Category: Sabremetrics

Did Chemistry Mix Up the Red Sox and Rays? – Part One

With the release of the movie “Moneyball” and the Oakland Athletics’ struggles in recent years, new “champions” for statistical analysis on how baseball teams should be run have emerged in the Boston Red Sox and the Tampa Bay Rays. Statisically, the Red Sox (as well as the AL East Champion New York Yankees) were favored, though not heavily, over the Tampa Bay Rays in the preseason. In addition, in terms of Pythagorean Winning Percentage which Red Sox analyst Bill James popularized, the Red Sox performed better in 2011 than the Rays, having outscored their opponents more than the Rays did (138 to 98).

Yet, one of the neat features of baseball is that whenever you think you’ve seen everything in baseball and analyzed that “everything” down to its minutae, there’s always something new event to surprise you. The epic, historic, traumatic, dramatic (and every other adjective you deem appropriate based on your fan affiliation) collapse of the Boston Red Sox in the American League Wild Card race to the Tampa Bay Rays concluded a month of flailing and failing for the Red Sox faithful. Statistically, it was virtually impossible. Turns out, “virtually” does not equal “actually”. As a kicker, the turnaround also makes for a neat looking graph too.

Consider in the course of the Red Sox collapse, the Red Sox “featured” a 7-20 record in September, the Rays pulled off some magic of their own. On September 6th, the Rays had a 0.6% chance of making the playoffs as the Red Sox held an eight game lead in the wild card. As the Red Sox slumped in september, the Rays went 17-10 in September, even managing to turn a triple play during the hot streak. Then on September 28th, the Rays clinched the Wild Card by overcoming a 7-0 deficit in their game with the Yankees as the Red Sox lost to the Orioles.

As Jack Moore of FanGraphs shows (and writes about in minute-by-minute detail), just within that last day, the playoff odds swung back and forth not just once, but twice like a fall breeze. Then the leaves settled and the Rays were in the playoffs.

The Red Sox, despite the fifth-best record in the American League, found themselves shaken. Long-time manager Terry Francona and the Red Sox parted ways (or politely ousted) as the Red Sox ownership and media tried to come to terms with what had happened. Some, such as Gordon Edes from ESPN, suggest that Francona had lost control of his clubhouse and that a “new” voice was needed. It appears the players were more interested in arguing over who could drink beer during ballgames than staying prepared for games. Other players, such as ex-Ray Carl Crawford, were unproductive and in some aspects, symbolic of the Red Sox collapse as the Orioles Robert Andino’s game-winning hit slipped out of Crawford’s glove to close out the Red Sox season. David Ortiz, seen as a leader (and often a mouthpiece to the media about all Bostonian baseball things) felt that “there were certain things in the clubhouse that no one can control”. His only feeling towards Francona besides being “fine with Tito” was reminiscing on when Francona had benched him in 2010. With David Ortiz’s voice catching most of the media attention, ownership’s voice and players shocked and saddened after the fact, Edes wondered if “an inmates-running-the-asylum environment” caused things to spiral out-of-hand.

The playoff-bound Rays however are now rejoicing. Before the 2011 season, longtime Rays Carl Crawford, Carlos Pena and Matt Garza had left the team and the Rays had 25 new players report to their spring training camp. Though the Rays were defending their 2010 AL East championship and had, in the eyes of executive vice president Andrew Friedman “As much physical talent in this camp as he ever had”, and an effort was made to build team chemistry. With some jokes sprinkled in here and there and members of the front office joining manager Joe Maddon on the field, a “light mood in the clubhouse with constant team-building moments” was encouraged. In the words of ex-Red Sox turned Ray Johnny Damon “We’re here to cause no trouble — we’re not here to do the Super Bowl Shuffle, but we’re here to cause no trouble. But it’s good everyone has the same common goal.” Not everything worked out for the new Rays, however. Not everything worked out for the Rays, even among the Red Sox imports. Five games into the Rays season, ex-Red Sox Manny Ramirez abruptly retired before he was suspended for 100 games for failing a second drug test. Yet, the Rays kept on producing. Except for a poor 11-15 record in July, the Rays played solidly all year coupled with a scorching 18-10 August and their blazing 17-10 September that enabled them to overtake the Red Sox.

Was it talent, luck, or chemistry that made the difference between the Red Sox and Rays?

Stay tuned for Part 2 as we try to untangle (and maybe learn) from this conundrum a bit.

