Statistics the Statistical Analysis Shows

¶ … Statistics

The statistical analysis shows that the strongest correlation is between salary and attendance. There are two fundamental links between salary and attendance: the star power of the players and the ability of the team to pay a higher payroll. Higher salaries are associated with greater star power, which can be expected to have a high correlation with attendance. In addition, teams with a higher attendance will earn more revenues, and therefore be able to afford a higher payroll. This high level of correlation does not imply a causal effect, since there is a two-way feedback loop created by the elements of attendance level and payroll.

There are minor correlations noted with wins and stadium age and attendance. Wins was expected to be a strong correlating factor, since the competitiveness of the team increases its attractiveness to fans. This is especially true towards the end of the season, when teams still in playoff contention can expect a boost in fan support during the pennant race. The age of the stadium was not expected to have a strong correlation, but it does. The surprising element is that it appears as though the older the stadium, the higher the level of fan support. While McEvoy (2002) hypothesized that stadium age should be inversely correlated with attendance, this was not borne out by our analysis. The analysis indicated that some of the strongest attendance came from the oldest stadiums, such as Yankee Stadium, Wrigley Field and Angel Stadium. A more minor correlation, a negative one, was noted for ERA and attendance. This was speculated to be relevant because ERA is believed to be strongly correlated with winning.

It is worth noting that batting was not strongly associated with attendance. High-scoring, offensive baseball did not resonate with baseball fans in a statistically significant way. Stadium size also did not have a strong correlation. This is a critical insight because stadium size has often been viewed as a constraint for attendance, especially when a team has a high degree of popularity. In fact, our data shows that teams in smaller stadiums were more likely to have sellouts, even when that team had a poor won-lost record, such as the Pittsburg Pirates. Most teams fell well short of maximum capacity, indicating that for all but a small minority of teams stadium capacity was not a significant constraint on attendance.

The mean attendance in major league baseball in 2005 was 2,496,457.9 for a team, or 30,820.5 per game. There was no mode. The median attendance was 2,523,082 or 31,149 per game. The difference between the median and mean is 328.5 fans per game, which is only marginally significant (1.06% of the mean). The data is therefore considered to be well-dispersed. The standard deviation from the mean is 661,569.7 people, which indicates a fairly high degree of disparity in the results. The range has no skew.