SABR Analytics Notebook: Tying Up Loose Ends

PHOENIX—A smattering of interesting topics discussed by panelists and research presenters during the three-day SABR Analytics Conference in Phoenix. Make use of the hashtag #2014 Analytics Conference for a complete archive of content at Baseball America.

• From “Which Numbers Athletes Love, Which Numbers Athletes Hate,” with commentary from ESPN analyst (and former Nationals and Indians manager) Manny Acta, ESPN analyst (and Astros bench coach in 2013) Eduardo Perez and ESPN analyst Aaron Boone.

SABR 2014 LogoActa stole the show at this panel, saying that he first learned to appreciate baseball analytics while working as a coach with the Mets in 2005. “A clubhouse kid convinced me that a team had a greater scoring opportunity with a runner and first and none out as opposed to a runner on second and one out after a sacrifice bunt,” he said. “Once he showed me the numbers to back it up, I was hooked. But then I began to think about all the minor league hitters I had managed in the past—Why did I not let them swing the bat?”

Acta said his metrics of choice are WHIP (“traffic on the bases equals runs allowed”), batting average on balls in play, or BABIP (“it can be telling if a pitcher without power stuff might come down to earth next year”) and Wins Above Replacement, or WAR (“it can settle the argument about who’s the best player—not the best hitter, but the best player.”)

He’s no fan of a defensive alignment in which the third baseman guards the line. “It sickens me to see a third baseman standing on the line because nine out of 10 groundballs will go to the shortstop hole, not down the third-base line.”

Players are not always receptive to many of the new metrics.

“Players hate the new numbers because so many of them begin with a minus sign or a zero,” Acta said. “It’s easier to see that a .260 average with seven homers and 53 RBIs looks better than a -0.2 WAR. They like the good, old, traditional numbers.”

Acta related another baseball pet peeve of his, which occurs when players high-five the batter who drives in the runner from third with no outs on a lazy sacrifice fly to left field. “I’m thinking, ‘What about the guy who scored the run?’ Maybe he worked the count and got on with a walk or a single, went first to third on another single, then scored on the batter’s sacrifice fly.”

Boone talked about how prospects of the mid-’90s analyzed the numbers.

“When I was minor league player, the first thing I looked at for the opposing pitcher was his innings, hits, walks and strikeouts,” he said. “Those numbers painted a picture. These days, players can watch video, but back then I would watch the pitcher in a side session and take note of a slider that might look ordinary in the bullpen, but with all those strikeouts on the stat sheet, I bet that’s not the case.”

Like Acta, Boone uses BABIP and WAR, while also mentioning fielding-independent pitching ERA, or FIP, because it incorporates many of the inputs—strikeouts, walks and innings—he relied on 20 years ago.

Boone also praised the work of Athletics general manager Billy Beane and manage Bob Melvin.

Bob Melvin

Bob Melvin (Photo by Bill Mitchell)

“After being in the booth for their playoff games last year,” Boone said, “I’ve come to appreciate the A’s, in that they receive similar value from the sixth through 25th players on the roster. That makes everybody important.

“The Rays do this, too. Those teams will take three players to fill two positions, with the right manager in charge to match up players—not just left/right platoons, but maybe this guy is a better matchup against a sinkerball pitcher.”

Boone doesn’t dispute the validity of research indicating that value supplied by a catcher’s pitch-framing skills can be on the magnitude of 20 runs or more per season.

“Growing up in a catcher household (his father Bob caught for 19 seasons, winning seven Gold Gloves), I am receptive to the value of a catcher’s pitch-framing,” he said. “Can he steal strikes for his pitcher? Does he receive well night in and night out? Those are important considerations.”

Perez mentioned briefly an Astros proprietary metric he called BABIP+, which he said incorporates velocity off the bat into its calculation. A consistently high exit velo for a batter means his high BABIP results probably are not a fluke, much the same way a high line-drive rate correlates positively with a high BABIP.

All members of the panel reserved some skepticism for the readings from the advanced defensive metrics because they don’t always agree with one other, and they often don’t pass the eye test.

• From “An In Depth Study Of Team Chemistry,” a research presentation by SABR president Vince Gennaro. 

Even though team chemistry tends to be undervalued by the analytics community, Gennaro said, that doesn’t prove its nonexistence. After all, players believe in chemistry, and they view it as integral to team success, he learned through interviews with players, former players, coaches and front office personnel.

He singled out the Cardinals as an ideal organization for his examination of team chemistry. The front office sets the tone, which is relayed to players by manager Mike Matheny and his coaching staff. St. Louis also has visible alumni in prominent roles, and a fan base that feels the impact of the club’s success or failure. This, according to Gennaro, creates a sense of higher purpose for Cardinals players.

