What’s in a projection? | Disciples of Uecker

Disciples of Uecker

We'd like to go to the Playoffs, that would be cool.

With the Brewers’ 2013 ZiPS projections available, I gather that scores of Brewers fans already concerned about the starting rotation suddenly became worried about the club’s excellent offense. As our own Ryan Topp noted with side-to-side comparisons, ZiPS’ conservative projections result in declining offensive production from the 2013 Brewers. Beyond the typical “slash” numbers (AVG/OBP/SLG) — inevitably the first place most fans look when reading projections — there are some unique trends in the ZiPS projections that pertain to the contact and discipline approaches of Brewers batters.

Rather than analyze how the various “slash” lines from the projections affect the club, I would like to look at basic shifts in stats that deal with the rate that batters knock the ball in play (or, strike out, walk, or homer). These rates will help to define the approaches of our batters, and rather than get into an argument about what OBP so and so will actually have, we can analyze the aspects of batting that will eventually influence those final slash numbers.

Now that ZiPS is published at FanGraphs, here is an explanation comparing ZiPS to other projection systems. Furthermore, I highly recommend reading methodological explanations at the bottom of the Brewers’ projections. My goal is not to criticize the method or results of ZiPS, but to show how the 2013 Brewers projections differ from previous playing record in some ways.

Ryan Braun’s Strike Out Rate
As I noted in my analysis of the Brewers’ batting approaches last week, as Ryan Braun‘s surface stats remained rather consistent over his first few years, his approach shifted wildly toward a contact/power player. Shifting to a contact hitter means that his strike out rate fell, and his walk and home run rates remained low enough that he produced strong power performances while batting the ball in play at (or more than) the typical league average of 68% to 70%.

In 2012, Braun’s strike out rate skyrocketed from 14.8% to 18.9%, shifting his approach away from his previous contact hitting developments. What I found interesting about the ZiPS projection is that Braun’s strike out rate stands at 16.5% — so, not as high as his early career strike out rates, but not as low as his shift to the contact-hitting Braun we know and love. Right in between those approaches for Braun.

This suspends one of the main questions about Braun for 2013 — not whether he will produce, but how he will approach his plate appearances. Will his sudden 2012 strike out increase continue, or will he return to the path of 2010 and 2011 contact rates?

Aramis Ramirez’s Contact Shuffle
Oddly enough, ZiPS projects Aramis Ramirez‘s basic contact approach to be exactly the same as his 2012 approach. In 2012, Ramirez’s HR, BB, and K totaled 24.3% of his plate appearances; in 2013, ZiPS projects the same. However, in this case, Ramirez featured very slightly fewer home runs, slightly fewer walks, and slightly more home runs. Here’s a good exercise about how a few ticks one way or the other can influence batting production:

A: 550 PA, 74 K/36 BB/23 HR
B: 550 PA, 72 K/39 BB/24 HR

These totals look similar enough; in fact, to argue that they’re significantly different would be nitpicking. Yet, if we assume a typical BABIP (say, .300 for this purpose), and present Ramirez with his typical percentage of doubles on hits in play, we get these different lines:

A: 550 PA, 74 K/36 BB/23 HR; 514 AB, 148 H, 33 2B, .288/.335/.486
B: 550 PA, 72 K/39 BB/24 HR; 511 AB, 149 H, 33 2B, .292/.342/.497

Again, just a few more times on base, just one more home run, and yet, we’d be inclined to argue that “Player B” is preferable to “Player A.” Sometimes analyzing ballplayers is splitting hairs, and even when two players have similar batted-ball-in-play rates, their distribution of home runs and walks can notably influence their overall production (even when the apparent difference is minute).

Jonathan Lucroy’s Walk Rate
Over the course of his career, Jonathan Lucroy fluctuated between a strong contact bat and a high strike out/moderate power bat. In 2012, his great improvement included more line drives and flyballs as a reward for fewer strike outs and better overall discipline. Lucroy batted the ball in play more frequently, and also hit with more power.

Lucroy’s ZiPS performance is not unlike Braun’s in terms of strike out rate; given two extremes, ZiPS takes the mean. Lucroy also keeps his moderate power. However, even with more strike outs in his 2013 projection, ZiPS presented Lucroy with a significantly higher walk rate. Whereas Lucroy’s previous walk rates inched forward from 6.1% to 6.4%, ZiPS sees our franchise catcher jump to a 7.0% walk rate in 2013. In this regard, Lucroy’s discipline advances in 2012 translate into more walks, rather than more contact. It will certainly be interesting to see if Lucroy maintains his contact approach into 2013.

