Monday, February 24, 2003

In Defense of Defense

Of the four main criticisms of Jim Rice’s talents, the attack on his defensive prowess is the most widespread.
  • "A dominant hitter for a decade, but he didn’t walk, didn’t steal, and didn’t field." Jon Heyman, The Sporting News, 12/21/2000.
  • "I voted for Rice three years ago but not the past two. Obviously, Rice did nothing in that time to diminish his candidacy. He certainly was a dominant player for 10 years, and I considered him a more feared hitter than first-ballot lock Dave Winfield. Why, then, did I change my mind? To me, a Hall of Famer should be a complete player. Rice's defensive inadequacies bother me. He didn’t run as well as Puckett or Winfield. He was ‘just’ a slugger – though a fabulous one, at that." Ken Rosenthal, The Sporting News, 1/1/2001.
  • "My reservation about Rice is that he was a one dimensional player…You essentially have to vote him in as a hitter only, because he DH-ed extensively. He gave you no speed, no Gold Gloves, no off-field ‘character-and-integrity’ points." Jayson Stark, ESPN.com, 1/15/2001
  • "[Rice was] a one-dimensional, no-speed, no-defense left fielder…" Dan Shaughnessy, The Boston Globe, 1/9/2002

One common theme among the criticisms is a complete lack of supporting documentation. These are opinions only, some not even based upon first-hand observation. In addition, each of these comments was written more than a decade after Rice’s retirement. Do they represent the opinions held at the time Rice played, or those expressed by his teammates, managers, coaches, or those who observed him play? In a word, no.

  • "…He plays The Wall almost as well as Yastrzemski did in his prime. His arm is as good as any left-fielder’s in the league, and he routinely holds balls off The Big Green Thing to singles." Steve Wulf, Sports Illustrated, 8/6/1984
  • "While Rice’s hitting commands attention, he also has mastered the art of playing balls off the left-field wall. On several occasions this year, he’s either thrown runners out (he has 10 assists) at second or held the hitter to a single with his barehanded grabs of balls coming off the Green Monster." Joe Giuliotti, The Sporting News, 9/20/1982
  • "With George Scott injured and Yaz filling in at first base, Rice has played more in the outfield this season, and played it well." Time, 6/12/1978
  • "He took over for me in left field and really learned how to judge balls off the left field wall." Carl Yastrzemski, RedSox.com, 1/29/2001
  • "…He became a better-than-average left fielder who could really play that Fenway Park wall." Don Zimmer, RedSox.com, 1/29/2001
  • "He may have been the hardest working ballplayer I was ever associated with. We would go out to the park before anyone, and I would hit ball after ball to him in left field. He became a master of playing that tricky Green Monster wall." Johnny Pesky, RedSox.com, 1/29/2001
  • "In ’75, his rookie season, Rice played 98 games in left, didn't make an error, threw out six runners and played the unpredictable caroms off the Fenway Green Monster flawlessly." Mark Ribowsky, Sport, July, 1978.
  • "Rice always worked on his skills…Despite his speed, arm strength, and body control that allowed him to dive and cradle the ball without the jar of landing, he had to work hard on playing left field…Five days a week he would take fungoes from Pesky, first practicing charging grounders, then taking flies off The Wall. After eight seasons, he finally got out of the shadow of Yastrezemski, whose unique style of playing left field like a shortstop had paled anyone else who went out there. But by his ninth year in Boston, Rice, too, owned the new Wall, and in 1983 he had as many assists (twenty-one) as Yaz ever had in one season and probably should have won a Gold Glove for fielding excellence. Dwight Evans, who had the worst defensive year of his career, won one instead, proving clearly the value of a reputation." Peter Gammons, Beyond the Sixth Game, 1985.

The anecdotal evidence from first hand observers makes it clear that Rice developed a solid defensive reputation during his playing days. Only later did writers – usually writers who only witnessed a hobbled, aging Rice in his final three seasons – paint him as a defensive liability. It’s an easy, albeit incorrect, conclusion to draw for those who saw his waning days, when in reality Rice was worn down by a dozen years of playing through every injury that didn’t involve broken bones.

Besides, as Gammons pointed out, it took Rice a while to earn that good reputation. When he first arrived in Boston, he began his career playing more games at DH than most players his age since he was literally surrounded by Gold Glove winners who commanded more playing time in the field.