Ultimately, Gennaro singled out four necessary base ingredients for positive team chemistry: accountability, trust, belief in team and communication. Those attributes foster an atmosphere of confidence, motivation and empowerment for the club.

A number of players told Gennaro that they became better players when playing on a team with great chemistry because teammates motivated them to improve. On teams with lesser chemistry, the effort level can be inconsistent.

• From “Analytics From The Player’s View,” with commentary from Diamondbacks righthander Brandon McCarthy and retired righthander Brian Bannister.

McCarthy reprised his role on the player panel for the third time in the conference’s three years. (One wonders what would happen if he were to sign with a club that trains in Florida.) Joining him was Bannister, an early and visible adopter of advanced metrics during his big league career, which spanned 2006-10 with the Mets and Royals.

A resident of nearby Scottsdale, Ariz., Bannister told the audience he felt lucky to have been around so many great pitchers in his life. His father Floyd went No. 1 overall in 1976 draft and spent 15 seasons in the big leagues. Growing up in Houston, Brian witnessed Nolan Ryan throw bullpen sessions between starts. When he moved on to Southern California, Bannister pitched in the same rotation as Mark Prior. As a rookie with the 2006 Mets, he pitched in the same rotations as Tom Glavine and Pedro Martinez at the end of their storied careers.

Bannister focused most of his time on the panel to discussing pitching styles, the major league strike zone and how to improve utility of today’s advanced technology for young pitchers.

“If you’re not an elite talent, like I wasn’t, you have to optimize everything you do,” Bannister said. “For me, I determined that I would need to throw more groundballs in order for my ERA to go down.

“It’s really simple: If a pitcher can’t throw the ball past the batter up in the zone, then he’s only doing the batter a favor by throwing a high four-seam fastball. That helps him hit under the ball and loft it. That’s why I teach kids to throw a two-seamer or a cutter to try to prevent the batter from lifting the ball.”

Bannister said that the refinement of two pitch types had the largest impact on his career.

“I finally found success with my changeup when I stopped trying to throw it slower,” he said. “Every pitching coach will tell you that you need the 10 miles per hour of separation, but as long as the changeup sinks you can use it to get groundballs. That’s how I had success at the end of my career with the Royals.”

He also learned to trust his cutter because the pitch produced reliable vertical drop in the zone. In fact, he credits pitch movement for helping to drive a pitcher’s groundball rate or rate of home runs per flyball.

“More sink on a fastball or changeup induces the batter to hit a softer groundball,” he said, “so that (retired Diamondbacks ace sinkerballer) Brandon Webb would give up a five-hop groundball to the second baseman, while a pitch with less sink would produce two hops and maybe get through the infield.”

Bannister talked about how Giants righthander Matt Cain is able to consistently allow a low, and seemingly unsustainable, rate of homers per flyball.

“Cain has a fastball with late life up in the zone, a pitch that batters just can’t get on top of,” he said. “That means that instead of hitting it square, they tend to hit the very bottom of the ball, which explains why so many of his flyball stay in the park and die on the warning track.”

Bannister said that Glavine told him he had to change pitching styles after the implementation of Questec pitch-tracking systems, the predecessor of Pitch f/x.

“The umpires would expand the strike zone in the minors,” Bannister said, “but that went away in the big leagues. Glavine told me that couldn’t pitch the way he used to because the umpires called a narrower zone. He said Questec made umpires accountable.

“The major league strike zone is so much more narrow than in the minors that you lose parts off the corner of the plate. You end up favoring pitches with downward action—sinkers, changeups, splitters, sliders. Anything that breaks down will be more popular with a narrow zone.”

Echoing a theme heard more than once during the Analytics Conference, Bannister said that while the advanced technology in today’s game is great, it needs to filter down to the minors, where it can help developing players.

“You have to integrate the new technology all the way down,” he said, “to change the way players train and practice, because the big leagues are not the place to make fundamental changes.”

Brandon McCarthy

Brandon McCarthy (Photo by Mike Janes)

McCarthy avoided doubling up on most of the topics he discussed in 2013, but he also touched on advanced metrics as a diagnostic tool, the trend toward more defensive shifting and the future of the closer’s role in the majors.

“I observed what I thought was a problem and diagnosed it with analytics,” McCarthy said of recent struggles to find consistency. “Batters were making contact with more pitches in the zone and not swinging at stuff out of the zone as much.

“As a pitcher, I know general batter tendencies, but it’s good to have the backup confirmation.”

