Randy Wolf is a Better Pitcher than Zack Greinke | Disciples of Uecker

Disciples of Uecker

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

In 2009, Zack Greinke pitched one of the greatest seasons in the “live ball era” by a pitcher 25 or younger. Facing 915 batters, Greinke struck out more than 240 batters, walked only 51, and allowed 11 home runs. Allowing 64 runs in nearly 230 innings, Greinke’s value was off the charts — against his league and park, Greinke was approximately 55-60 runs better than the average pitcher.


Greinke’s 2009 campaign marked another milestone for the young pitcher. Working in front of an inefficient Kansas City Royals defense, Greinke allowed fewer hits on batted balls in play than expected. Given approximately 604 batted balls in play, Greinke allowed 184 hits; the Royals’ defense would be expected to allow 187 hits over the same number of batted balls in play.

Heading into 2009, Greinke was not yet the strike out machine that Brewers fans know and love. Joe Posnanski provided an exceptional biography of the Royals’ pitcher, and the personal challenges he faced (including a stint in which he nearly gave up professional baseball). Still, heading into his 2009 season, Greinke showed flashes of brilliance, and against a moderate home run rate, he boasted a particularly strong strike out-to-walk ratio.

The only trouble was, he pitched in front of a porous defense. During Greinke’s first season with the Royals, he allowed 117 hits on approximately 437 batted balls in play. The American League average in that 2004 season would suggest that 130 hits on batted balls in play would be an average rate. The Royals’ inefficient defense suggested that Greinke should allow 5 more hits on batted balls in play, so his mark of 117 hits allowed was rather remarkable.

18 hits better than his team defense would be as good as it got for Greinke. Throwing aside his extremely brief 2006 campaign, Greinke would allow 20 more hits than his expected team defensive support between 2005, 2007, and 2008. This is significant because the Royals’ defense itself was already 18 hits worse than the American League average over those years; granted, the Royals’ defense improved from 2005 through 2008, when they were basically average on batted balls in play. Greinke allowed approximately 38 more hits on batted balls in play than the league between 2005, 2007, and 2008.

Against this history, his slightly-better-than average performance against batted balls in play in 2009 seems remarkable. After all, Greinke spent his formative years pitching in front of poor defenses, and his overall performance suffered because of it. Greinke rates as an above average pitcher between 2004 and 2008, but there is a real sense that he was not as valuable as he should have been — despite markedly improving his strike out and home run rates after 2004, between 2005 and 2008, Greinke allowed approximately 15 more runs than one might expect (given his fielding independent performance).

RANDY WOLF LEADS THE PHILLIES
By age 27, Randy Wolf was consistently one of the best pitchers for the pre-dynasty Philadelphia Phillies. He anchored rotations featuring Vincente Padilla, a young Brett Myers, Robert Person, pre-sock Curt Schilling, and Kevin Millwood; I gather than Terry Francona and Larry Bowa prayed for rain more often than not. Mind you, this is not to sell Wolf short — he adroitly exceeded his fielding independent elements and turned in an overall performance that was above average between 1999 and 2004.


Wolf’s fortune was completely different than that of Greinke. Believe it or not, the turn-of-the-century Phillies played an efficient brand of defense, and the young Wolf learned his lessons under their guard. By the time his injury troubles hit, he was indeed an above average pitcher, and he also allowed notably fewer runs than one might expect. Wolf may not have known it then, but he was planting the seeds for his ability to limit the damage once his strike out rates were to fade late career.

During his first few seasons, the Phillies’ defense was 3 hits better than the National League average, prorated against Wolf’s batted balls in play. Wolf consistently allowed more hits than his Phillies’ defense suggested during this time, resulting in 14 more hits than expected between 1999-2001. Then, it clicked: the Phillies’ defense improved to 12 hits above average between 2002 and 2006, and Wolf more-than-doubled that support, allowing 26 fewer hits on batted balls in play than expected (during that time). Wolf produced one of his very best seasons during this stretch, preventing more than 20 runs against his league and park during his 2002 campaign.

PREVENTING HITS VS. FIP
Comparing regular MLB starters, there are few pitchers that yield average performances that are as different as Zack Greinke and Randy Wolf. Over the last couple of years, we can say that Greinke pitched better than his record actually shows, while Wolf seemingly confounds us with his ability to continually out-pitch his fielding independent elements.