Advertisements

Reframing Gregg Zaun

In my Baseball Chirps on Some Links (which is my fancy way of updating “Bird on a Wire”), I touched on Mike Fast’s study at Baseball Prospectus on a catcher’s ability to frame pitches and the way those framed pitches tended to switch close pitches from balls to strikes. One thing that stuck out to me was catcher Gregg Zaun’s name near the top of that list.

One interesting quirk of sabremetrics (which can lead to a lot of fun debate) is how new studies can lead to reevaluations of existing players. This concept was illustrated, in part, in the book and movie Moneyball. According to Oakland Athletics General Manager Billy Beane, Scott Hatteberg stuck out because his high on-base percentage (OBP) skills were undervalued by the rest of baseball. Thus, Beane brought Hatteberg to offset departing MVP Jason Giambi’s production for pennies on the dollar. Now, I am not a sabremetrician but I do love thinking about baseball in new ways. Thus, I decided to, for lack of a better pun, reframe my thinking about Gregg Zaun.

Now, Zaun was a bit of an oddity as a switch-hitting catcher without much power and a “small” 5’10” 170 lb frame. Most of the early part of his career was spent as a backup catcher to the likes of Chris Hoiles (who had the reputation as a good hitter) and Charles Johnson (who had a reputation as a great defender). A few years ago, he picked up the label as “The Practically Perfect Backup Catcher and later, as David Ross of FanGraphs notes, “However, as people looked more closely. they began to realize that a catcher who was close to league average offensively (Zaun has a career 94 wRC+) and non-horrible defensively would actually make a Pretty Good Starting Catcher. The Toronto Blue Jays noticed and were the only team to really give Zaun a full season of playing time in 2005.”

Mike Fast’s study indicates that since 2007, Zaun saved the sixth most runs of any catcher in baseball based on his ability to frame pitches, equivalent to 36 total runs. What makes it more interesting is that Gregg Zaun also did not play in 2011 and had durability issues from 2007 to 2010. Fast calculated that Gregg Zaun saved 19 runs per 120 games. Based on my rough, back-of-the-envelope calculations, it is possible that his framing skills were worth an extra win a season above an average major league catcher of that time period. To put that into scale a bit, an “average” major league catcher (as opposed to a Triple-A replacement level player) is generally worth three wins without counting runs saved from framing pitches.

So, we have Gregg Zaun, a person who was at first considered a backup catcher, then upon further analysis, would be considered a “Pretty Good Starting Catcher”. Does this new analysis on his catcher framing skills push Gregg Zaun into elite territory? I do not think so. Does it make him a star? Perhaps not. We do not know if he exhibited these framing skills early in his career. However, it does reinforce the idea that a potential opportunity was missed to see if Zaun would have been a “Pretty Good Starting Catcher”.

More importantly, if the results of Fast’s study are vetted, this could become a tool used by Major League Teams to analyze a catcher’s defensive performance. If that happens, I strongly believe that a future rookie that has Gregg Zaun’s skillset will no longer be undervalued but will stick out and thus warrant a full-time role to show that they really are a “Pretty Good Starting Catcher”.

