I posted earlier today about Ubaldo Jimenez's uncanny combination of strikeouts and the ability to induce ground balls and prevent home runs. In fact he gave up just 11 homers last year in 198.2 IP despite playing his home games in Coors Field where the ball carries and doesn't break as much. Dalton commented on the post (click the link and scroll down) that Jimenez's excellent home run rate was luck, because only 6.9 percent of his fly balls left the yard, third lowest rate in the majors.

His basis for claiming the 6.9 percent HR/FB rate to be luck is that on average around 11 percent of fly balls leave the park, and that pitchers don't typically have control over whether a fly ball surrendered gets out. We believe this because from year to year, different pitchers will typically lead the league in luckiest and unluckiest in terms of HR/FB - unlike strikeouts where we can expect Jake Peavy and other power pitchers to be near the top consistently.

This is similar to BABIP for pitchers which typically hovers around 30 percent (on average), but can deviate unpredictably for all pitchers from year to year.

However, some players are freaks for whom the normal laws don't always apply. Tim Lincecum can throw 97 mph fastballs for lots of innings despite a small frame, and doesn't have to ice his arm. Carlos Zambrano's BABIP the last six years: .277/.275/.265/.258/.284/.291, i.e., his "luck" has always been above league average, sometimes well above it. And lest you think Zambrano had a 1 in 64 chance of getting lucky for six years in a row, (2^6), that's not the case. It would be 50/50 merely to be slightly under 30 percent for six years in a row. But to be between .258 and .277 for four years is far more unlikely than that. That Zambrano has some of the nastier pure stuff in the league with a lot of movement also argues in favor of this being a skill.

So what do we make of this? We've discovered these amazing tools to remove noise from players' stat lines, but certain players don't seem to fit. Fret not, the existence of outliers might disprove the rule, but the rule still works because most pitchers' ability to control whether balls-in-play fall in for hits is so minor that you might as well ignore it.

I'm fairly sure the same will be the case for HR/FB - most pitchers have so little control over that you can do your analysis as if it's pure luck. But I'd argue that some will have this skill, and those you have to treat differently. Jimenez's sample is pretty small, but 11 HR (8 in Coors, 3 on the road) in 198.2 IP, combined with his stuff makes me think that the 6.9 percent HR/FB is a skill. In other words, he had the third lowest HR/FB percentage in the league |STAR|without adjusting|STAR| for his home park.

It's tempting to think that because most pitchers don't show a measurable ability to prevent balls in play from being hits or fly balls from leaving the park, that it's the case for all of them. But just as blackjack is a game of luck where the house has a 1-2 percent edge over any normal player using perfect basic strategy, the house's long-term edge does not apply to card counters like this fool. And just as the House learned to take outliers like Jeff Ma to a back room and threaten them, we need to be on the lookout for pitchers like Jimenez and Zambrano whose skills slip through the cracks, and cheaply roster them.

Incidentally, while doing some research, I stumbled upon an interesting in depth breakdown of Jimenez's 2008 season at Beyond the Boxscore.