WAR (Wins Above Replacement) drives our feelings about these pitchers. With 5.1 and 3.9 WAR between 2010 and 2011, Greinke looks a lot more valuable than his rotation-mate, as FanGraphs ranks Wolf at 0.8 and 1.4 WAR during 2010 and 2011.

Despite these WAR rankings, Wolf is arguably the more valuable pitcher to his ballclub over 2010-2011. For all his WAR, Greinke allowed 5 more runs than his league and park, as the only pitcher to work more than 200 innings for the hapless Royals. His injury-shortened 2011 campaign was 1 run worse than his league and park for the Brewers. Meanwhile, Wolf was the only member of the 2010 Brewers to pitch more than 200 innings, and aside from Yovani Gallardo, he was the only Brewers starter to turn in a campaign remotely close to average (2 runs worse than league/park). In 2011, Wolf was one of the Brewers’ most consistent starters, pitching 5 runs better than his league and park to strengthen the team’s middle rotation.

Unfortunately, WAR only considers pitchers based on their fielding independent elements. Over time, these elements might have a better chance at predicting future success — certainly, everyone consistently expects Greinke to outpitch Wolf. However, there are other elements to pitching that help players maintain their value.

In a classic response to Voros McCracken’s Defensive Independent Pitching shot heard ’round the world, Tom Tippett wrote a strong study that outlined pitchers’ abilities to prevent hits on batted balls in play. Nearly a decade later, the survey still rings true. Beyond their team defense alone, long term MLB pitchers indeed limited hits on batted balls in play. In some cases — pitchers such as Charlie Hough and Jamie Moyer, for instance — the ability to limit hits on batted balls in play is one of the main reasons pitchers stick around.

This is not to say that striking batters out and limiting walks and home runs is unimportant. Long term MLB pitchers do that, too. However, the suggestion about limiting hits on batted balls in play should resonate with our desire to explain the performance of pitchers such as Randy Wolf. Wolf doesn’t strike as many batters out as he used to, and yet limits the damage and manages to outpitch aces-in-training, such as Zack Greinke.

TRAINING THE BREWERS ROTATION
I worked through the Brewers’ rotation, after reading Tippett’s extremely logical and moderate conclusions, and I found something rather interesting. The Brewers’ pitchers who worked in front of good defenses early in their career became better at limiting hits throughout their career. Basically, this division pits Shaun Marcum and Wolf against Greinke, Yovani Gallardo, and Chris Narveson.

We should know, by now (especially after 2011), that defensive efficiency is neither distributed equally around the league, nor within each team. This is one of the main reasons that an individual pitcher’s actual performance can vary so greatly against his FIP during any particular season.


Even beyond this point, it is not necessarily the case that hits on batted balls in play are randomly distributed. If this were the case, it’d be difficult to explain how Zack Greinke, regarded as one of baseball’s better pitchers, consistently allows more hits than his league and team defense. Shaun Marcum throws a wrench in the other direction: pitching for some strong Blue Jays defenses, he was extremely successful in limiting hits while playing in the American League. While the Blue Jays’ defense prevented 15 hits on batted balls in play during Marcum’s time there, he went above and beyond, preventing 33 hits beyond his team defense.

Marcum and Wolf were the pitchers most successful with limiting hits on batted balls in play for the 2011 Brewers. Marcum allowed 16 fewer hits than the Brewers’ moderately efficient defense, and Wolf allowed 2 fewer hits. By comparison, both Gallardo and Narveson were basically even (both allowed 1 more hit than expected against their team defense), and Greinke allowed 12 more hits than expected.

One might simply write these off as one-year shifts, but that does not seem to be the case. Wolf was consistently better than the Phillies’ expected defense, after his first few years. During his transitional years, he remained better than his team defenses, while fluctuating between teams and parks. In Milwaukee, he was 24 hits better than the Brewers’ 2010 defense, and 2 hits better than the Brewers’ 2011 defense.

Marcum, 16 hits better than the Brewers’ defense in 2011, was 33 hits better than the Blue Jays during his time there. Narveson and Gallardo, both trained in front of below average to moderately efficient Brewers’ defenses, ultimately fluctuating a few hits here and there each year. Gallardo, during his time with the Brewers, is 5 hits better than his team defense, but 3 hits worse than the league; Narveson is 1 hit worse than the league, but 4 hits better than his Brewers defensive support.

