June 18, 2007
Watching the Detectives
Judging the Field of Field Judges
This is the second part of a series on umpires. Read the first part here.
"Any umpire who claims he has never missed a play is...well, an umpire."
"They expect an umpire to be perfect on Opening Day and to improve as the season goes on."
Don Denkinger umpired 31 years in American League, tying the league record (since broken) for longevity. His career is remembered for one call, a call he got wrong. It is a call so monumentally bad that it placed first in the worst calls of all time according to ESPN.
You know the call--Game Six of the 1985 World Series, the 'I-70 Series,' with the Cardinals ahead of the Royals three games to two. With the Cardinals leading 1-0 in the bottom of the ninth, Cardinals closer Todd Worrell faced pinch-hitter Jorge Orta, the leadoff hitter in the Royals' half of the inning. Orta hit a slow roller to St. Louis first baseman Jack Clark, who tossed to Worrell covering first for the apparent first out. However, Denkinger called Orta safe on a play that replays, photos, commemorative mugs, and Broadway shows have documented clearly that he was out, way out. What followed is a matter of history, as the Royals' comeback was capped by Dane Iorg's pinch-hit two-run single to right to win the game.
Denkinger drew the straw to umpire behind the plate in Game Seven. Amid the 11-0 pasting, Denkinger threw out Cardinals pitcher Joaquin Andujar and manager Whitey Herzog, solidifying his position in umpire purgatory among Cardinals fans. He also allegedly drew the ire of then-commissioner Peter Ueberroth, who waited for Denkinger outside of the umpires' room, and who responded when asked by Denkinger if he had gotten the call right, "No, you didn't." Thanks, boss.
Only slightly less famous was the impact of another field judge on Game One of the 1996 American League Championship Series. Umpire Rich Garcia drew the right field role--a cushy assignment for any ump, right? Well, not exactly: in the bottom of the eighth with the Yankees trailing 4-3 and one out, Derek Jeter lifted a ball to right field. Orioles right fielder Tony Tarasco camped under the ball and set up for the catch. As Garcia ran out to the right field wall, twelve-year-old fan Jeffrey Maier reached over the wall and deflected the ball into the stands. Instead of calling the clear-cut fan interference, Garcia gave Jeter the cue to circle the bases. Tarasco, gesticulating wildly at the fan, could not believe the call; irate O's manager Davey Johnson got ejected for his response to it. The game remained tied until Bernie Williams hit a walk-off home run in the bottom of the eleventh. Garcia later admitted that the replays revealed he made the wrong call; there's no word on whether or not Bud Selig admonished him after the game.
In the first two games of the 1999 ALCS, the Yankees, playing at home, edged the Red Sox by one run in each game. When the series moved to Fenway, the Red Sox pounded the Yankees 13-1 in a seemingly legendary matchup that pitted Pedro Martinez versus Roger Clemens, setting the stage for an all-important Game Four. With the Yankees leading 3-2 in the bottom of the eighth after leaving the bases loaded in the top half of the inning, leadoff hitter Jose Offerman singled to right with one out, driving New York starter Andy Pettitte from the mound in favor of closer Mariano Rivera. Red Sox third baseman John Valentin then hit a dribbler to second base. Chuck Knoblauch grabbed the ball and feigned a tag of Offerman as he ran by, and then threw to first.
To everyone's astonishment, second base ump Tim Tschida called Offerman out, even though replays showed that Knoblauch was not even at the same latitude as Offerman--some say his tag missed by at least a few feet. Instead of Offerman on second with two outs and Nomar Garciaparra at bat, the inning and rally were snuffed out. Red Sox manager Jimy Williams was then ejected for his understandable outrage--notice a trend? In the top of the ninth, with angry fans littering the field with debris, the Yankees poured six more runs over the stunned Red Sox, the rally being spurred on by a Knoblauch single (of course) and his scoring on an Offerman error two batters later (natch). The Yankees sealed their series victory the next day, gliding to a 6-1 win behind Orlando Hernandez to set up a sweep of Atlanta in the World Series.
In his post game mea culpa, Tschida further grated on the Red Sox Nation when he innocently referred to Knoblauch as "Knobby," as if they were intimates and there was a fix in. Tschida's call was named the worst ever by a sports official in an ESPN readers' poll.
Life as a field judge once seemed simpler. According to Ron Luciano, it was "heaven." Now, we can tell if a throw to first beat a batter, if a base stealer evaded a tag, a liner hit the line or not, or if a long fly ball hit the foul pole for a home run or missed it becoming just a souvenir. And we can see it in super slow-mo, at multiple angles, and in HD with Tim McCarver highlighting and circling salient features in the teleprompter illustrated by smarmy commentary. Once the goats were players with names like Merkle, Snodgrass, and Owen; now, they are the field umps who end up with a bad angle or mishear the ball hitting the glove, trying to make the best call they can while often sprinting around the field. In the process, they wnd up the bugbears of the local fandom.