Norichika Aoki’s Contact
The Brewers gained a surprising leadoff bat in Japanese veteran / MLB rookie Norichika Aoki. Aoki’s mid-to-late season power surge was especially promising, as he showed the ability to adjust to new pitching approaches, punishing pitchers with extra base hits when they challenged him to not slap the pitch into play. Even with his power, Aoki was a true, extreme contact hitter — including 13 HBP, Aoki knocked the ball into the field of play in more than 79% of his plate appearances.

In 2013, ZiPS projected Aoki to become an even more extreme contact hitter. Not only does Aoki’s projection show his home run rate dropping from 1.7% to 1.1%, but his strike out rate drops to 9.1%, and his walk rate drops from 7.3% to 6.2%. These significant shifts mean that even if Aoki maintains his strong HBP total, he will bat the ball into play 81.4% of the time in 2013. While Aoki certainly has the basic contact profile to become a more extreme contact hitter, it will be interesting to see how his contact approach advances.

Carlos Gomez’s Home Run Rate
Perhaps there is no better test for a projection system than a player like Carlos Gomez. Gomez maintained a similar discipline approach over the last three seasons, although that lack of discipline did not keep him from changing the type of contact he made. As is well-documented over the last season, Gomez shifted his contact from ground balls and bunts to fly balls (and line drives) over the last three years. On the surface, we might see the same “below average bat/all glove value” Gomez, but in fact, his power has increased to a level that makes other offensive shortcomings acceptable.

Can anyone blame ZiPS for not projecting Gomez’s home run rate to continue to improve? Certainly, it would be intriguing to say that his jumps from 20.1% flyballs to 26.7% flyballs to 29.0% flyballs from 2010 to 2012 sustains his home run increases of 1.6% HR to 3.1% HR to 4.2% HR. But, does the correlation of significantly more flyballs to significantly more home runs mean that we should continue to project that trend forward? Imagine Gomez knocking homers in 4.5% of his plate appearances; even if he struck out 25% of the time and walked 5% of the time (those are off the top of my head extremes), he might produce the following batting line over a full season:

560 PA, 140 K / 28 BB / 25 HR / 7 HBP; 360 BIP (Balls-in-play), 107 BIPH (Balls-in-play Hits); 525 AB, 132 H, 23 2B, 5 3B, 25 HR; .251/.298/.457

Notably, that AVG, OBP, and SLG are all worse than Gomez’s 2012 campaign, and in some ways, it’s not fair to even place extreme K% or BB% that I chose (that’s really just picking and choosing from the most extreme rates of his 2010-2012 seasons). At 21.7% strike outs and 5.8% walks?

560 PA, 122 K / 32 BB / 25 HR / 7 HBP; 374 BIP, 111 BIPH; 521 AB, 136 H, 24 2B, 5 3B, 25 HR; .261/.313/.470

There’s your career year (contract-season?) Gomez. One way or the other, from a poorly disciplined Gomez with a continuation of his power surge to a better-disciplined Gomez with a continuation of his power surge, can we hold it against a projection system for not continuing those trends?

SOURCES:
Baseball-Reference. Sports Reference, LLC. 2000-2013.
FanGraphs / ZiPS projections.

IMAGE: http://larrybrownsports.com/baseball/carlos-gomez-pimps-hr-down-15-0/18798/attachment/carlos-gomez-brewers

Share Our Posts

Share this post through social bookmarks.

  • Delicious
  • Digg
  • Newsvine
  • RSS
  • StumbleUpon
  • Technorati

Comments

Tell us what do you think.

  1. Jon R White says: January 16, 2013

    I have no problem reading the analysis of players, but to trianglulate stats to the nth degree in an attempt to get MORE out of Braun seems to be counterproductive at best.

    • Nicholas Zettel says: January 16, 2013

      I don’t think it’s for the purpose of getting more out of Braun, but rather, it’s for the purpose of understanding how Braun approaches the game at the plate.

      If he bats the ball into play more consistently, rather than striking out frequently, that will change the spectrum of results for his batting production. Now, he’s consistently an above average or elite bat, so the results usually look the same on the surface, but those contact trends could show how his approach will eventually shift (for example, we might prefer that Braun strike out, say, 14% of the time, rather than 18 or 19%; that could make it easier for him to maintain that strong batting average along with his HR power).

Trackbacks

Websites mentioned my entry.

  1. Daybreak Doppler: What Now – The Wide Receiver Edition | PocketDoppler.com

Add a Comment

Fill in the form and submit.