In 1902, 2.1% of all games played ended in ties, 24 out of 1117 games.  Nine of these tie games were second games of double headers.  Eight of these second games occurred in September. 

Here is a 439-word synopsis of Clifford Blaus article lifted from his direct quotations including his subtitles.  I recommend reading the fullness of Mr. Blaus 3,525-word story.   (Robert Bofors)

http://mysite.verizon.net/brak2.0/tiegames.htm


The History of Major League Tie Games by Clifford Blau

HOW COMMON WERE TIE GAMES?

From 1876 to 1917, on an annual basis, tie games generally comprised between one and three percent of all games played. The most tie games in a single season was (sic) thirty, in 1907. 

About 45% of tie games were extra-innings, and the average length of a tie game was about 9.7 innings. The longest tie game was twenty-six innings. The shortest was one out longer than 4  innings, on Sept. 16, 1906 between St. Louis and Cincinnati. 

WHY DID TIE GAMES OCCUR?

The primary reason that tied games have been stopped is darkness. Through 1938, before most teams were hosting night games, about three-quarters of tie games were called due to darkness... Consequently, tie games have occurred most frequently in September, October, and April, when sunset occurs early, and least often in June and July. Also, second games of doubleheaders have seen more than their fair share of draws. 

The second most common reason for ties is rain. Rain accounted for about 10 to 15 percent of tie games up to 1938.

It has frequently been noted that baseball is a game without time limits. However, this has not always been so. In the days before teams traveled by airplane, they would often enter into agreements before the final game of a series. These agreements provided that the game would end by a specified time so that one or both teams could catch a train to a distant city for their next series. Such agreements caused about 10 percent of tie games until the 1950's. (Actually, they were rare in the 19th Century, when fewer than 154 games were played in most seasons, leaving more off days for travel.) This means that the teams believed that their next game was more important than the one they were playing. 

Still another reason for drawn games, and another example of how fans have been cheated, occurred several times. For many years, most doubleheaders featured morning and afternoon games with separate admissions. Tie games arose several times because the morning game of a doubleheader took longer than expected and management wanted to clear the park to admit the ticketholders for the second game. It was somewhat like going to a movie and being forced to leave so the next showing can begin, although the film hasn't ended yet. 

A Clifford Blau tidbit hid in his footnotes:

I was surprised to learn, in the course of researching this article, how frequent forfeits were in the 19th Century. Many times a team would simply refuse to continue playing when it disagreed with the umpire's decision. This is another way that the "cranks" were cheated. 

