Smart Baseball
Page 11
Most important, though, our stories about the stars will be driven by fact, not wishful thinking, which makes the stories more compelling, and the greatness of the game’s best players that much easier to appreciate.
PART TWO
Smart Baseball
8
OBP Is Life:
Why On-Base Percentage Is the Measure of a Hitter
If on-base percentage is so important, then why don’t they put it up on the scoreboard?
—Jeff Francoeur, 2009, failing to realize that OBP did indeed appear on the Turner Field scoreboard for each batter
If you knew nothing about baseball statistics, traditional or modern, how would you try to measure the value of a baseball player’s performance? In other words, what if you had to start from zero and build a new vocabulary of baseball stats?
This is the sort of question that the earliest baseball analysts, from Branch Rickey and Earnshaw Cook to Bill James and Pete Palmer, tried to answer, sometimes deconstructing older statistics and sometimes just building their own new ones from scratch. So much of what I’ll discuss in this section of the book builds on their work and the work of others, writers and thinkers who realized things that today seem obvious—for example, that outs are bad and form the “cost” of offense, or that pitchers control some outcomes more than others. In the world in which most of us grew up, prior to radar and computer systems that could identify each pitch and its velocity and spin, we had to take the most basic accounting information about each game and try to figure out what data really mattered.
I’ll start here, with the most useful of all basic offensive statistics, on-base percentage, which tells us the simplest and yet most important thing of all about a hitter: How often does he make an out? If you asked the average baseball fan what a hitter’s job is at the plate, he or she would likely answer with something along the lines of “to get a hit.” Maybe something more nuanced, like “to drive in runs.” This is what we’ve been told since we could first sit up on our own in front of the TV set. Hitters hit.
It sounds good, like Baseball Populism, but it’s not right. A hitter’s job is to not make an out. If you’d prefer to phrase it in a positive way, then his job is to get on base. The result of every at bat is that the hitter reaches base or he makes an out. If we’re just talking performance, there is no particle smaller than this one. Getting on base, or not making outs, is a definable, repeatable skill for hitters. And nothing measures that skill better than on-base percentage.
Sabermetrics started to creep into mainstream baseball writing in the early 2000s, after the publication of Moneyball, the Red Sox’s hiring of Bill James as a consultant, and other smaller changes alerted the less retrograde journalists covering the sport that they needed to adapt to keep up with the teams and front offices they covered. (The sport’s media were and still are infected by a strong Luddite streak that won’t go away until that generation of writers has died off.) We didn’t get rid of the bad stats, but we started to see new ones appear in print.
The most obvious change was the inclusion of what is commonly called the “triple-slash line”—batting average, on-base percentage, and slugging percentage, in that order. The triple-slash format is common enough that you will frequently see those three stats (separated, actually, by just two slashes, but let’s not be pedantic here . . . there are better targets for our pedantry to come) without labels. Bryce Harper, the unanimous National League MVP in 2015, hit .330/.460/.649 in that season, leading the National League in the final two categories. But regardless of how we present these stats, my belief here is simple: you can’t tell the story of a hitter’s season without knowing at least his on-base percentage, and while slugging percentage isn’t as essential, it tells a lot of the story as well.
The proverb “OBP is life. Life is OBP” has been going around sabermetric circles for a while now; writer Joe Sheehan used it as the title of a 2006 piece on Baseball Prospectus’s site, and in that article he credits BP cofounder Gary Huckabay with coining the phrase. There is simply nothing more important to powering a baseball offense than putting men on base—and, conversely, not making outs.
On-base percentage is the most complete of all basic hitting stats, because it includes almost everything the hitter does, and the excluded events are justifiably thrown out. The common formula for OBP is:
OBP = (Hits + Walks + times HBP) / (At bats + Walks + times HBP + Sacrifice Flies)
While that’s a lot of terms, the underlying logic is simple: take the number of times the hitter got on base, and divide it by the number of times he came up to the plate in total. In fact, you could also express OBP as:
OBP = (Times on base) / (Plate appearances − Sacrifice Bunts − Catcher’s interference)
Those last two items are excluded from OBP for good reason: as discussed, the sacrifice bunt is a deliberate (and stupid) waste of a plate appearance, and reaching on catcher’s interference has nothing to do with the hitter. (It’s also quite rare. I go to a lot of games, and if I see it twice a year, that’s a lot.)
That makes OBP the measure of the player’s frequency in reaching base. We express OBP as a three-digit decimal, as we do batting average, but it would make just as much sense to express it as a percentage, so Harper’s .460 OBP in 2015 becomes 46.0 percent—that is, he reached base in 46 percent of his plate appearances, which led all of Major League Baseball.
The converse of OBP is never seen, but its importance might be more apparent to readers who haven’t spent much time reading or thinking about OBP. If you subtract a hitter’s OBP from 1, you get his out percentage: How often did his plate appearances result in an out. If I tell you that Wilson Ramos had a .258 OBP in 2015, the worst among qualifying hitters, that probably sounds bad. If I tell you that Ramos made an out in 74.2 percent of his plate appearances, you’ll wonder how the heck he wasn’t demoted to triple-A.