The 1975 Sox featured more defensive talent than any team in the franchise’s history. Catcher was manned by Carlton Fisk, second base by Doug Griffin, shortstop by Rick Burleson, first base by Carl Yastrzemski, center field by Fred Lynn, and right field by Dwight Evans. Each of those men won one or more Gold Gloves in their careers. Even the Sox primary DH, Cecil Cooper, went on the win two Gold Gloves at first base. Only in left field and at third base did the Sox lack a former or future Gold Glover, though neither Rice nor Rico Petrocelli was a slouch with the glove. Beyond the everyday defensive excellence, the Sox also featured two reserve outfielders, Rick Miller and Juan Beniquez, who eventually went on to win Gold Gloves.

With this much defensive prowess surrounding him, it’s no wonder that Rice came off looking like a poor defender. Perhaps, at that stage of his career, he really did pale in comparison. The reality, of course, is much more complex than that. Calling Rice the worst defender on the late-1970s Red Sox is like calling a guy slow because he finished last in the Olympic 100-meter dash. It’s simply not a fair comparison.

Beyond the problem of failing to measure up to his stellar teammates, Rice faces the problem of all left fielders. Bluntly, left fielders get no respect. At least, that’s the case when it comes to defense.

With the exception of a privileged few, like Barry Bonds, left fielders are always thought of as these massive clods who can’t play any other position. Their managers just stick them out in left to keep them from hurting the team and hurting themselves. To most people, Greg Luzinski and Ron Kittle are the prototypes for left fielders.

And it isn’t just the fans that hold this view; it’s the view that is officially condoned by major league baseball. The decision several decades ago to allow Gold Gloves for outfield play to be handed out to any three outfielders, regardless of position, has resulted in left fielders being virtually ignored for their defensive contributions. Year after year, three center fielders are awarded, with an occasional strong-armed right fielder thrown in. Left fielders get only the scraps.

From the inception of the Gold Glove in 1957, through the 2001 season, 135 Gold Gloves have been given to American League outfielders. Just 19 of those went to left fielders, and never more than one in any year. Of those, 15 are accounted for by just four men - Carl Yastrzemski (7), Minnie Minoso (3), Joe Rudi (3) and Dave Winfield (2 – the rest of his awards were for playing right field).

When Darin Erstad was awarded a Gold Glove in 2000, he was the first American League left fielder to be so awarded since Winfield in 1983. Even then, Winfield was a guy who made his reputation elsewhere, having already won two Gold Gloves as a right fielder in the National League. He didn’t actually deserve the award in 1983, or the previous year for that matter. Several AL left fielders were better, but Winfield had the reputation that he brought with him from right field.

When the situation is really considered, it’s clear that left fielders are in a silly position. Even pitchers get their own Gold Gloves. Does anyone in their right mind think pitchers make a bigger contribution with their glove than left fielders do? That’s ridiculous.

Consider the best-fielding pitcher ever, Jim Kaat, owner of sixteen consecutive Gold Gloves. During those sixteen seasons, theoretically playing the greatest defense of any pitcher ever, Kaat averaged just 1.48 chances per outing. That’s it. Even the worst-ever left fielder, let’s call him Lonnie Smith for argument’s sake, averaged 1.90 chances per game. And that’s a guy with truly awful range compared to the (supposedly) best fielding pitcher in baseball history. We’ve got to assume that the gap between an average pitcher and average left fielder is about a full chance per game. Multiply that by 140 to 160 games that a regular left fielder will play each year compared to the 35 starts a pitcher gets and we find that a left fielder will handle the ball about 250-275 more times per year than a pitcher. But somehow, it’s the pitcher that is awarded for his glove work.

So, what if this dumb rule didn’t exist? What if a Gold Glove were awarded to each outfield position, just as each infield position is awarded? Which left fielders in history would have received those awards each year, thus escaping the usual left field stereotype?

Of course, we’re not going to be able to reproduce the real voting process. And thank God for that. After all this is the system that gave Rafael Palmeiro a Gold Glove for playing just 28 games at first base, and one of those awards to Joe Rudi was for playing that grand total of 44 outfield games in 1975. Obviously this isn’t a system worth replicating, so we’re going to have to dig into the wonderful world of defensive statistics. There are a couple of ways to go about this.

First, the time consuming one. Using the fabulous Baseball-Reference.com web site, it’s simple to determine each team’s regular left fielder over the past 45 years, since Gold Gloves have been in existence. In some cases teams didn’t have a true regular. They would have one guy who played 45 games in left and another two or three who played 20-30 games each. So, largely to avoid some stupid statistical anomaly that would give the award to a guy who played 25 games in left field, for the purposes of this discussion eligibility will be limited to players who played at least half of their team’s games in left field.