How do we know which umpires perform best and worst when they are not behind the plate? The rulebook doesn't even define their roles: "A field umpire may take any position on the playing field he thinks best suited to make impending decisions on the bases" reads Rule 9.04(b). It's like little league: the umpires can pick any position they want. One would assume that they take up the traditional positions at first, second, and third, and if there are a few extra umps-say, in a playoff game-they know to go to left and right field.
The field umps' roles are subordinate to and circumscribed by the plate ump, i.e., the game's umpire-in-chief, as defined in the rule Rule 9.04(b): "Make all decisions on the bases except those specifically reserved to the umpire-in-chief." If two or more field umpires have different opinions, then the plate ump will call a umpires-only conference and determine the correct ruling (Rule 9.04(c)). Actually, they are initially defined as a simple negation of the plate ump. "If there are two or more umpires, one shall be designated umpire-in-chief and the others field umpires." (Rule 9.03(b)) They do, however, get to share duties with the plate ump to call 'time,' balks, illegal pitches, or defacement or discoloration of the ball by any player."
That is all the authority that the rule book provides to the field umps. Apparently, originally only plate umps were employed, and as the others were gradually added, the responsibilities for each role was defined. I mean, what if the plate ump retained the shared responsibility of calling outs on the bases? On the Denkinger play, he would have made a concurrent call and potentially overruled the bad call. Or let's say that the second base ump backed up the first base ump on calls at first when the bases were empty. With the added angle, two different calls could have been made on the Denkinger play, and after a conference, the plate ump could have made the correct call.
The way that the roles are defined may be the better approach, but that is just not how the roles developed. Each field ump handles the calls for his particular quadrant of the diamond. Imagine how much longer and less interesting the game would be if there were a conference on every close play. Fixing the egregiously bad calls like Denkinger's does not outweigh the damage to game that shared roles would do.
That said, can we use statistics to evaluate the job that field umps do? I know that a number of the more infamous plays don't necessarily leave a statistical footprint that can be easily analyzed. For example, given that the Garcia call would either result in a home run or fan interference, something that is not tracked by an official statistic, there really is not much that can be revealed by statistical analysis of whether Gracia called more home runs than the average umpire in the handful of games which employed left or right field umps.
However, there are some calls that field judges make that can be evaluated using the statistical record. Even though some of the calls are attributed to the crew as a whole, looking at career trends can be illustrative of an individual's tendencies.
For first base umpires, we will take a look at career tendencies in calling balks, calling pickoffs at first (i.e., via stolen base percentage), and calling runners out on close plays (i.e., as part of a double play groundout). At second base, we will look at how he makes stolen base calls (via stolen base percentage) and how he handles calls at second on double plays. For first and third base umps, we will look at the umpire's tendency on balls hit down the line (via doubles and triples per plate appearances). Given that there are so few games for left and right field umpires, there were no meaningful trends that I could detect for these umps.
First Base Umpires
"Any time I got those 'bang-bang' plays at first base, I called 'em out. It made the game shorter."
The first recorded game in which a first base umpire was used (August 15, 1876) dates back almost as far as the first recorded game with a plate ump (April 22, 1876). On that afternoon in Lousville, Kentucky, John Morris, in one of four games that he umpired in the majors, became the first recorded first base umpire in the local Grays' 17-5 victory over the Cincinnati Reds.
First base umpires were used more and more regularly over time, but it was not until the 1910s that their use became universal. Here's a breakdown of games umpired at first by decade:
Decade Tot G w/FB ump Percentage 1870s 2,031 2 0.1% 1880s 8,758 166 1.9% 1890s 9,468 2,792 29.5% 1900s 11,342 4,003 35.3% 1910s 13,324 13,128 98.5% 1920s 12,323 12,316 99.9% 1930s 12,311 12,311 100.0% 1940s 12,376 12,376 100.0% 1950s 12,374 12,374 100.0% 1960s 15,961 15,961 100.0% 1970s 19,806 19,794 99.9% 1980s 20,337 20,337 100.0% 1990s 21,594 21,594 100.0% 2000s 17,002 17,002 100.0% Total 189,007 164,156 86.9%
As has been pointed out to me in the e-mails I've received since the first article, home plate umps are not the only ones that call balks--a good deal are called by first base umps. Below are the umps that had the highest balk occurrences while they were umping at first (adjusted for era):
When a player is leaning towards second and a pickoff throw comes from the pitcher (or catcher) are some umpires more likely to call the runner safe than others? Here are the first base umpires most likely to call the runner safe:
Finally, I wonder if the sentiments in the quote above are actually representative of the average umpire. Do some umpires call those bang-bang plays out just to shorten the game? Here are the umps who were most likely to call the trailing runner out at first on a ground-ball double play. Frank Wilson was twenty times as likely to call these runners out than the average ump (keep in mind that his career totals are incomplete):
Second Base Umpires
"Most plays that are missed by the umpire are caused by the umpire not reading those cues early enough and making the proper adjustments."