Gregg Zaun’s Minor and Major League Statistics

Year Age Tm G AB HR RBI BB SO BA OBP SLG OPS OPS+
1990 19 BAL-min 98 284 3 28 30 32 .239 .315 .310 .625  
1991 20 BAL-min 113 409 4 51 50 41 .274 .353 .369 .722  
1992 21 BAL-min 108 383 6 52 42 45 .251 .324 .376 .700  
1993 22 BAL-min 100 336 4 49 33 37 .295 .357 .384 .740  
1994 23 BAL-min 123 388 7 43 56 72 .237 .337 .353 .690  
1995 24 BAL-min 42 140 6 18 14 21 .293 .367 .529 .896  
1995 24 BAL 40 104 3 14 16 14 .260 .358 .394 .753 95
1996 25 BAL-min 14 47 0 4 11 6 .319 .441 .362 .802  
1996 25 TOT 60 139 2 15 14 20 .245 .318 .367 .685 75
1996 25 BAL 50 108 1 13 11 15 .231 .309 .352 .661 67
1996 25 FLA 10 31 1 2 3 5 .290 .353 .419 .772 106
1997 26 FLA 58 143 2 20 26 18 .301 .415 .441 .856 130
1998 27 FLA 106 298 5 29 35 52 .188 .274 .292 .566 53
1999 28 TEX 43 93 1 12 10 7 .247 .314 .323 .637 60
2000 29 KCR-min 9 25 0 3 4 3 .280 .379 .400 .779  
2000 29 KCR 83 234 7 33 43 34 .274 .390 .410 .800 102
2001 30 KCR-min 17 61 1 9 10 8 .213 .329 .328 .657  
2001 30 KCR 39 125 6 18 12 16 .320 .377 .536 .913 131
2002 31 HOU 76 185 3 24 12 36 .222 .275 .319 .594 53
2003 32 TOT 74 166 4 21 19 21 .229 .309 .349 .658 67
2003 32 HOU 59 120 1 13 14 14 .217 .299 .300 .599 56
2003 32 COL 15 46 3 8 5 7 .261 .333 .478 .812 97
2004 33 TOR-min 7 23 0 2 2 5 .304 .346 .348 .694  
2004 33 TOR 107 338 6 36 47 61 .269 .367 .393 .761 96
2005 34 TOR-min 2 6 0 0 2 2 .333 .500 .500 1.000  
2005 34 TOR 133 434 11 61 73 70 .251 .355 .373 .729 94
2006 35 TOR-min 1 4 0 0 0 1 .000 .000 .000 .000  
2006 35 TOR 99 290 12 40 41 42 .272 .363 .462 .825 112
2007 36 TOR-min 3 11 0 0 1 2 .091 .167 .091 .258  
2007 36 TOR 110 331 10 52 51 55 .242 .341 .411 .752 98
2008 37 TOR-min 2 8 1 1 0 2 .250 .250 .625 .875  
2008 37 TOR 86 245 6 30 38 38 .237 .340 .359 .700 88
2009 38 TOT 90 262 8 27 31 48 .260 .345 .416 .761 99
2009 38 BAL 56 168 4 13 27 30 .244 .355 .375 .730 92
2009 38 TBR 34 94 4 14 4 18 .287 .323 .489 .813 112
2010 39 MIL 28 102 2 14 11 12 .265 .350 .392 .743 101
16 Seasons 1232 3489 88 446 479 544 .252 .344 .388 .732 91
162 Game Avg. 162 459 12 59 63 72 .252 .344 .388 .732 91
Provided by Baseball-Reference.com: View Original Table
Generated 9/27/2011.

Simply Enjoying Sabremetrics

If you’ve followed baseball a bit over the last ten or twenty years, you might’ve heard a term bandied (and in some cases, ridiculed) called sabremetrics. You might even hear more about the term creep into the media since Michael Lewis’s book “Moneyball” is scheduled to hit the big screen on September 23rd starring Brad Pitt. Yet you might still wonder what it is, why people care, and how much of it is “much ado about nothing”.

Sabremetrics is, to put it simply, the study of the measurable elements of the game of baseball. In other words, you got a group of people who really love baseball and for fun (and in recent years, as a career), dive as deep as they can into the bits and pieces of the game to figure part of it out. Just as there are people who try to predict the stock market or to figure out how to build a better mouse trap and what parts of a mouse trap are better than others, sabremetricians try to find concrete evidence in baseball. Now, that isn’t a really easy thing to do. Sure, baseball as elements of the hard sciences such as physics, geometry and mathematics. Yet, it’s also a game played by people and the best player in the world can strike out four times in a row or the best pitcher can have problems finding the strike zone with a GPS and a trebuchet on a given day. Then, think of all the things not in a typical physics equation that can fiddle with a baseball thrown to the plate like how the wind speed and humidity can flatten a curveball in Colorado and sabremetricians got a lot of noise that they have to wade through. And that’s even before “lucK” tosses a wrench in the equation!

So what are some of the concepts that sabremetricians use? Well, since sabremetricians try to study what they can measure (with some reliability), what they do is break the events in baseball down to a series of events that have a chance of leading to a subsequent set of events. As an example, sabremetricians try to calculate if a player gets on first base, what would be their chance to score? How does their chance to score change with one out in the inning? What if there is one out with a left-handed groundball pitcher throws against a right-handed flyball hitter when the infield is shifted to the third base side? Sabremetricians can get even deeper into the scenarios and the probabilities of certain outcomes than that, but you get the idea.