With each of these players identified, their individual defensive numbers were examined. These numbers included each player’s Putouts per Game, Assists per Chance, Double Plays per Chance, Fielding Percentage times Chances, and Range Factor. These were used to level out the advantages that would have gone to those players who played 150 games instead of 120 or 110. But, to give credit for durability, each player was also given points for the number of total outfield games played.

Here’s the math, using Jim Rice’s 1980 season as an example. Rice played 109 games in the outfield, and totaled 233 putouts, 10 assists, 2 double plays, a .988 fielding percentage and a 2.23 range factor (RF). These translate to 2.14 putouts per game (PO/G), .041 assists per chance (A/C), .0081 double plays per chance (DP/C), and a 243.048 score in fielding percentage times chances (FP/C). With these calculated, they are compared to all of the other eleven left fielders who qualified for the award that year. Rice ranked seventh in games (this was the year he missed almost two months with a broken wrist), seventh in PO/G, fourth in A/C, third in DP/C, sixth in FP/C, and seventh in RF. After all of the ordinals are added, Rice has a total score of 34, which ranks fourth among the twelve eligible left fielders. Ben Oglivie’s total of 21 was the best. We’ll call this one the Defensive Statistics Method, or DSM since all baseball stats have to have an acronym. It’s a federal law.

Kind of ugly, but ultimately it works. We know this because the folks who publish Total Baseball have their own handy-dandy statistic, called Fielding Runs. Without getting too detailed, Fielding Runs (FR) calculates the numbers of runs each player contributed to his team each year based solely on defense. That math won’t be replicated here, just trust that they’ve done a ton of work to make FR a meaningful measurement of defensive value.

By comparing the highest FR scores each year we find that they track pretty closely with the simplistic DSM calculations described above. Looking at all American League seasons from 1957 through 1998, for 28 of those 42 years, the Gold Glove winner determined by DSM matched with the winner determined by FR. In eleven of the remaining fourteen seasons, the winner in DSM finished second or third in FR. So there is agreement two-thirds of the time and near-agreement for about 80% of the remainder. There is only one true outlyer, the 1970 season, during which FR favors Roy White while DSM favors Felipe Alou, who was a subpar outfielder (-3 FR) according to Total Baseball that season. Hey, defensive statistics are notoriously nebulous, so if there had been complete agreement between the two methods, someone probably would have uncovered it long ago.

Still, as a whole, we’ve got a pretty good basis for determining who deserved a left field Gold Glove in those years. To break ties between the two methods, the player with the best general reputation as a defender was chosen as the winner. Those winners were:

  • 1957 Charlie Maxwell
  • 1958 Bob Cerv
  • 1959 Minnie Minoso
  • 1960 Minnie Minoso
  • 1961 Rocky Colavito
  • 1962 Rocky Colavito
  • 1963 Carl Yastrzemski
  • 1964 Chuck Hinton
  • 1965 Bob Allison
  • 1966 Carl Yastrzemski
  • 1967 Carl Yastrzemski
  • 1968 Carl Yastrzemski
  • 1969 Lou Piniella
  • 1970 Roy White
  • 1971 Carl Yastrzemski
  • 1972 Lou Piniella
  • 1973 Johnny Briggs
  • 1974 Johnny Briggs
  • 1975 Roy White
  • 1976 Larry Hisle
  • 1977 Carl Yastrzemski
  • 1978 Jim Rice
  • 1979 Willie Wilson
  • 1980 Rickey Henderson
  • 1981 Willie Wilson
  • 1982 Gary Ward
  • 1983 Gary Ward
  • 1984 Jim Rice
  • 1985 Mickey Hatcher
  • 1986 Jim Rice
  • 1987 Jose Canseco
  • 1988 Dan Gladden
  • 1989 Rickey Henderson
  • 1990 Dan Gladden
  • 1991 Tim Raines
  • 1992 Brady Anderson
  • 1993 Albert Belle
  • 1994 Brady Anderson
  • 1995 Marty Cordova
  • 1996 Tony Phillips
  • 1997 Garrett Anderson
  • 1998 Albert Belle

There were three tough judgement calls in there that went against Roy White, Ben Oglivie and Tim Raines, respectively, but there they are. Notice how much reputation comes into play in the actual voting. Of the few left fielders who actually did receive multiple Gold Gloves in real life, all of them had their totals decreased using these methods. Yastrzemski and Minoso each lost one award, while Rudi and Winfield lost all of theirs.

