Where Luck Ends and Skill Begins
- By: Chris Liss
- On: 4/9/2009 12:23:00 AM
- View Comments : 7
Related: Chris Liss Chicago Cubs Colorado Rockies MLB
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 *without adjusting* 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.
reddit
Comments
On: 4/9/2009 9:36:00 AM
On: 4/9/2009 11:33:00 AM
Now if he was an older pitcher with a track record and saw that ratio suddenly jump out of nowhere I would probably say he will regress. However, given Jimenez's skill progression through the minors and into last year, I think his 2008 performance is probably a function of his evolution as a pitcher, and a direct result of his core skills, and thus a repeatable skill not an outlier from a temporary and unsustainable run of luck.
The same cannot be said of sinkerballer Aaron Cook though. That's the prime example of a pitcher who's results could regress.
Cook's infield FB% in 2008 was 12.9%.
in 2007 7.4
in 2006 5.6
Cooks's FB/HR% in 2007 11%
in 2006 9.6%
in 2008 7.6%
On: 4/9/2009 11:56:00 AM
In any event, it's also intuitively obvious that hitting a 98 mph sinking/moving fastball into the air with authority is not easy to do, as opposed to an 90 mph fastball that relies more on location than movement.
On: 4/9/2009 12:20:00 PM
On: 4/9/2009 1:58:00 PM
On: 4/9/2009 2:44:00 PM
On: 4/9/2009 7:52:00 PM
As far as predicting the the skill of pitchers based on GB/FB and FB/HR % I think my point is essentially that simply skimming the numbers won't give you a real answer. Let's take the case of Derrick Lowe and Jimenez. The difference isn't just the fact that they induce GB but how they accomplish it. Lowe throws a FB that rarely tops 90 mph with a late diving action that makes most hitters swing over the top. He uses a CH with the same arm action and the same break but about 6-8 mph slower. He then uses his nasty slider as a wipeout pitch. What Lowe doesn't do is generate enough velocity to overpower hitters with his FB, and when he's not on his game or is tired his FB loses it's break and flattens out. Although some more may have to do with this http://baseballevolution.com/keith/gbpitchers.html.
With Jimenez you have a FB that comes in at 95 mph with movement (not necessarily sinking). His GBs aren't a function of the late downward path of the ball, which requires pin point accuracy, but the overwhelming inertia generated by the velocity and spin of the ball. In tennis this is referred to as a heavy topspin and if you've ever tried to hit one it makes your wrists feel like they are going to collapse because to return the ball you have to overcome the generated velocity of the ball and it's heavy rotation. This is the "heavy" ball that Rollins was referring to (heavy=rotation+velocity). With Lowe the ball needs to stay down in the zone or you can get under it and drive it. With Jimenez he can throw the ball all over the place and you'll still pound it into the ground. That's a repeatable skill.
As far as picking up GB pitchers bceause the don't give up homers... well I prefer the pitchers who don't even let the batter get the chance.
Leave a comment
Commenting is restricted to registered users only. Please register or login now to submit a comment.