At a team level, on-base percentage is the best predictor of a team’s run production, which should be obvious with a moment’s thought: put more runners on base and make fewer outs and you will score more runs. With just 27 outs at your disposal each game, hitters who reach base more frequently (and thus make outs less frequently) will improve your offense.
In MLB history, the correlation between team OBP and team runs scored per game is huge—stronger than the correlation between team batting average and team runs scored. Using the common statistical measure of the correlation between two variables we discussed earlier, in chapter 1, where 1 is perfect positive correlation and 0 is no correlation, we find the following results:
All MLB teams, 1901–2015
Batting average to team R/G
0.827
OBP to team R/G
0.894
All MLB teams, 1960–2015
Batting average to team R/G
0.828
OBP to team R/G
0.897
In fact, in the 232 seasons of the major leagues from 1901 to 2015 (that’s 115 each for the AL and NL plus the two seasons of the Federal League), the team with the best OBP in its league also led its league in runs scored in 135, or 58 percent of the time, versus 53 percent for teams that led in batting average.
Useful as OBP is on the team level, it’s when you move it to individual players that you begin to see just how powerful it is at showcasing players’ actual ability to avoid getting out. One of the worst MVP votes in history came in the “rabbit ball” year of 1987, when home runs spiked without warning throughout the game, only to drop back to contemporary levels the following year, fueling rumors that MLB had altered the ball to try to boost offense and thus increase fan interest. (The league has never admitted to doing so, although it’s commonly accepted that the ball was different that year.)
That year also came in the midst of MLB’s collusion scandal, where a bunch of dim-witted owners, egged on by Commissioner Peter Ueberroth, decided to ignore American labor law and the existence of the league’s collective bargaining agreement w
ith the players’ union by coordinating their interests in free agents, driving down prices, and in some cases eliminating the markets for certain players entirely. One of the players hit by this cartel-like behavior—for which owners eventually paid $280 million in fines and lost salaries—was Andre Dawson, who had hit a respectable .284/.338/.478 with 20 home runs in 1986 for the Montreal Expos (#RIP). Dawson was a three-time All-Star, six-time Gold Glove winner, former Rookie of the Year, and two-time runner-up for MVP, so he was certainly well regarded as a player, and at thirty-two appeared to have plenty of production ahead of him. He found no takers whatsoever in free agency in the winter of 1986–87, and finally told the Cubs he would play for them for any amount of money—the “blank check” offer of legend—because he felt that the artificial turf in Montreal’s home ballpark was destroying his knees. The Cubs signed him to a one-year, $500,000 deal, maybe a third or a quarter of what he would likely have gotten in a clean free-agent market.
Dawson provided the Cubs with a career-high 49 home runs, leading the National League in that category and in RBI, winning a Gold Glove (primarily on reputation at that point), a Silver Slugger, and the NL MVP award, the last one by a significant margin over St. Louis defensive wizard Ozzie Smith and Smith’s teammate Jack Clark. The vote was silly at the time but has since become a standard example for the sheer idiocy of past awards voting, where the MVP usually just went to the league leader in RBI and the Cy Young to the league leader in pitching wins, because not only was Dawson not the National League’s most valuable player, he wasn’t even in its top ten.
To hit all those home runs and drive all those other runners in, Dawson had to get to the plate a ton of times, and in the process he made a lot of outs—the fifth-most outs at the plate by any hitter in the National League that year. Think of those outs as the cost of the production: To “buy” the 49 homers and the 24 doubles and the hits and the walks and so on, the Cubs had to “pay” with 445 outs. Dawson’s OBP of .328 was the 42nd best in the NL that year out of 56 hitters who had enough plate appearances to qualify for the batting average title; that OBP means that he made an out in 67.2 percent of his trips to the plate. The beat writers who voted on the MVP award focused on what Dawson did in the one-third of his plate appearances where he did something good and ignored the two-thirds where he didn’t.
The voting results are even harder to fathom today because the NL player who actually was the most valuable in 1987 was a household name at the time and remains one today, so it’s not a case of voters overlooking an unknown or disliked player in favor of a more popular one. The real MVP of the NL in 1987 was future Hall of Famer Tony Gwynn.
Gwynn and Dawson had nearly identical totals of plate appearances in 1987, with Gwynn getting 680 to Dawson’s 662, yet Gwynn did much, much more with his trips to the plate than Dawson did. Other than home runs, where Dawson’s total of 49 dwarfed Gwynn’s total of 7, Gwynn led Dawson in commanding fashion in every other category that mattered, especially the one that most explicitly addressed a hitter’s fundamental job: don’t make an out.
