each scored around 30 percent. Pitching prospects faired even more poorly: 54 percent were busts, only 3.8 percent stars, and the balance in the middle shifted toward the lower end. The take-away is somewhat more complicated than these numbers suggest: given that a team will get six years of WARP out of a prospect, and at low cost, there's still considerable advantage in retaining them. A trade of prospects for a 1-year rental is generally not worth the trade, unless what comes back pushes a team into title contention. Worth the read.

Also valuable: an analysis of the "gyroball" from Alan Nathan and David Baldwin. They argue that the pitch does exist, explaining that it is thrown by generating spin on the ball that is perpendicular to the flight path of the ball. The result is a pitch that is fast but not quite as fast as a fastball, and curves but not quite as drastically as a curve or slider. The advantage is that it works as something of an off-speed pitch, and the spin can be deceptive. They also used some pretty advanced pitch tracking technology to determine whether Daisuke Matsuzaka actually throws the thing. By examining a graph that plots pitch velocity and break, they conclude that Matsuzaka does, indeed, throw the gyroball, but only with extreme rarity.

Comments

[Lar]

They should also plot other pitcher's velocity/break to see percentage - it might not be statistically significant (thrown it by accident or something..)

[LocklandSF]

Tis the season of minutia, I love me some minutia.

Good stuff YF.

[Nick-YF]

I wonder what the likelihood of any player developing into a star, let alone a reliable player. What percentage of non-prospects (eg. Piazza) develop into stars or even reliable players?