Greinke, the odd man out when the Brewers were distributing their 2011 defensive shares, was 38 hits worse than his league in Kansas City (but he ended up even, overall, with his Royals’ defense).

How on earth can this be, if batted balls in play are simply randomly distributed, and defensive efficiency is randomly distributed?

PET THEORY TIME
This is not a conclusion, but a call for more research and investigation:

(a) Strike outs, walks, and home runs define the boundaries for pitching. They are extremely important elements, and one definitely wants to see pitchers strike out lots of batters, while limiting walks and home runs. Furthermore, there is conclusive research that pitchers that open their careers with strong strike out totals have longer careers, given that they have more room for error as their ability to strike batters out fades. On the other hand, pitchers that don’t strike out anyone have to do everything else very well, in order to limit the damage.

(b) Beyond striking out batters, and limiting walks and home runs, pitchers have some ability to limit hits on batted balls in play. Do pitchers that learn to pitch in front of efficient defenses learn how to limit hits? Is there something to working in front of a good defense that teaches pitchers how to effectively use their defensive support?

That is, if we compare Marcum and Wolf to Greinke, can we explain the differences in their 2011 pitching performances to the fact that Greinke learned how to pitch in front of inefficient defenses, while Marcum and Wolf spent their formative years working in front of defenses that were typically average or better?

The reason I ask this is, even when Greinke received defensive support that was near the league average, in Kansas City and Milwaukee, he typically allowed more hits than expected. In 2007, he was 3 hits worse than a team defense that was 1 hit worse than league average; 6 hits worse than a basically average defense in 2008; 12 hits worse than an average defense in 2011. By comparison, after 1999-2001 campaigns in which he was 14 hits worse than basically average team defenses, Randy Wolf was even with average defenses in 2004 and 2006, 14 hits worse between 2007-2008, and 5 hits better in 2011. Marcum’s performance in front of average defenses jumped to 13 and 16 hits prevented in 2010 and 2011, after allowing 4 more hits than his defense during his brief 2006 campaign.

Oddly enough, after an early career full of efficient seasons in front of efficient defenses, Randy Wolf’s best season prevented hits was his 2010 campaign in front of a below average defense. Furthermore, each time he received notably above average defensive support, he capitalized on that support (minus his brief 2005 campaign).

Zack Greinke has not once pitched in front of a notably above average defense. From his formative years, beyond, the best defensive support he received was from an average 2008 Royals defense, and the basically average 2011 Brewers defense. I’ll ask again: What effect does pitching in front of bad defenses early in a career have on a pitcher’s ability to limit hits later in his career?

(c) I believe this is something that we can study, but will be difficult to prove. We can place each team’s defense in the context of its league and park, and we can place each pitcher within that context, too. We can analyze the types of batted balls in play they allow, we can analyze their fielding independent elements, and we can analyze how those elements interact to build a full pitching profile.

Logic / Rationality Sidenote: I believe that there are mystical or unspoken elements within rationality. No matter how far we can push logic, no matter how much we can carry identity through strong laws (such as principles regarding internal contradiction of logical elements, or principles of sufficient reason), we can investigate the boundaries or limits of reason in each logical element that we examine. By interacting with logical elements and material elements alike, our soul can double back and trace its psychological and spiritual development (climbing the incline of physics to metaphysics, as Henri Bergson might say). These elements might be completely unspoken or unquantifiable, but we can learn that they exist by fully investigating each rational, logical, and material element that we encounter.

That is, we might know how to fully maneuver in a world that opposes [a/~a], and by investigating that, we might move to oppose [a/b] or [b/c], ascribing positive identities to fill in the “not-a” that opposes “a.” Furthermore, I believe tracing these examinations, we can find alterity alongside those very logical elements, eventually understanding that there are real things that are outside [a/~a], or even real, unspeakable or unquantifiable elements ([    ] / [a/~a]). That doesn’t make what we know any less real; rather, it allows us to see the boundaries of what is real that we can know.

CONCLUSION
So it goes with baseball stats. We have a complete record of pitching elements in front of us. We can stop at fielding independent elements, and forever proclaim pitchers such as Zack Greinke to be eternal victims of bad luck. We can scoff at unexpected performances by Randy Wolf as just another lucky season — some might try and convince themselves, “he’ll regress, he will, just you wait” (don’t worry, I used to think this, too). Yet, beyond WAR, beyond those fielding independent elements, we can work with the progression of each pitcher’s career, and learn about the circumstances in which they learned to pitch.

