The Games That Broke Retrosheet

By Tom Ruane

Back in 2001, when I started developing the software to display box score and play-by-play data on our web-site, I thought I knew how major league baseball was played and so I made assumptions. Most of these seemed pretty straight-forward: nine fielders, three outs to a half-inning - that sort of thing. Of course, there are several other applications designed to parse Retrosheet data and I'm sure many of those handled the games described below with little or no difficulty, but what follows are games that caused my software to misbehave. In many cases, the only way I was able to made peace with these games was to add code that read like "IF (this game is that weird one in 1949) THEN forget everything you thought you knew about the rules of substitution."

Let's start at the beginning. Or at least close to the beginning. On July 5, 1872, the Cleveland Forest Citys were in Brooklyn to take on the Atlantics and were traveling with a squad of only ten men. According to the New York Times: "The Forest Citys went through the game with only eight men, Pratt being ill and Wolters absent for some reason peculiarly his own."1 Playing with only two outfielders, they lost 10-3. Cleveland was also obliged to play with eight the next day when they faced the Brooklyn Eckfords, but this time the under-manned team routed the locals 24-5. We do not have all of the National Association box scores (we are missing 1873 and 1875), but there was only one other league game in our collection that featured an eight-man team: on July 10, 1874, the Mutuals' first-baseman Joe Start left the team suddenly, requiring the visitors to take on Hartford short-handed. They lost 13-4.

The next group of troublesome games came about due to something that was allegedly in short supply during the rough and tumble brand of baseball played in the old days: courtesy. Courtesy runners and fielders was the practice of allowing what would ordinarily be illegal substitutions as long as both teams agreed to permit it. It was outlawed in 1950, but before (and in one case, after) that, they resulted in some curious results.

The first example I encountered occurred during the second All-Star game in 1934. Back in the early days of the Midsummer Classic, the rosters were relatively small. Instead of the 80-100 players who normally squeeze into each dugout these days (making each inning resemble a presidential bill signing more than a game of baseball),2 each team had only twenty players at their disposal in 1934. So when Frankie Frisch sprained his foot in the fifth inning, the only player on the National League's bench was pitcher Fred Frankhouse. Billy Herman, who had pinch-hit for Carl Hubbell in the third inning, was allowed to re-enter the game to replace Frisch.

There was a related game eighteen years later. Once again, a player re-entered the game, but this time there were a few important differences. First of all, this was a regular season game, not an exhibition. And more importantly, the game came two years after the practice has been expressly prohibited. Nevertheless, when catcher Clyde McCullough was injured making an out in the bottom of the eighth inning, Ed Fitz Gerald, who had pinch-hit in the sixth inning, was allowed to go behind the plate for the top of the ninth. Admittedly, neither team was in a pennant race at the time, but the game was close and the Cubs' decision to permit the substitution increased Pittsburgh's chances (albeit slightly) of winning the game.

Both of the games above involved courtesy fielders, but courtesy runners were far more common. The first ones I stumbled upon (and remember Retrosheet typically adds games working backwards in time) came from 1949. On June 14, 1949, Cleveland's Lou Boudreau was hit by a pitch in the top of the first inning and the Red Sox allowed the Indians a courtesy runner. One weird thing is that, rather than some player off the bench, the substitute runner was third-baseman Ken Keltner, who had just scored on the grand-slam home run that preceded (and perhaps caused) Boudreau's plunking. Keltner would eventually score his second run of the inning. So now you know the answer to the question: how can a player score two runs in an inning if only ten men come to the plate?

Cleveland would get the major league's last courtesy runner less than a month later, when on July 2nd, Ray Boone was hit by a pitch in the bottom of the ninth inning. Again, the Indian chose a player who was already in the game, catcher Jim Hegan, and again, he would score a run. Since the game ended after nine innings, it would not have been considered a courtesy runner had Cleveland chosen a player who wasn't or hadn't already been in the game.