Some pitchers don’t like to pitch in front of an infield shift—Rays lefty David Price being a notable example—because they feel like it removes pitches from their repertoire, but McCarthy is in favor of the maneuver.

“If David Ortiz hits a one-hop rocket to short right field and our second baseman is playing right there,” he said, “it’s rewarding to get him out in a smarter fashion, even if I didn’t make a good pitch.”

McCarthy saved his strongest opinion for the practice in which teams designate and condition closers to pitch only the ninth inning, even if the game’s most crucial situation occurs earlier. “Throw out the save stat,” he said, “and money is no longer the driving issue (for pitchers to pursue the closer’s role).”

McCarthy believes change could only be effected with a bottom-up approach. “If teams eliminate the closer’s role in the minors,” he said, “then it might one day go away in the majors. But as it is, teams immediately divide their minor leaguers into bullpen roles—closer, set-up man, and so on.

“I would like to see all relievers in the minors be available from the fifth inning on. If a reliever is brought through the system being used by his manager in game situations with the highest leverage in which he can succeed, then we might see some change in the majors.”

 • From “Decision Making In The Front Office,” with commentary from Mariners general manager Jack Zduriencik, Rockies senior vice president and assistant GM Bill Geivett and Giants vice president and assistant GM Bobby Evans.

While the panel focused on major free agent acquisitions such as the Mariners’ Robinson Cano and Cuban first baseman Jose Abreu—whom Geivett said had “tremendous bat potential” and probably would have been a Rockie had the White Sox not narrowly outbid them—they did also discuss the role of analytics in decision-making.

While noting that “heart so often determines success or failure for a prospect,” Zduriencik boasted of the expansion of Seattle’s analytics department.

“We have increased our analytics department to the point where we have eight full-time members of the team,” he said. “Two work on the amateur side, two for pro scouting and two are on the advance scouting side. I don’t know how any GM can function without them, because analytics bring everything together.”

Geivett spelled out the various arms of analytics the Rockies use prior to making a player acquisition. “We’ve been deep in analytics for a long time,” he said. “We have four main branches: medical, contract/money, scouting reports and analytics.” When all four come back positive, as they did for Abreu, the club enters full pursuit.

The Rockies valued Jose Abreu, but he signed with the White Sox. (Photo by Bill Mitchell)

The Rockies valued Jose Abreu, but he signed with the White Sox. (Photo by Bill Mitchell)

Geivett also said the Rockies formulate their own proprietary version of WAR to help determine whether a player would be a good fit for Coors Field.

Evans said the Giants front office, one of the longest-tenured in baseball, has evolved with the times.

“We’re perceived as an old-school organization, but the truth is that every club goes through transition,” he said. “We began applying analytical methods first with amateur players, then brought that to the majors.

“We prepare GM summary reports for Brian (Sabean) with the metrics that he prefers, the ones that tell the best stories. In the end, we weigh everything—scouting reports, numbers, comments we have on file—and everyone in the baseball ops department plays a role in analytics.”

• From the panel “Inside the SABR Defensive Index,” with commentary from Sports Reference president Sean Forman and SABR president Vince Gennaro.

The panel shed light on the nature of the system used for the 2013 Gold Glove balloting. For the first time, American and National league managers were presented with finalists at each position who had been vetted by the advanced defensive metrics.

Gennaro spelled out the five measures, each weighted equally, that determined last year’s Gold Glove finalists: 1. Defensive Runs Saved, or DRS, by Baseball Info Solutions; 2. Runs Effectively Defended, or RED, by Chris Dial; 3. Ultimate Zone Rating, or UZR, by Mitchel Lichtman as presented at FanGraphs; 4. Total Zone by Sean Smith as presented at Baseball-Reference; and 5. Defensive Regression Analysis, or DRA, by Michael Humphreys as presented in his book “Wizardry.”

Forman believes that defensive metrics “capture 85-90 percent of what’s going on defensively,” and that even if MLB Advanced Media does not make data from its forthcoming, hyper-detailed tracking system available to the public, having access to historical data down the line could benefit researchers.

“Even if we don’t get BAM data for current year, he said, “if we begin to receive it retroactively, we can then compare its results with popular defensive metrics in a correlational study to see which of the existing metrics is most accurate.”

Gennaro raised an interesting point in that, while many defensive models can estimate overall defensive effectiveness, the can’t isolate a defender’s range, pre-pitch positioning or arm strength and credit each of the three component separately.

“We’ll get component parts with the BAM data,” he said, “with which we can measure things like first-step quickness, acceleration, arm strength and outfield release times on throws.”

He envisions a day when a defender’s reliability and sure-handedness will take precedent over pure range because advanced data will help to optimally position the defense.