Just looking for a good game: competitive balance in the Korean Professional Baseball League

ABSTRACT

Rottenberg’s uncertainty of outcome hypothesis is about preferences that can vary across fans and sports. We provide the first evidence of the empirical power of the hypothesis in the Korean Professional Baseball League (KPBL). In a panel data set, team-level aggregation over years shows that fans of this league attend in ways consistent with the hypothesis at the level of game uncertainty and consecutive season uncertainty, but only the first is statistically significant. KPBL fans appear to just be looking for a good game. This is consistent with the nearly complete concentration of post-season participation among a very few teams outside of the major population centre in Seoul and a lack of local team identification among KPBL fans. Our work adds to the diversity of league structures in the competitive balance literature since talent levels in KPBL are the result of parent company advertising choices rather than geographic identification. There are research questions and a policy implication.

KEYWORDS:

• Sports economics
• uncertainty of outcome hypothesis (UOH)
• competitive balance
• attendance
• baseball

JEL CLASSIFICATION:

• L83
• L80
• Z20

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

¹ Lee (2006) documents a clear effect on KPBL attendance when MLB games involving Korean stars are broadcast back to Korea.

² The panel we follow adds to the findings in other studies, cited in Section IV, which used either annual measures of league-wide competitive balance variables or uncertainty measures of daily games.

³ This section draws heavily on Fort, Kang, and Lee (2015).

⁴ The Korean Baseball Association governs amateur baseball separately.

⁵ Heroes are group-owned, much like an MLB team. The naming rights are sold by this owner group and have belonged to the Nexen tire company since 2009.

⁶ Lately, the bottom-line team profit may be getting less important for North American leagues as franchises become part of major real estate developments and media endeavours like regional sports networks.

⁷ Alterations in the post-season format of KPBL would represent exciting natural experiments if any had lasted long enough to generate very many distinct observations. Originally, the season was split into 40 spring games/55 fall games and the winners of each met in the Korean Series. In 1985, Samsung Lions won both splits and the rule crowned them champion without a Korean Series. The next year, the top two teams in each split played each other to determine each season’s champion and the winners moved on to the Korean Series. A single-season, two-league play-off was in place by 2001.

⁸ We leave monetary comparisons in South Korean Won (KRW). As of 8 August 2014, 1 USD = 1156.69 KRW and 1 Euro = 1244.20 KRW.

⁹ Lee (2004) compared RSD for KPBL, NPB, the AL and the NL, 1981–2000. The AL (RSD average = 1.68) and NL (RSD average = 1.69) in MLB are the most balanced, followed by the NPB’s Pacific League (RSD average = 1.74) and Central League (RSD average = 1.77). KPBL (RSD average = 2.15) was the least balanced league among the five baseball leagues.

¹⁰ On ‘Quality of the Fan Experience’, we find little variation in either seat quality or amenities offered at any of the KPBL ballparks. Neither could we find any data attesting to such variation. This view is also justified by the statistical insignificance of stadium age up to 4 years, a usual quality variable in the literature, in any of our regression work, below.

¹¹ Knowles, Sherony, and Haupert (1992), Rascher (1999) and Lemke, Leonard, and Tlhokwane (2010) included both probability that home team wins and its squared value in their daily game attendance regression. We also tried WIN²_it as a robustness check. The results were essentially the same as those reported for $\left|WI{N}_{it} - 0.5\right|$, below.

¹² There are not enough years in our sample to perform the time series assessment suggested in Fort and Lee (2006). However, both the Fisher-ADF and Fisher-PP unit-root tests were performed and both rejected the null hypothesis of unit-root process.