Dawson made 70 more outs at the plate than Gwynn did in 1987, even though Gwynn had a few more plate appearances. Gwynn had more singles, more doubles, more triples, and more walks than Dawson did, reaching base in 44.7 percent of his trips the plate, compared to Dawson’s 32.8 percent. The cost of Dawson’s production relative to Gwynn’s was all of those extra outs. Gwynn made fewer outs and provided those singles, doubles, triples, or walks, each of which increased rather than decreased the Padres’ chances of scoring runs.*
Gwynn was also a better baserunner and defender than Dawson in 1987, which should have made the voters’ choice easier—and he was a beloved figure, witty and popular and good with the press, so even some of the usual nonsense that might have hurt a candidate (like Clark) didn’t apply here. Instead, 11 of the 24 ballots listed Dawson first overall, while none had Gwynn at the top, and Dawson garnered 269 voting points* to win the award, while Gwynn only earned 75 points and finished eighth. It probably didn’t help Gwynn’s cause that the Padres finished with the worst record in the league that year, but the Cubs didn’t fare much better, going 76-85 to finish last in the National League East.
Egregious as this vote was, there are plenty more like it, often between players even more beloved and revered in Cooperstown. Another botched MVP vote came in 1941, pitting a popular Yankee who set a record that still stands as of this writing by collecting a hit in 56 straight games against a less-popular member of the Red Sox who set a record for the highest OBP in a single season that stood until 2002.
Both Joe DiMaggio (Joltin’ Joe, the Yankee Clipper) and Ted Williams (the Kid, Teddy Ballgame, the Splendid Splinter) stand in the inner circle of all-time greats in the history of baseball. Both are obvious Hall of Famers, on merit and on the whole “fame” thing. DiMaggio inspired his own song while still playing, and then appeared in an iconic line from Simon & Garfunkel’s “Mrs. Robinson.” Williams became a highly decorated pilot, serving in World War II and the Korean War, and his retirement inspired one of the most famous pieces of literary baseball writing ever, novelist John Updike’s New Yorker essay “Hub Fans Bid Kid Adieu,” which itself opens with an iconic line: “Fenway Park, in Boston, is a lyric little bandbox of a ballpark.”
But in 1941, the twenty-six-year-old DiMaggio and the twenty-two-year-old Williams were the two best position players in the American League, rising stars in a country not yet at war, playing for bitter rivals who would eventually finish 1-2 in the standings, with the Yankees taking the pennant by 17 games. DiMaggio’s hitting streak made him front-page news for much of the summer and enhanced his stature as a celebrity beyond the field, but even those 56 games couldn’t get him close to Williams’s level of production.
Stats from 1941 look like typos today given how strikeouts are now at record levels in the majors, but Williams and DiMaggio were especially adept at putting the ball in play. Williams punched out only 27 times in 1941, the lowest total in any of the 13 seasons where he qualified for the batting average title. DiMaggio struck out just 13 times in 1941, the lowest figure of any of his 13 seasons; in April and May 2016, Colorado shortstop Trevor Story struck out at least once in 24 consecutive games, 41 strikeouts in total, more than DiMaggio ever struck out in any single season of his career.
DiMaggio’s hitting streak helped him to a fine season, .357/.440/.643, 43 doubles, 11 triples, 30 homers, and, if you care about such things, 125 RBI. In many years that might have made him the (lowercase) most valuable player, but in 1941 he couldn’t touch Williams’s line: .406/.553/.735, 33 doubles, 3 triples, 37 homers, and 120 RBI. The two players’ power outputs were similar, especially when you consider that Fenway Park was an easier place to hit for power than Yankee Stadium was in 1941, but DiMaggio made 78 more outs that season than Williams did in just 16 more plate appearances. That gap is how Williams led DiMaggio by more than 100 points of OBP, and no player other than Barry Bonds has ever posted a higher on-base percentage than Williams did that year.
The hitting streak and the players’ relative popularity with the press—Williams’s antagonistic relationship with writers was well documented during and after his career, including in his own memoir—trumped Williams’s commanding advantage in, you know, actual value. DiMaggio took 15 first-place votes to Williams’s 8 (one writer, likely from Chicago, put White Sox pitcher Thornton Lee first) and won the award 291 points to 254. And that is how the player who put up one of the 30 most valuable seasons by any hitter in MLB history ended up without an MVP award to show for it.
One reason why fans don’t seem to consider the “cost” of production in terms of outs is that we don’t really count outs as an individual stat. We count hits, and walks, and specific types of hits, and we count certain outs like strikeouts and sacrifice flies and double plays (which aren’t even an individual stat, really), but there’s no column even on most online sortable stat pages that says “Outs.” Some sites, like Baseball
-Reference, provide “Outs” but calculate it in peculiar fashion, double-counting those times the hitter grounded into a double play, adding in times caught stealing (that’s not his at bat), even adding in sacrifice hits (often a manager’s call, which is why we exclude them from plate appearances entirely).
In fact, just looking at the all-time leaders in outs casts a bit of a new light on some of the all-time leaders in certain offensive categories:
Player
Outs*
Pete Rose
9876
Hank Aaron
8714
Carl Yastrzemski
8674
Cal Ripken Jr
8494
Eddie Murray
8209
Robin Yount
7989
Dave Winfield
7988
Rickey Henderson
7973
Brooks Robinson
7920
Craig Biggio
7897
Omar Vizquel
7803
Derek Jeter
7788
Willie Mays
7689
Luis Aparicio
7629