Retrosheet's web-site contains an article describing all of these courtesy fielders and runners (and many more), but I did want to mention one more game that caused my software to misbehave: the June 9, 1922, game between the Yankees and White Sox. In the sixth inning that day, New York catcher Wally Schang was hurt sliding into second and was replaced temporarily by courtesy runner Al DeVormer. So far, nothing too unusual. But in the eighth inning, Schang hit a one-out single and DeVormer was sent in to pinch-run for him for a second time. It is the only instance so far in our released event files in which a player pinch-runs for another twice in the same game.

That day is also notable as it is the only one we've discovered so far with two courtesy runners. The Athletics' Jimmy Dykes was hit on the head by a Syl Johnson pitch in the second inning of their game that day, knocking him unconscious. Fred Heimach was permitted to run for him and by the end of the inning Dykes had recovered enough to re-enter the game, ensuring his eventual enshrinement in the Tough Guy Hall of Fame.

The next group of games fall under the heading of Umpire Mistakes.

The first took place on May 14, 1938, and if you look at the box score, you'll notice something very unusual: the game went ten innings despite the fact that the Reds were ahead 6-5 after the regulation nine. The reason: Dusty Cooke's sixth-inning home run was ruled a triple on the field. The Reds protested the call, the protest was upheld and a 7-6 ten-inning loss became a weird 7-7 tie. You might ask why the game wasn't simply replayed from the point of the protest, with Dusty Cooke credited with a home run and Cincinnati ahead 6-1 in the top of the sixth? I suppose that the league president felt that wiping out the Cards apparent game-tying four-run rally in the bottom of the ninth would have been unfair. As it is, not only does the game go into extra-innings with the Reds holding a one-run lead, but the bottom of the tenth ends with no one out following Enos Slaughter's game-tying two-run home run (which appeared to all the fan's in the park that day to be a game-winning walk-off blast).

An even weirder game, also involving the Cardinals, was played on July 20, 1947, and also centered around a disputed home run. With two outs in the top of the ninth inning, Ron Northey hit a long fly ball to center field that either did or didn't leave the park before bouncing back into play. The third-base umpire ruled it a homer, causing Northey to slow up as he rounded the bases. Unfortunately, the first-base umpire ruled it in play, and as Northey coasted toward home plate, the ball was waiting for him. So instead of a 3-0 lead (with perhaps more to come in the inning), St. Louis headed into the bottom half of the inning with a two-run lead. That lead disappeared as the Dodgers proceeded to rally, and with two outs they (apparently) won the game 3-2. When the Cards won their protest, NL President Ford Frick ruled the game a 3-3 tie, giving us the odd result of a game in which both teams finished their halves of the ninth inning with only two outs. Again, the game was probably simply not suspended at the point of the protest because such a ruling would have unfairly disadvantaged the home team, who would have seen their rally wiped out.

The next game doesn't involve a disputed home run, but instead a misunderstanding of the rule governing batters hitting out of order. On September 24, 1964, Cubs' manager Bob Kennedy decided to give Ernie Banks the day off and so posted a lineup in the dugout with rookie John Boccabella playing first. While filling out the official version, however, he got distracted, habit kicked in, and he wrote Banks' name on the lineup given to the umpires. Dodgers' manager Walter Alston noticed the error when Boccabella took the field in the top of the first, but waited until the time was right to point it out. That time came in the bottom of the sixth inning when Boccabella tripled to score a run, cutting Los Angeles' lead to a single run. Alston complained that he had hit out of order and the umpires agreed, ruling Banks (the proper batter) out, and ending the Cubs' rally.

The only problem was that the umpires were completely wrong. The only thing the Cubs had been guilty of was an unannounced substitution: once Boccabella took the field in the top of the first, he replaced Banks in the lineup and so for the rest of the game was hitting in the proper place. Kennedy wasn't aware of the mistake and so failed to protest, something which would have been moot anyway when the Cubs rallied to win the game. So just about the only consequence of the series of mistakes is that Banks got charged with an at-bat in a game in which he didn't appear. Officially, he started the game at first, was replaced immediately, somehow re-emerged to make an out in the sixth inning, and then left again.

