(but in an interesting way in the AL). As the opportunity cost of time falls, chances of attendance increase (Friday and weekends). Chances of attendance rise with population; apparently, scale effects outweigh the larger substitution variety in larger cities. Baseball attendance is an inferior good in the AL ( Noll, 1974, found this, as well) but not in the NL. In fact, for the NL attendance is a normal good in 1990. Whether or not play against a division opponent effects attendance chances appears to depend on the season. People usually like to see home teams that are winners and visiting teams that are less successful. The evidence on games between evenly matched teams is mixed, but a game against a more evenly matched division opponent decreases the chances of attendance (an unexpected outcome). Fans do not like old stadiums. Playing at home in the east increases chances of attendance. Attendance chances are inelastic with respect to stadium capacity. Attendance usually picks up as the season progresses. There is evidence for one year that double-headers increase the chance of attendance. Finally, some teams definitely are a visiting draw ( Boston and Oakland, compared to Toronto, in the AL, and Chicago, Cincinnati, Los Angeles, New York, Pittsburgh, and San Francisco, relative to Atlanta, in the NL).
The punch-line for this chapter concerns the impacts that streaks have upon attendance chances. Winning streaks for both the home and visiting teams, when they matter, almost always increase the chance of attendance. The only exception is in the NL, 1989, where home winning streaks do not matter but visiting team win streaks lower attendance. Losing streaks do not matter as often as winning streaks, but there is evidence that they decrease the chance of attendance regardless of whether the home or visiting team has hit the skids.
These results imply that streak management is valuable. Fans are more (less) likely to attend when their team is on a roll (has hit the skids) and gate revenues can be increased through effective streak management. The goal is clear. Win in streaks at home. If losing streaks occur, see to it that they occur while your team is on the road and try to end any losing streak at home.
In addition to the resource misallocation that follows when streaks are too long for streak managers of poor teams (found in the previous section), we speculate that the misallocation during the season may be even worse. The setting seems ripe for a lemons problem. First, it is difficult to determine whether or not a manager is a streak manager just by observing their behavior. Second, especially since even some streak managers run streaks too long for poor teams, it is easy for any other manager to claim responsibility for streaks that do occur and blame the cruel world of "bad breaks" for any unhappy outcomes. This seems especially problematic in pro sports, where every contest is "a game of inches" and wins and losses can turn on twists of fate. To the extent that streaks pay and no market mechanism protects owners, we suspect that any manager can claim to be a streak manager and misallocate talent on their team during the season.
In MLB, by and large, streaks are random, but some managers appear to be "streak managers." While we have not established how or why a given manager is a streak