The first recorded use of a second base umpire occurred in a game in the one-year rival Players League in 1890. On July 14, the hometown Ward's Wonders from Brooklyn beat the Pittsburgh Burghers, 6-2, with Bob Ferguson in the second base role, Lon Knight at first, Bill Holbert at third, and Charley Jones behind the plate in the first game in recorded history to employ all four infield umps. It was also the only game for the next 29 years in which all four were used. The next occurrence was on July 6, 1919, in Cincinnati when Cy Rigler was the second second base ump. Second base umps were not employed regularly until the 1950s, and still there is the odd game in which one is not used:
Decade Tot G w/SB ump Percentage 1870s 2,031 0 0.0% 1880s 8,758 0 0.0% 1890s 9,468 1 0.0% 1900s 11,342 0 0.0% 1910s 13,324 1 0.0% 1920s 12,323 25 0.2% 1930s 12,311 148 1.2% 1940s 12,376 1,483 12.0% 1950s 12,374 10,108 81.7% 1960s 15,961 15,508 97.2% 1970s 19,806 19,261 97.2% 1980s 20,337 20,006 98.4% 1990s 21,594 21,429 99.2% 2000s 17,002 16,969 99.8% Total 189,007 104,939 55.5%
For second base umps, we want to look at each umpire's proclivity for calling runners at second. To do this, we look at tendencies to call base stealers out on stolen base attempts and to call base runners out on groundout double plays (both adjusted for league average). Here are the umpires who were most likely to call potential base stealers safe while umping at second:
Here are the umpires who were most likely to call a player out as the lead runner in a double play groundout:
Here are second base umps who are most likely to call any runner out at second, whether by a caught stealing or groundout double play (i.e., by adding the two together after inverting stolen base rate):
"Why is it they boo me when I call a foul ball correctly and they applaud the starting pitcher when he gets taken out of the ballgame?"
On August 13, 1888, Chick Fulmer, in his only game as an umpire that year, became the first man to be used as a third base ump in an American Association game between the old Baltimore Orioles and the then-Red Stockings in Cincinnati. Again, turning to the data, the use of third base umps was very slow in coming:
Decade Tot G w/TB ump Percentage 1870s 2,031 0 0.0% 1880s 8,758 5 0.1% 1890s 9,468 18 0.2% 1900s 11,342 15 0.1% 1910s 13,324 82 0.6% 1920s 12,323 6,057 49.2% 1930s 12,311 9,231 75.0% 1940s 12,376 11,388 92.0% 1950s 12,374 12,356 99.9% 1960s 15,961 15,961 100.0% 1970s 19,806 19,796 99.9% 1980s 20,337 20,337 100.0% 1990s 21,594 21,594 100.0% 2000s 17,002 17,002 100.0% Total 189,007 133,842 70.8%
For corner umpiring, we combine each ump's statistics at first and third base. The main job for a corner ump is to make safe-foul calls on balls down the line. If the umpire calls this ball fair it may end up a double or triple; if he calls it foul, it is just another strike--that is, if the batter has fewer than two strikes already called. Those are tremendously divergent outcomes in just one call. Below are the corner umps who were most likely to call doubles and triples (adjusted for era). I know that these calls include hits that were not necessarily down the line (especially the doubles), but it is the best we can do with the data available.
Here are the most likely to call a double:
And the most likely to call a triple:
Finally, there are the rarest of umpires--those who draw an assignment in the outfield. Outfield umpires were first used in a game on August 3, 1919 between the Giants and the Reds in-you guessed it-Cincinnati. I have no idea why the Queen City witnessed so many umpiring firsts. However, Barry McCormick was used in left and Pete Harrison in right to go with Hall of Famer Bill Klemm behind the plate and Bob Emslie at first on that somewhat historic day. (By the way, no umps were used at second or third.)
The first regular-season game in which all umpires were employed was on October 1, 1949, in Yankee Stadium (oddly not in Cincinnati) for a Red Sox-Yankees game. The team that day was Bill Summers behind the plate, Hall of Famer Cal Hubbard at first, former pitcher Eddie Rommel at second, Charlie Berry at third, Eddie Hurley in left, and Jim Honochick in right. With the Sox leading New York by one game with two to play, the game was handled like a playoff. The Yankees, who initially trailed 4-0 wound up winning 5-4 on a Johnny Lindell home run, won the league title the next day with a 5-3 Vic Raschi win, and went on to win the World Series in five games over the Dodgers. Oddly, no outfield umps were used in the clincher; maybe it was because the first game was on Joe DiMaggio Day.
Here are all the breakdowns for the outfield umps:
Decade Tot G w/LF ump % w/RF ump % 1870s 2,031 0.0% 1880s 8,758 0.0% 0.0% 1890s 9,468 0.0% 0.0% 1900s 11,342 0.0% 0.0% 1910s 13,324 1 0.0% 1 0.0% 1920s 12,323 0.0% 0.0% 1930s 12,311 0.0% 0.0% 1940s 12,376 1 0.0% 1 0.0% 1950s 12,374 26 0.2% 2 0.0% 1960s 15,961 159 1.0% 0.0% 1970s 19,806 1 0.0% 0.0% 1980s 20,337 0.0% 0.0% 1990s 21,594 4 0.0% 3 0.0% 2000s 17,002 0.0% 0.0% Total 189,007 192 0.1% 0.0%
In the next entry in this series, we will look at breakdowns by home/away team to determine if some umpires have a bias towards the home team in close calls.