You’re probably saying, "Yeah, but how good could these guys be compared to REAL Gold Glove winners? Center fielders and right fielders are just more valuable as defenders, so who cares if these guys got snubbed?" Well, these guys actually compare very well. The real, live Gold Glove recipients average just about 9 Fielding Runs in their awarded seasons. The left fielders above average just about 13 FR in these seasons, so who was more valuable defensively to their respective teams?

Of these forty-two seasons, just seven were actually awarded with a Gold Glove – Minoso in 1959 and 1960, and Yastrzemski in 1963, 1967, 1968, 1971, and 1977. The rest of these men were essentially stiffed by the lords of baseball, with the result that many of them are generally remembered as bad defenders or as slow designated hitters.

It’s simply not true. Under a fair awards system, each of these men would have likely received one or more Gold Gloves. In fact, in Jim Rice’s case, he would have received three Gold Gloves, tied for the second most among all American League left fielders, trailing only Yastrzemski’s six. Due to a silly rule and an unfair reputation, he received none.

Since defensive statistics are still evolving, it makes sense to also examine the measurement that has evolved the most. In 2002, Bill James published his latest work, Win Shares, in which he claims to have developed the most accurate method for measuring defensive prowess yet. In his book from the previous year, "The New Bill James Historical Baseball Abstract", he went so far as to claim that Win Shares measures defensive contributions, "vastly better than any previous analysis." In general, sabermetricians seem to agree with him.

Again, without explaining all of James’ math, Win Shares works like this – the players on every team in history are awarded shares of each of the team’s wins. Each win can be divided into three shares, so a team that wins 100 games has 300 Win Shares to divide among its roster. James has devised a method of allocating those shares based on each players’ offensive, defensive, and, if applicable, pitching performances.

During his career, Jim Rice accumulated 35.6 Win Shares for his defensive play. That compares well to other outfielders, particularly corner outfielders. For instance, nineteen different retired Gold Glove winners posted lower career totals. To be fair, Rice played a relatively long career, so his raw total of defensive Win Shares should be higher than most players’ totals. A more fair comparison would be to examine his Win Shares per 1000 defensive innings played. Rice posted 2.63 Fielding Win Shares per 1000 innings in the field. Thirteen different Gold Glove-winning outfielders, men who combined to win 36 outfield Gold Gloves, posted equal or lower figures:

  • Hank Aaron 2.63
  • Dave Parker 2.63
  • Sixto Lezcano 2.62
  • Frank Robinson 2.58
  • Roger Maris 2.56
  • Ellis Valentine 2.48
  • Rickey Henderson 2.42
  • Joe Rudi 2.41
  • Wally Moon 2.37
  • Barry Bonds 2.26
  • Tony Gwynn 2.26
  • Jay Buhner 2.04
  • Dave Winfield 1.92

Another group of 15 Gold Glovers finished just ahead of Rice:

  • Carl Yastrzemski 2.65
  • Raul Mondesi 2.66
  • Dusty Baker 2.70
  • Al Kaline 2.70
  • Jackie Jensen 2.72
  • Tony Oliva 2.76
  • Minnie Minoso 2.76
  • Rick Manning 2.79
  • Bobby Murcer 2.79
  • Dale Murphy 2.80
  • Bobby Bonds 2.81
  • Dwight Evans 2.82
  • Pete Rose 2.85
  • Roberto Clemente 2.87
  • Al Cowens 2.90

James took the extra step of assigning letter grades to each player for their defensive prowess. Rice received a C+, which certainly appears to be only slightly better than mediocre until we examine the others who received the same or similar grade. Carl Yastrzemski and his seven Gold Gloves playing the same left field as Rice also scored a C+. Dwight Evans and his eight Gold Gloves across the Fenway outfield scored a B-. Hank Aaron and Frank Robinson were C+ defenders. Roberto Clemente was a B-. So was Al Kaline. We’re talking about perhaps the finest corner outfielders, in terms of defense, in the game’s history.

In fact, very few corner outfielders rank higher than the C+/B- range because James lumped them in with center fielders when passing out grades. Obviously, center fielders are of higher defensive value because more balls are hit to them. James asserts, and there is really no reason to doubt him, that the data simply wasn’t there to allow the corner outfielders to be separated. Fair enough. But what that means is that the scale for corner outfielders’ grades has to be shifted downward a bit. Those in the D+/C- range were probably average. Anyone in the C/C+ range was a good defender, anyone in the B-/B+ range was outstanding, anyone higher was a freak of nature.