In 2008 there was an incident involving an umpire error that was (for the most part) corrected. It took place on September 26, 2008, less than a month after baseball ruled that teams could appeal home run calls. The fun began in the bottom of the sixth when Bengie Molina hit a ball that was ruled a non-homer on the field. Giants manager Bruce Bochy sent in a pinch-runner for Molina before realizing that the call may have been incorrect. He then appealed the decision and, upon further review, Molina was credited with a home run. But not, in the opinion of the umpires, with a run scored. Despite the fact that Molina was only taken out of the game because of the blown call, the umpires ruled that the substitution could not be reversed. And so officially, Molina hit a home run and was replaced by a pinch-runner as he rounded the bases.

This is not the only example in our files of a pinch-runner appearing in the middle of a play, but the only case in which the batter was taken out while rounding the bases on a homer.

The last game in the section on umpire error is also the most recent one. And in this case, the umpires can probably be forgiven for their mistake because it took the major league baseball rules experts nearly a week to figure it out. On April 19, 2013, Juan Segura was caught off second base, kicking off a run-down that ended with both Ryan Braun, who had been on first, and Segura, standing on second. In the confusion, Segura mistakenly thought he had been tagged out and started jogging off the field, toward first base. Realizing what had happened, the first base coach yelled for Segura to run to first, where he was allowed to remain. This is only case so far in our files in which a base-runner moved clockwise around the bases. And it turns out that the umpires shouldn't have allowed the retreat, since the comment to rule 7.01 trumps the comment to rule 7.08(i).

By the way, when we get around to processing the event files for the 1911 American League, we will see this again. On August 4, 1911, Germany Schaefer stole second and then went back to first. The game was scoreless in the bottom of the ninth inning and Schaefer was attempting to give the runner on third, who represented the winning run, a chance to score in the resulting confusion. It didn't work, but it made Schaefer sort of famous.

Next up, two games involving another type of appeal, one occurring in 1975 and the other in 1980. In both cases, a pitching change was made and then a successful appeal was made that a runner on the previous play should have been called out for missing a base. Officially, the new pitcher gets credit for the out (the third of an inning pitched) while the previous pitcher gets credit for facing the batter, something my software was (and still is) very reluctant to do.

The last section involves my attempt to establish limits on how often something in a baseball game could happen. I know it's a good practice to avoid setting arbitrary limits like this. I really do. But sometimes the temptation to do something quick and dirty is overwhelming. After all, how often could a player change positions in a game? Well, on a few occasions, the first in 1965, a player (like Bert Campaneris) has played all nine positions in a game. So a limit of sixteen positions in a game seemed like more than enough.

At least until Toronto manager Jimy Williams decided to try to get by without a real second-baseman during the first of week of May in 1988. In one game, he alternated Kelly Gruber and Cecil Fielder between second and third, depending upon the batter's tendencies to hit to right or left field. This resulting in eighteen position changes for the two players. And it wasn't exactly a new strategy. When Retrosheet worked their way back to 1959, I discovered that on June 27th and June 28th of that year, Cleveland manager Joe Gordon did much the same thing with Woodie Held and Granny Hamner, this time shuttling them between second and short, sixteen times in one game and seventeen times the next.

The final case of a poorly chosen program limit was not foiled by a managerial strategy, but rather by a succession of managers (and others) getting really pissed off. When we started adding ejection information to our box scores, I decided that I would never have to handle more than sixteen of these in a single game. Of course I was wrong, but I was almost right, being foiled by a single game, one in which (you guessed it) seventeen players, managers and coaches were tossed. Second-year home-plate umpire Steve Rippley had ejected only two in his career prior to that game, but the seventeen he sent on their way that day are more than any other umpire has tossed in a full season since 1950.

And in case you were wondering about games in which a fed-up umpire "cleared the bench," those aren't really ejections (sorry Bill Sharman) and the players so cleared are allowed to remain in the game (they just can't stay on the bench when they aren't playing).

So that's it. Of course, this isn't a complete list of troublesome games. As we add play-by-play accounts for nineteenth century contests to our web-site, I suspect we will run into quite a few more examples of games that don't meet our expectations of how civilized baseball ought to be played. As a matter of fact, I'm looking forward to them.


1"Defeat of the Forest Citys by the Atlantics - Score 10 to 3," The New York Times. July 6, 1872. Page 8.

2This is an exaggeration: I'm pretty sure each team's roster today is no more than fifty or sixty.