Jfish, have you read "The Book" by Tango, Lichtman and Dolphin? Based on their empirical data, Dan is right on. I quote: "A right-handed hitter needs around 2,000 appearances against left-handed pitchers before his measured platoon split can be considered reliable. ... For lefties, the number is about 1,000, which means that only veteran starters have reliable platoon splits" (163). For pitchers, the same holds true, though the numbers are smaller: 700 for righties facing lefties, and 450 for lefites facing righties (168). You'll see a difference after 500 ABs, but what you see may not be a good predictor of reality.

I think Bill Staffa over at Skeetersoft came to the same conclusion regarding splits.