Using Pitch F/X Data to Track Rookie Regression: The Michael Pineda Experiment

facebooktwitterreddit

The recent trade involving Jesus Montero and Hector Noesi being dealt to Seattle for Michael Pineda and Jose Campos blew a lot of people out of the water. Granted, a lot of the strong reaction was mostly in part of what you see on the surface and not all of the factors involved from both sides. Bottom line is this: Seattle has a terrible offense, and they have real trouble attracting free agents to come there even when matching the market value. They have to overpay to get quality bats to go there, and if they decide to do that they’re just making the situation even worse. So instead of putting themselves in a bad position financially they look to trade for the pieces they need. There is a ton of information written about this deal, including an article by Keith Law on ESPN that pretty much mirror’s my opinion of it, so I decided to hold off on analyzing it.

What this trade did was push me to analyzing Michael Pineda a bit further using Pitch F/X data. There has been a ton of talk about how Pineda struggled through the second half of the season last year, with varying opinion’s on why that was the case. In fact if you have a moment, read Dave Cameron’s article on FanGraphs which digs a little deeper to explain the rise in Pineda’s ERA in the second half. My questions were these: What is the reason for this? Is he getting hit harder? Is he missing less bats? Why did his GB% rise, when his LD% really didn’t change from the norm? Is this learning on the job or is the league as a whole being fooled less?

Every year we see the new talented rookie come up and dazzle. Often times when they fail, we’re always wondering why? The real answer is way too complex to answer even though we would like to put some type of stat on it, these are human beings after all. But as the world of data at our hands increases, there are less and less questions that cannot be answered. This example uses an MLB-level pitcher, but the questions I were asking myself can be applied to any pitching prospect that has some success and then tails off. The fact of the matter is that tendencies are just what they are, and they can be used to not only judge a past performance but also to make sense of these results and try to figure out the future.

Pineda’s BABIP in the second half rose and consequently his ERA rose with it. This is normally a good stat to look at and say “oh, now I know why!” however his BABIP numbers are actually relatively low across the board. For the 2011 season, Pineda’s collective BABIP was .258, league average in 2011 for qualifying starters was .291, which counts for average defense behind and average ballpark factors. However, during the second half Pineda’s BABIP %’s were .294, .262, .275 which are all well below league average. His first half rates were completely unsustainable, so was his second half “regression” really that or was it the level that we can come to expect under normal circumstances? If that’s the case, how bad will it look when he suffers from higher than normal BABIP’s in the .300’s? You may see Yankee fans running him out of town, but it’s essentially not his fault.

Now before you get all excited and ask Seattle to take the trade back, look a little deeper. Pineda’s second half showed drastically higher GB%, while sustaining his LD%. Using the Pitch F/X tool, you can see that Pineda’s stuff never really dropped off. Below is an example of two individual games, but they serve as a good example of the data as a whole for 2011.

The general consensus after reviewing the data from 2011 is that his stuff never really regressed or lost anything. The only answer that I can hypothesize is that the league was making adjustments. But wait, wouldn’t that mean that his K-rate would drop if everyone is swinging and missing less and making contact more? Well, that would apparently be too easy! Pineda posted a 9.48 K/9 in the second half of 2011, trumping his 8.81 K/9 in the first half all with sustained LD%’s throughout!

Now with all of this said we can summarize that Pineda is maintaining his K-rate, inducing more ground balls but for some reason he stinks. I mean, he was just getting lucky in the first half with such low BABIP rates anyhow right? The second half he got what he deserved! Well, not so fast. As you can see, although

he posted a normalized BABIP in the second half, he was actually let down by his defense that made him look like a superhero in the first half. It’s quite possible Seattle’s defensive performance dropped in the second half, which in turn was a disaster for Pineda as he was learning to induce more ground balls as the league was making adjustments. Using the xFIP stat shows you that although his performance “suffered” due to fatigue and other elements everyone wrote about, he really wasn’t doing anything different other than getting more ground balls and improving as a pitcher.

Whew, lets take a breath! That’s a lot of information to go through just to say “Pineda’s really good”, but now hopefully some of you can answer some of the questions you may have when trying to figure out if the young stud on your home team is for real, or if he’s headed for AAA next season. Pineda will more than likely have a few frustrating moments in 2012 due to an aging defense behind him but we can be sure that this young flamethrower is good and getting better. Watch out AL East, Michael Pineda is here to stay.

You can follow us on Twitter @Seedlings2Stars and yours truly @MLBPerspectives. You can also keep up to date with all things S2S by liking our Facebook page.