Advanced baseball statistical analysis: Game-changer or fuzzy math?
05/11/2013 7:14 PM
11/12/2014 2:47 PM
A yes-or-no question turned into a six-minute lecture by Derek Lowe. The topic of advanced statistics and their increasing popularity struck a chord with Lowe, who saw this past off-season how influential they can be.
Do players pay attention to them? “God, no.”
He took a breath and shook his head. “No.”
Lowe then gave an analogy: What do parents ask their children when they come home from a game?
“Did you win or did you lose? That’s it,” Lowe said. “Nobody asks, ‘What was your WHIP?’ ‘Uh … what was your WAR?’ It’s always, ‘Did you guys win?’ ”
Started by Bill James in the 1980s, the advanced statistics movement now includes formulas that seem more suited for engineers and physicists than pitchers and outfielders.
Batting average, RBIs and earned-run averages aren’t enough these days. Now there’s wOBA (weighted on-base average), UZR (ultimate zone rating), FIP (fielding independent pitching) and DIP (defense independent pitching).
Lowe talked about his tumultuous off-season when he tried to find a job. He had pitched well at the end of last season and had a long, respectable career. But it didn’t add up to offers.
Lowe, who turns 40 on June 1, didn’t find a team until the first week of March when the Texas Rangers signed him to a minor-league contract with an opportunity to become the team’s long reliever.
Lowe won the job with the Rangers and has since found out that at least three teams wanted to sign him in a similar capacity. However, he didn’t pass the “stats test.”
“If you pump my numbers into the system compared to, let’s say, Tanner Scheppers, of course his stuff is going to outscore my stuff, I’m not naive,” Lowe said. “He’s a young kid who throws 98 mph with a great breaking ball. Listen, I know I don’t pass the test.
“But it doesn’t take into consideration the human element of sports. Don’t get me wrong, I think those stats can be beneficial. But I use more of a human element. Where has the guy had success? What cities has he had success? What cities has he failed at? Has he performed well when it matters?
“I’m a big believer in winners and losers. Guys that know how to win, and they’re not always the superstar players. There are guys that don’t perform well when push comes to shove, but those things don’t show up in statistical data.”
As Rangers catcher A.J. Pierzynski said, “At the end of the day, if your team wins, everything else takes care of itself. As a catcher, if I stink at the plate but get our pitcher through six, seven innings, it’s a good day.”
While players might not pay close attention to advanced metrics, there’s no question about increasing relevance.
Bill James, the father of sabermetrics, got the movement started in the 1980s, and it’s taken off ever since.
Hollywood made a movie on it two years ago, based on the book Moneyball, which took an in-depth look at how Oakland Athletics general manager Billy Beane constructed his team, based largely on numbers.
Rangers manager Ron Washington was the A’s third-base coach during the Moneyball years, and remembers a few animated conversations with Beane. The most vivid was when Beane tried to convince Washington that Frank Menechino was a better defensive second baseman than Mark Ellis.
“You can look at Mark Ellis and Frank Menechino and there’s no comparison,” Washington said. “Mark Ellis doesn’t miss anything hit his way, turns the double play as well as anybody. … Defense, to me, is still an eye test.”
Trying to quantify defense is one of the flaws of arguably the most popular advanced stat in the game today, WAR — wins above replacement. The stat, which is computed different ways by baseball-reference.com and fangraphs.com, is used to determine the overall worth of a player by factoring in his offense, defense and base running compared to that of a “replacement” player.
Baseball-reference.com has a lengthy explanation on the intricacies of WAR, complete with graphs and charts of how it’s calculated.
Part of it reads, “The idea behind the WAR framework is that we want to know how much better a player is than what a team would typically have to replace that player. We start by comparing the player to average in a variety of venues and then compare our theoretical replacement player to the average player and add the two results together.”
What is a replacement player?
As Yahoo! Sports’ baseball blog, Big League Stew, writes: “A replacement player is defined as someone who is below average and should be easily obtainable, the sort of fringy cup-of-coffee guy you can find in Triple A, on the waiver wire, or acquire for a PTBNL, a warm body who hurts the team the more he plays.”
In other words, WAR is a little more involved than batting average ( hits divided by at-bats) or on-base percentage ( hits plus walks plus hit by pitch divided by plate appearances).
But WAR has made its way into the mainstream, as shown by last year’s debate on who should be the American League’s Most Valuable Player.
The stats-based segment of fans felt the award should go to Angels outfielder Mike Trout, even though the Tigers’ Miguel Cabrera became the first player to win the Triple Crown since 1967. Cabrera, who won the award, led the league in batting average (.330), home runs (44) and RBIs (139).
Trout had a 10.9 WAR, though, while Cabrera had a 7.3, according to baseball-reference.com.
“The AL MVP race last year was the watershed moment when some of these advanced metrics became a major part of the debate to determine who was the best player,” said Sean Forman, founder of baseball-reference.com. “Ten years ago, it would have been Cabrera in a landslide, but now there is much more debate about it.”