You’ll have a hell of a time proving that it exists, and so will I, but something is there that allows Randy Wolf to consistently limit the damage and work with his pitching abilities. Furthermore, something is there that keeps Zack Greinke from preventing hits on batted balls in play, even when he pitches in front of average defenses.

My hypothesis is that pitchers can learn how to pitch given the contexts in which they are required to work. Does this not explain Greinke’s vocal love for FIP? Wouldn’t you strive after every element you could possibly control if you worked in front of generally poor defenses? Wouldn’t that also affect your pitching approach? Same for Randy Wolf: if you knew that you wouldn’t be able to strike out a lot of batters, wouldn’t you learn to pitch to limit the damage in every way possible?

I can’t quantify this yet. I am sure that defensive luck has something to do with it, but I think it goes beyond that. Now, that’s a call to investigate pitching records within defensive contexts.

RESOURCES:
Baseball-Reference. Sports-Reference, LLC, 2000-2012.
“FIP Glossary.” FanGraphs.

James, Bill. The New Historical Baseball Abstract. New York:  Free Press, 2001.
Posnanski, Joe. “Zack Greinke is in Total Control.” Sports Illustrated, 4 May 2009. Retrieved from http://sportsillustrated.cnn.com/2009/writers/joe_posnanski/04/28/zack.greinke/index.html.
Tippett, Tom. “Can pitchers prevent hits on balls in play?” 21 July 2003. Retrieved from: http://207.56.97.150/articles/ipavg2.htm.

Images:
Greinke: http://articles.businessinsider.com/2010-12-19/sports/30076430_1_milwaukee-brewers-top-stars-trade
Wolf: http://bleacherreport.com/articles/749031-philadelphia-phillies-the-20-most-beloved-players-in-team-history
Marcum: http://mopupduty.com/would-marcum-even-get-drafted-in-2010/

 

STATS LOG

Zack Greinke:
2004 (437 BIP): 117 H / 135 team / 130 league
2005 (622 BIP): 210 H / 199 team / 182 league
2006 (19 BIP): 6 H / 6 team / 6 league
2007 (442 BIP): 110 H / 107 team / 106 league
2008 (583 BIP): 181 H / 175 team / 175 league
2009 (604 BIP): 184 H / 187 team / 181 league
2010 (651 BIP): 201 H / 200 team / 191 league
Even with team, 38 hits worse than league (team 38 hits worse than league)
2011(445 BIP): 142 H / 130 team / 132 league
12 hits worse than team, 10 hits worse than league (team 2 hits better than league)
Overall: 48 hits worse than league, 12 hits worse than team

Yovani Gallardo:
2007 (315 BIP): 95 H / 98 team / 94 league
2008 (65 BIP): 19 H / 19 team / 20 league
2009 (466 BIP): 129 H / 139 team / 139 league
2010 (508 BIP): 166 H / 159 team / 152 league
2011 (564 BIP): 166 H / 165 team / 167 league
3 hits worse than league, 5 hits better than team (team 8 hits worse than league)

Chris Narveson:
2006 (21 BIP): 5 H / 6 team / 6 league
1 hit better than team and league (team even with league)
2009 (131 BIP): 38 H / 39 team / 39 league
2010 (494 BIP): 151 H / 155 team / 148 league
2011 (486 BIP): 143 H / 142 team / 144 league
1 hit worse than league, 4 hits better than team (team 5 hits worse than league)

Randy Wolf:
1999 (343 BIP): 106 H / 101 team / 103 league
2000 (605 BIP): 185 H / 180 team / 180 league
2001 (449 BIP): 135 H / 131 team / 132 league
14 hits worse than team, 11 hits worse than league (team 3 hits better than league)
2002 (584 BIP): 149 H / 167 team / 171 league
2003 (558 BIP): 149 H / 161 team / 165 league
2004 (429 BIP): 125 H / 125 team / 127 league
2005 (238 BIP): 73 H / 68 team / 71 league
2006 (166 BIP): 50 H / 51 team / 50 league
26 hits better than team, 38 hits better than league (team 12 hits better than league)
2007 (304 BIP): 100 H / 93 team / 92 league
2008 (346 BIP): 109 H / 103 team / 104 league
2008 (207 BIP): 61 H / 61 team / 62 league
2009 (612 BIP): 154 H / 171 team / 182 league
4 hits better than team, 6 hits better than league (team 12 hits better than league)
2010 (663 BIP): 184 H / 208 team / 199 league
2011 (661 BIP): 191 H / 193 team / 196 league
26 hits better than team, 20 hits better than league (team 6 hits worse than league)

