Dan Rosenheck discovers that spiking ZIPs projections with spring training numbers improves the ZIPs projections!

The answer was an unequivocal yes. In every peripheral category, forecasts that included a finely calibrated dose of spring-training numbers outperformed ZiPS by itself. The impact was particularly strong for first-year players (“rookies”), for whom spring training is their first taste of proper big-league competition. After adding the peripherals back together to get an all-in-one value measure, incorporating spring training improved the correlation between preseason projections and final results from .578 to .593 for hitters (using OPS) and from .354 to .387 for pitchers (using ERA).

Via Rob Neyer, show says this about Dan:

Dan sorta flies under the radar in these parts, because he doesn’t publish often, and when he does publish it’s just under his initials (because of editorial policy where he works). But he’s crazy smart and enthusiastic and opinionated, which makes for a delightful combination. For me, the worst thing about missing the MIT-Sloan Analytics Conference last week was not getting to see Dan.

I know Dan as well and I agree. I don’t agree, however, that this is as huge a break through as Dan claims in the article. Let’s revisit this line:

The impact was particularly strong for first-year players (“rookies”), for whom spring training is their first taste of proper big-league competition.

In other words, player for whom the projections are very close to league average. If spring numbers move these projections off average in the right direction, low and behold the correlation improves. What I’d like to see from Dan how much the numbers improve for players with at least 600 PA in the majors. I suspect it’s much smaller, and for them spring training doesn’t matter as much.