Still, Forman understands that most casual fans don’t entirely grasp the concept of WAR or the other advanced statistics that have become popular such as weighted on-base average (wOBA), ultimate zone rating (UZR), fielding independent pitching (FIP) and defense independent pitching (DIP).
“The casual fan will find it unnerving to have somebody say that batting average or fielding percentage isn’t really important compared to other things, but that’s why we present both types of numbers on the site,” Forman said. “But once you see teams making decisions based on these numbers, that’ll push fans to accept it.”
Jon Daniels, the Rangers’ president of baseball operations/general manager, recently spoke at a Baseball Prospectus gathering and recalled being a guest for a Society for American Baseball Research event last spring where he and a few other GMs were asked about their stats departments.
To the surprise of many, Daniels revealed he doesn’t have a statistical team in place.
Instead, most in the Rangers’ front office have an understanding of the advanced statistics and blend it together with the scouting reports. So far, that structure has worked out well for the Rangers.
“We rely more heavily on scouting reports than statistical analysis,” assistant general manager Thad Levine said. “We have guys that can crunch the numbers, but that’s not all they do.”
When it came to signing Lowe during spring training, for instance, Levine said it was a combination of scouting reports, the makeup of the player and recommendations from those within the industry.
Another example of the Rangers going with scouting over stats is when they drafted starter Justin Grimm out of the University of Georgia in the fifth round of the 2010 draft.
Grimm had gone 3-7 with a 5.49 ERA his junior season, but the Rangers saw something they could develop. Grimm has since reached the big leagues and was recently named the AL Rookie of the Month for April.
The Rangers have stats-based signings, as well. Levine said reliever Jason Frasor might not blow you away on the scouting reports, but he had a long track record of success as a reliever in one of the toughest divisions (AL East).
The advancement of stats has taken Levine by surprise. He got his first job in the game back in 1999 as someone who was ahead of the curve from an analytical standpoint, but it has grown substantially in the past 10 to 15 years.
“It’s changed dramatically,” Levine said. “There are always people who are looking to produce the next new formula. And it tells you something when teams hire people away from FanGraphs and Baseball Prospectus to work for them.”
The Houston Astros might be at the forefront of the stats movement with a front-office staff that includes Sig Mejdal, a former NASA engineer, as director of decision sciences. Houston also hired Kevin Goldstein, a Baseball Prospectus writer, as its pro scouting coordinator.
Other teams, such as the Brewers, find themselves straddling the line between scouting and stats, similar to the Rangers’ approach.
Milwaukee general manager Doug Melvin has a baseball research department that works with the scouting department. But Melvin is hesitant to read too much into advanced stats such as WAR.
“The problem with the newer stats is everybody has different formulas,” Melvin said. “We won 96 games with Prince Fielder [in 2011] and somebody told me his WAR was 6. So, if we lose him, we should still win 90 games, right? And they said, ‘No, that’s not the way it works.’
“So how do we value WAR? I don’t understand what that actually means or how to implement it in putting a team together. And the numbers are only based on that year’s performance; it doesn’t mean it’ll be predictive of next year’s performance.”
Josh Hamilton is a good example. He had a 3.7 WAR last season for the Rangers, but that number had dropped to minus- 0.8 through his first 35 games with the Angels.
As Melvin said, “Someone once told me that baseball is a series of random events that provide an unpredictable outcome, and I still think there is a lot of truth to that.”
Stats that matter
While WAR and wOBA and UZR might not mean much to players, there are a handful of stats that managers and players study before every game.
Ron Washington pays attention to how his batters have fared against right- and left-handed pitchers. He’ll also look at stolen bases and the tendencies when teams hit and run.
Pierzynski likes to look at an opposing pitcher’s count and location charts, but said, “There are so many things you can look at, but until you step into a batter’s box it’s hard to quantify what they’re trying to do.”
Said Rangers outfielder David Murphy: “I just like to know what percentage of the time they throw each pitch. I don’t want to go up there looking for a fastball if they throw a changeup 60 percent of the time. Other than that, I keep things simple up there. I’m just a see-it-and-hit-it kind of guy.”
For a veteran pitcher such as Lowe, he can rely more on his previous history with a certain batter. He tends to remember what worked well and what didn’t, for the most part.
Lowe also likes to know a batter’s hot zone — where he is hitting .380 compared to .110, for instance.
“Once you get all this information, you pick and choose what works for you,” Lowe said.
That’s probably the same approach your casual baseball fans will have with advanced metrics. Some of the numbers will register and mean something to them, and others won’t.
Lance Berkman summed it up nicely.
“I don’t even know if your casual fans are going to care about batting averages, they just come to the game and want their team to win,” Berkman said. “The difference between a .275 and a .300 average … it’s just numbers to them.
“People have been trying for years to try to quantify what they see. To me, the beauty of baseball is you can’t. It’s an impossibility.”
Join the Discussion
Fort Worth Star-Telegram is pleased to provide this opportunity to share information, experiences and observations about what's in the news. Some of the comments may be reprinted elsewhere on the site or in the newspaper. We encourage lively, open debate on the issues of the day, and ask that you refrain from profanity, hate speech, personal comments and remarks that are off point. Thank you for taking the time to offer your thoughts.