Shaun Marcum
2005 (24 BIP): 6 H / 7 team / 7 league
2006 (234 BIP): 73 H / 69 team / 71 league
2007 (454 BIP): 122 H / 128 team / 138 league
2008 (425 BIP): 105 H / 122 team / 128 league
2010 (568 BIP): 157 H / 170 team / 167 league
33 fewer hits than team, 48 hits fewer than league (team 15 hits better than league)
2011 (580 BIP): 153 H / 169 team / 172 league
16 fewer hits than team, 19 fewer hits than league (team 3 hits better than league)
55 hits better than league, 49 hits better than team (team 18 hits better than league)

Share Our Posts

Share this post through social bookmarks.

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

Comments

Tell us what do you think.

  1. Bradley says: March 29, 2012

    What methodology did you use to determine what made a defense (relatively) good or bad?

    • Nicholas Zettel says: March 29, 2012

      To determine hits against batted balls, I simply used BABIP. It’s extremely limited, as defensive inefficiencies can occur in ways that don’t result in hits.

      For 2011 NL, defensive efficiency accounts for approximately 184 times on base per 600 PA; BABIP accounts for approximately 178 hits per 600 PA (this is also consistent with the plain error rate of approximately 1% per PA).

  2. Matt Tracy says: March 29, 2012

    Interesting article. As you may have been told before writing this, there is an extreme challege to write about math and logic. One thing i wish you have done is not to vacillate between hits allowed and hits prevented. Often, when the reading got tough, you arbitrarily switched between which measurement you were using. I think it would have been easier to follow if you had simply used negative integers of the same measurement (either all “allowed” or all “prevented” and let the measurements be plus or minus).

    Very interesting though. that is worth repeating. And I can’t help but wonder if your missing “not-logic-based” measurement/element that you are looking for is some form of “baseball Heisenberg Uncertainty”, where the pitcher learns what he learns about the play simply by watching it unfold. Unfortunately, to ever prove that, we’d need to have pitchers that never watch the defense behind them. That’s not going to happen.

    • Nicholas Zettel says: March 29, 2012

      Thanks for the comments, Matt. I especially appreciate the feedback on the writing style. I am constantly working on my presentation, and I tried to write this more as a narrative, rather than a presentation of numbers. Sometimes I lean too far in the other direction, too, and simply write the numbers without the narrative.

      I agree with your last paragraph. That’s a very strong observation.

      • Matt Tracy says: March 29, 2012

        Hey man, you did better than I think anyone short of Stephen Hawking or Neil deGrasse Tyson cum fangraphs could have done.

  3. Matt says: April 2, 2012

    BABIP has some considerable limitations as a judge of luck vs. skill in a pitcher’s results. There needs to be a quality study beyond FB% and LD% and GB% as none of these determines the quality of the FB or GB. A more accurate measure would be velocity of the batted ball as it passes through the infield or even time it takes the ball to leave the infield. I understand that this is a very difficult number to measure in every at bat, but it would give a better estimation of whether a batter is able to make solid contact against a pitcher. I would be willing to bet that Greinke’s 2009 season could be explained by a more than average number of balls that were hit slowly either on the ground or in the air.

    • Nicholas Zettel says: April 2, 2012

      I think that’s right on.

      On a related note, once those fielding-related f/x stats are available, including batspeed and other measurements around the diamond, it should blow everything else out of the water.

  4. Zack Greinke says: May 17, 2012

    How’s this analysis looking now, Nick?

    • Nicholas Zettel says: May 17, 2012

      I was wondering how long it’d take for someone to call me on this.

      Thanks for reading, Zack!

Trackbacks

Websites mentioned my entry.

  1. Resource: Fun with FIP, 2011-2013 | Disciples of Uecker

Add a Comment

Fill in the form and submit.