There is a thing in statistics call an outlier. This is something that doesn't agree with the normal range of statistics, and when using statistics to predict future performance, outliers are commonly thrown out since they aren't representative of what is likely to happen. Shawn Chacon's half-season with the Yankees falls into this category. Normal fluctuation in pitcher's BABIP is around +/- .020 from season to season. Chacon's career average BABIP away from Coors' Field is .280 (I looked at his numbers on the road because the drier/thinner air can affect his curve, his best pitch). BABIP is the batting average on balls in play against Chacon, Team is the team average without Chacon pitching, DifC is the difference from his career average, DifT is the difference from the team average without him.
Split BABIP Team DifC DifT
COL-Road .280 .300 .000 -.020
w/ NY .239 .302 -.041 -.063
Chacon has demonstrated a consistent ability to limit batters to less hits on balls in play than the team he plays for, but his 2005 numbers from NY are completely ridiculous. When his DifT comes back to reasonable numbers, even despite his ability to outperform the rest of the team, his ERA is very close to the team ERA without him.
Split ERA Team DifT
COL-Road 4.84 4.91 -.07
w/ NY 2.85 4.62 -1.77
Now, this is baseball, not pure statisitics so throwing out his performance after coming to NY wouldn't necessarily be the right thing to do, but it certainly isn't wise to base expectations on it. My prediction, Chacon falls back to earth this year, hard.