April 1, 1997

Simulating Away Baseball's Ennui

By PETER WAYNER

he best defense may be a good offense -- unless you happen to be Major League Baseball. For the last several years, both fans and owners have expressed concern that the games were lasting too long because the players were hitting too much, bringing more players to the plate and delaying the nation's bedtimes.

A Virtual Ball Game The Simulator Test how different slugging percentages and times spent at bat affect the length of a baseball game with the baseball simulator. FAQ How do the simulated players hit? What's wrong with this approximation? What about errors? Check the simulator FAQ. The Code Explore the Java code that runs the simulator. Last Time For more on computer simulations, see New-Media Tools for Online Journalism, published Oct. 9, 1996.

A new computer simulation, however, suggests that the number of hits is far from the most important determination of game length -- and that what must be reigned in is the average time spent at the plate. This is good news for the game because some owners and officials worried that only low-hitting scores of 1-0 would get people home in time for school or work the next morning.

The numbers are startling. Today's average game is now 30 minutes longer than the average game of 20 years ago. One game in last year's World Series was the longest on record. At the same time, hot-hitting teams like last year's Baltimore Orioles filled the stands by breaking records for the most home runs hit by a team in a year. Many assumed that there had to be some relationship between hot bats and long games.

Identifying the symptoms is easy, but finding a way to fix them can be difficult. Two years ago, in response to a request by the owners, the former umpire Steve Palermo issued a set of proposals for speeding up the game. Some are procedural changes to tighten up action by cutting the dead periods in the game. Pitchers used to get 20 seconds to throw the ball after the batter was set. Now they get 12. Batters were also forbidden to take long time outs and stray from the batter's box.

But other proposals were more radical and aimed at changing the flow of the game. Palermo suggested enlarging the strike zone and raising the pitcher's mound from 10 inches to between 12 and 13 inches. Both of these give the pitcher more of an advantage and would probably lead to lower batting averages. After the 1968 season, the mound was lowered from 15 inches to 10 inches and the averages soared from .230 in the 1968 American League to about .260 in recent years.

These more fundamental proposals send shivers up the spines of many fans -- and send marketing agents into convulsions. Hot bats bring excitement. Do fans really want a shorter game if it means sacrificing runs?

But how much does the batting average affect the length of the game? This article comes with a simulator that runs through numerous hypothetical ball games following mythical players around the bases and timing the results. You can set how much time is spent per batter and then vary the hitting.

The answers can be surprising. Imagine that each batter takes one and half minutes, on average, if there is no runner on first with a chance to steal second, and two and half minutes if the pitcher must throw to first to keep the base runner in line. If the hitters get on base at the same rate as the 1968 American League (season batting average .230), then the games take about two hours and 48 minutes on average. But if the simulated players hit according to the 1983 American League's statistics (batting average of .266), then the games took an average of two hours and 52 minutes -- four minutes longer.

This suggests that the batting average is not a very large factor in the length of the game. In fact, in one simulation where the team batting average was set to.394, the average game only increased to three hours and 13 minutes. Some of these games had more than 50 total hits between the two teams. The latest evidence to this effect was the seventh game in the 1996 National League playoffs. The Atlanta Braves won with a score of 15-0, but the game only took two hours and 25 minutes.

The average time that each batter spends at the plate is much more important. At least 52 players must come to the plate in a minimal game where the home team wins by one home run. Adding 10 seconds to an average trip adds at a minimum of eight minutes to the time of the game. But many games may have 80 or more trips to the plate. Ten more seconds for an average trip to the plate adds about 13 more minutes to the game.

This simulator can't factor the length of the trip to the plate. Each batter in the simulator comes to the plate and after a set amount of time emerges with either a hit, a walk or an out, based upon the statistics used to set the odds. The only attempt to add more realism occurs when a player is on first with a chance to steal second, in which case some additional time is added to account for the pitcher's attempts at throwing the runner out.

A great simulator would be able to explore the dynamic tension between the pitcher and the batter. This one doesn't attempt it. Bruce Markusen, a researcher at the Baseball Hall of Fame in Cooperstown, says that the number of balls and strikes makes a big difference to the game. A larger strike zone or a higher mound could give the pitcher the advantage to get ahead in the count by throwing two strikes. When this happens, he says, the batter can't be choosy and wait for the perfect pitch. That means less pitches are thrown and the batters are retired faster.

This observation is also backed up by recent facts. The National League umpires are known for giving the pitcher a more generous strike zone, and this was reflected in the games played for the league championship. The American League games averaged 310 pitches per game, while the National League games (despite the Atlanta Brave's penchant for running up the score) averaged 261.

So baseball can have its cake and eat it, too. Many people greeted the proposals of a higher mound and a larger strike zone with fear. A 1-0 final score sounds more like a soccer match than the type of slugfest that produces challenges to Roger Maris's record.

But baseball can't have everything. Many modern writers have produced long-winded, lyrical tributes to the game that invest the batter-pitcher duel with almost mystical qualities. To sports writers, these players are not just two guys who are lucky enough not to need a day job. Now, the pitcher contemplates his move with Zen-like concentration. The batter plans his strategy with the iron will of Napoleon and the brilliance of Gary Kasparov. When people step to the plate with such ideals in their minds, the average time at bat can only soar. Maybe, the solution for cutting the time of the game is to cut the hype as well.