That scale would hold to form if we examine just the left fielders currently in the Hall of Fame.

  • Al Simmons - A
  • Fred Clarke - A-
  • Joe Kelley - A-
  • Stan Musial - B
  • Jesse Burkett - B
  • Joe Medwick - B-
  • Ed Delahanty - B-
  • Zack Wheat - B-
  • Carl Yastrzemski - C+
  • Goose Goslin - C+
  • Jim O’Rourke - C+
  • Ted Williams - C
  • Billy Williams - C
  • Chick Hafey - C
  • Heinie Manush - C
  • Willie Stargell - C-
  • Ralph Kiner - C-
  • Lou Brock - C-


Only eight of the eighteen posted better defensive grades than Jim Rice, with seven ranking lower and three being tied with a C+. So, in short, Jim Rice would be an average defensive left fielder compared to others in the Hall of Fame. That’s not a bad label to have.

Of course, this fact didn’t stop Bill James from making false statements about Rice’s defense either. James took a great deal of time and space in his comments about Roy White to demonstrate why he felt White was a better overall left fielder than Jim Rice. When he compared them as defenders, he was very short on facts and very long on opinion. With just one sentence, he judges the competition in favor of White, with no supporting documentation whatsoever.

"Jim Rice wasn’t a bad outfielder, but Roy White obviously was better."

It might be obvious to James, but it certainly isn’t obvious to the Total Baseball authors. Their statistic for measuring defense, Fielding Runs, ends in a dead heat. Each man compiled 71 Fielding Runs in his career. Since Rice played slightly fewer outfield games, he actually accumulated a better rate of Fielding Runs per game played. To put those 71 Fielding Runs in context, only twelve Hall of Fame outfielders have exceeded that total. In addition, Rice’s figure of 0.046 Fielding Runs per outfield game played is bettered by just seven Hall of Famers – Richie Ashburn, Max Carey, Roberto Clemente, Willie Mays, Kirby Puckett, Tris Speaker, and Carl Yastrzemski. It’s a who’s who of outfield defense.

The distinction between Rice and White shouldn’t be obvious to casual observers using traditional statistics either. In fewer games and fewer outfield plays, Rice threw out almost 60% more baserunners than White and participated in more double plays. He totals nearly as many putouts despite playing in the smallest left field in baseball, one with literally no foul ground, while White played in one of the largest. His putout totals were further depressed by the fact that the Red Sox usually had a pitching staff dominated by right-handed pitching, again due to the perceived vagaries of Fenway Park. During Rice’s full career, less than one-quarter of all innings thrown by Red Sox pitchers were accounted for by lefties, which of course resulted in fewer balls being hit toward Rice in left field. He did make more errors and consequently had a lower fielding percentage (.980 to White’s .988), so a review of the facts might indicate that the two players had equal defensive value. It certainly wouldn’t indicate that either player was "obviously" better.

And, in James’ case, it really shouldn’t have been obvious because he had just developed a defensive evaluation tool that was, in his own words, "vastly better than any previous analysis." It was this analysis that he used to develop his rankings, so obviously he had the results in hand when he claimed White was a better defender than Rice. Well, according to Win Shares, he wasn’t. In fact, the opposite was true. While the previous methods for evaluating defense ended in a dead heat between the two, James’ method gives Rice the advantage. Rice had a letter grade of C+ while White got a C. Rice totaled 2.63 Win Shares per 1000 innings in the outfield. Roy White totaled just 2.44.

So, you see, even Bill James, paragon of baseball research and analysis, isn’t above making statements that fly in the face of facts – even when he’s the one who analyzed the facts. More often than not, when it comes to commentary on Jim Rice’s defensive abilities, those statements unfairly chip away at his reputation and his claim to a place in the Hall of Fame.

1 comment:

Anonymous said...

I think outfielders are underated by the Bill James types because maybe halve the times they handle the ball it is not considered a chance because noone is put out while when an infielder cleanly fields a ball and makes a good through 90% of the time someone is out. Still many infield assists are as routine as an outfielder fielding a base hit and tossing to the cutoff man. A good outfielder can turn potential doubles into singles and I think this still falls between the cracks of our beloved statistics gurus like Bill James--Geoffrey Feldman