Carlos Zambrano has been the undeniable ace of the Cubs' staff the past three seasons. But those three seasons have seen a disturbing trend in his home runs allowed rate:

Zambrano has gone from allowing 0.38 home runs per 9 innings in 2003 to 0.85 in 2005, a number in any event well above the league average. He was still, of course, very good in 2005, posting a 131 ERA+ (following a 136 ERA+ in 2003 and a 165 ERA+ in 2004). And Zambrano's 2003 season was, for much of the season, historic for its lack of gopher balls. It is unreasonable to expect such things on a consistent basis.

I think I know, but can't prove, why Zambrano's home runs allowed rate has increased. It seemed to me that the radar gun on Zambrano's fastballs hit the high 90s much more in 2005 than in 2003. In 2003, the gun was almost always in the low 90s. I don't think Zambrano's become faster. I think he's always had the high 90s ability. In 2003, however, he made sure he got the necessary downward movement on his sinking fastball, and that meant taking off a few miles per hour. (That extra movement also, however, translated into a higher walk rate.) Zambrano's high 90s fastball, on the other hand, necessarily requires a straighter ball.

One of the reasons Zambrano continued to have success in 2005, despite the upturn in home runs allowed, was that he allowed a career low .212 batting average against. With similar strikeout rates in 2004 and 2005, Zambrano's having allowed fewer hits in 2005 was the result of allowing fewer hits on balls in play (BABIP):

(The league average BABIP is usually roughly .300.) Research has shown that a pitcher's BABIP is largely, though not completely, attributable to defense and luck (in other words, factors outside the control of a pitcher). We can expect Zambrano to return to the .280-.290 range in 2006. If he does, in order to have the kind of elite performance we have come to expect from Zambrano, he will need to return to inducing ground balls at will, or finally fix his one shortcoming, his walk rate.