One fundamental concept of sabremetrics and probabilities is that a team that scores more runs than it allows is more likely to win more games than it loses. Sounds pretty simple, right? Yet, tt wasn’t always something that was regularly thought about. Bill James, one of the forefathers of sabremetrics, popularized that insight (along with many others) around thirty years ago. If you ever check the baseball standings and see run differential, represented as the difference between runs scored and runs allowed, you can thank him for that. What made him so innovative was that he asked simple questions that challenged “common knowledge” assumptions and went about trying to find answers to those questions. In turn, he’d publish what he found in “The Bill James Baseball Abstract” and encourage others to look into what he was investigating. Thus, more people asking more questions lead to more knowledge about baseball. In case you’re wondering, he’s been working for the Red Sox since 2003 and while he obviously can’t claim all the credit for breaking the Babe’s curse, it’s very possible his insight on baseball has helped turn the Red Sox into a winning team.

If you want to talk about the importance of runs, you also need to talk about the components that make up runs. Sabremetrics, in recent years, has helped to look closer at those elements. If you looked on the backs of most baseball cards and in most baseball yearbooks before this millenium, the statistics you would see for a player would be their batting average, runs, home runs and runs batted in. Until sabremetrics, there was little true appreciation of how much walks helped an offense no matter how many times the adage “a walk is as good as a hit” was repeated ad nauseam by Little League coaches across America. Heck, some major league hitters would even get chastized for taking a walk.. even by some of the players-turned-broadcasters who were valued by sabremetricians precisely for their ability to get on base! Nowadays, the value of a walk has gone more mainstream and you can find on-base percentage and walks included in box scores and television broadcasts.

On the flipside, the perspective on what components make a good pitcher have also changed. It used to be that wins and ERA were the only factors that mattered when evaluating Cy Young Award candidates. Nowadays, ERA has diminished in importance among the sabremetric community. Research by Voros McCracken and others suggest that a pitchers tend to have some control over their walk rates, their strikeout rates and their home runs allowed. Whether batted balls that are not home runs are converted into outs are more dependent on a team’s defense than a particular skill the pitcher has besides, perhaps groundball/flyball ratio. This increased understanding of what pitchers can influence has permeated through mainstream baseball thought, to the point where the voters in 2010 awarded Felix Hernandez the Cy Young Award though he won only thirteen games and was only one game over .500 (13-12). Why did he win? Because he was a monster at preventing runners from reaching base but received so few wins because of the anemic offensive performance of the Seattle Mariners offense.

Another nice thing about sabremetrics is it makes it possible to compare players from different parks, leagues, eras and even competition levels. WHen new innovations come to sabremetrics, they are often applied to past years to arrive at new insights. Baseball is a great game for that because, to paraphase Kevin Costner from “For the Love of the Game”, they count everything in baseball and those counts can always be reevaluated and put into a new context. Tools like Pitch Fx can tell you about the movement and velocity of a pitcher’s pitches in such a way that you can see how a Stephen Strasburg fastball differs from Aroldis Champan’s. There are projection systems to predict what major league players would do like PECOTA and Marcel. You can even use Brian Cartwright’s Oliver projection system to look at Japanese, Korean, semi-pro, Cape Code and college leagues and estimate how those  players would perform if they were in the major leagues.

As sabremetricians identify components to scoring runs and winning games, they developed the idea of a replacement-level player to measure the talent on a given major league roster. A replacement-level player is basically a freely available Triple-A talent that any team could acquire for cheap. The idea is to identify players who provide as much value to your team as possible above replacement level with as low of a cost as possible. Measurement tools such as EqA, VORP and WAR all are used in such a fashion and can even be used to see if a pitcher is more valuable than a hitter or vice versa. Besides identifying value, keeping costs low can also be used to get the most “bang for your buck”. This concept was most illustrated in Moneyball where Oakland A’s GM Billy Beane had to figure out how to field a competitive team with a small payroll. To do so, he utilized sabremetrics to find market inefficiencies in how the rest of baseball evaluated players and used those efficiencies to acquire players who could provide as much value as possible for as cheaply as possible.

Yet, beyond increasing our understanding of baseball, the best aspects of sabremetrics are that it encourages passionate discussion among fans about the game, then provides those fans a framework to do their own research. Ever wonder if a certain pitcher is injured? You can look at the velocity and movement of his pitches for a clue. Do you really think Hank Aaron was better than Babe Ruth? Sabremetrics gives you tools to do comparisons. Just like a baseball commentator during a game or comparing the backs of baseball cards with your friends, sabremetrics provides just one more way of looking at players and thus, another way to appreciate the game.

So how do you get started? Browse the web. They’re all out there. Pick a name from this post and look up some of what they have written. Check out some of the baseball links on the right hand side of this post and click on any article you find there that looks interesting. Go to google and look up a term like WAR or on-base percentage. Take a peek and you might just see something about baseball you didn’t before. And once you do, make sure to tell your friends!