Title: The distribution of wins
1Competitive Balance
- The distribution of wins
- in professional team sports most consistent with
- the maximization of
- league profits.
2Neal, Walter. The Peculiar Economics of
Professional Sports Quarterly Journal of
Economics 78 (February, 1964) 1-14.
- Louis-Schmelling paradox
- The inverted joint product or the product joint
- Joint product - two products technologically
resulting from a single process. - Product joint - An indivisible product from the
separate processes of two or more firms. - Professional baseball produces several
interrelated streams of utility - In-person viewing of the game
- Broadcasts of games
- League standing effect
- Fourth Estate Benefit
- Conclusion The several joint products which are
products joint of legally separate business firms
are really the complex joint products of one
firm, and this firm is necessarily an
all-embracing firm or natural monopoly.
3Neale (1964), cont.
- Four possible cases of interleague competition
- Major League Baseball solution the joining of
two leagues into one monopoly. - The professional football solution of the 1940s
The bankruptcy of one league. - Survival of two or more leagues that are not
economically competitive due to geographic
distances or the institutions of sport and
culture. - Survival of two or more leagues that are
economically competitive and which could be
sportingly competitive. - The second case is the most common solution.
Geographic distances and culture institutions
seem to be overcome overtime (exception CFL and
Japanese baseball). - In general, additional leagues bid up costs and
reduce revenues, hence reducing the profitability
of each league.
4Uncertainty of Outcome
- Why is balance important? Uncertainty of
outcome is crucial to the demand for sporting
events. The works of Knowles, Sherony and
Haupert (1992), and Rascher (1999) found that
Major League Baseball attendance was maximized
when the probability of the home team winning was
approximately 0.6. These studies suggest that
consumers prefer to see the home team win, but do
not wish to be completely certain this will occur
prior to the game being played. - Knowles, Glenn, Keith Sherony, and Mike Haupert.
1992. The Demand for Major League Baseball A
Test of the Uncertainty of Outcome Hypothesis.
The American Economist, 36, n2, Fall 72-80. - Rascher, Daniel. 1999. A Test of the Optimal
Positive Production Network Externality in Major
League Baseball. Sports Economics Current
Research, Edited by John Fizel, Elizabeth
Gustafson and Lawrence Hadley. Praeger 27-45.
5Perfect Competitive Balance
- But do fans want perfect competitive balance?
- Teams in larger markets generate greater revenue
from an additional win than teams in smaller
markets. If marginal cost is the same for all
teams, then teams in larger markets would
maximize profits at a higher winning percentage
than teams in smaller markets. - This leads to four questions
- how do we measure competitive balance?
- what determines competitive balance?
- how much competitive balance does a league need?
- and do fans really care?
6Measuring Competitive Balance
- Standard deviation S(PCT actual - PCT
mean)2/number of teams0.5 - Idealized standard deviation (if every team was
equal) (.500)/ N0.5 - Where N number of games each team plays in a
season - With a normal bell shaped distribution
- 2/3 of league will be within one standard
deviation - 95 will be within two standard deviations
- 99 will be within three standard deviations
7The Noll-Scully Measure
- CBit s(wp)itactual/ s(wp)itideal
- with s(wp)itideal µ(wp)it / vN
- Where
- s(wp)it is the standard deviation of winning
percentages within league (i) in period (t) - µ(wp) it is league (i)s mean
- N is the number of teams
8What determines competitive balance in
professional sports?
- Schmidt, Martin B. and David J. Berri. (2003)
On the Evolution of Competitive Balance The
Impact of an Increasing Global Search. Economic
Inquiry, 41(4) 692-704.
9Competitive Balance in the AL
- Figure 1 - Noll-Scully Competitive Balance (CBt)
Measures - American League (AL)
10Competitive Balance in the NL
- Figure 1 (cont.) - Noll-Scully Competitive
Balance (CBt) Measures - American League (NL)
11A Simple Empirical Model
- Table I - OLS Trend Estimates for (CBt)
- Dependent Variable (Sample - 1911 2000)
-
- CBt AL Constant Time
- 2.742 -0.012 (0.109) (0.002)
- CBt NL Constant Time
- 2.439 -0.008 (0.108) (0.002)
- Standard errors are beneath each coefficient.
- The results indicate that competitive balance
improved throughout the 20th century.
12What explains the pattern?
- Institutional factors
- The Reserve Clause
- Reverse-order draft
- Free agency
- Revenue sharing
- Payroll and Salary Caps
- Expanding Populations or Talent Compression
13The Belief in Institutions
- Consider Commissioner Bud Seligs comment on the
recent (2002) Major League Baseball labor
agreement - . . . the issue here was competitive balance and
I feel this deal clearly deals with that. - From the news conference announcing the
agreement, Friday, August 30th, 2002.
14Major League BaseballsBlue Ribbon Panel
- Convened to examine
- The Economic Stability of MLB
- Competitive Balance in MLB
- Four members
- Yale president Richard C. Levin
- Former Federal Reserve chairman Paul Volcker
- Former Senator George Mitchell
- Columnist George Will
15The Blue Ribbon PanelThe Conclusions
- a significant disparity exists in the resources
member teams - For example, for the 2000 season the salary of
the highest paid player in MLB exceeded the
entire payroll of the Minnesota Twins (BRP 2000
p. 9). - such differences in market size has allowed a
collection of teams to consistently field
playoff-contending teams. - teams located in smaller-markets are incapable of
fielding teams that can challenge for post-season
success. Consequently, for the smaller-market
teams the outcome of the season is known before
the season is started.
16The Blue Ribbon PanelRecommendations
- The Blue Ribbon Panel recommended a number of
changes - recommended that foreign players be subject to
the draft - teams maintain the rights of draftees beyond the
one-year period they currently hold. - an annual "competitive balance draft," under
which the eight clubs with the worst records
could draft players not on the 40-man roster of
the eight playoff teams. - Most of these seem to limit player movement or to
convey greater rights to Major League Baseball
teams
17The Coase-Rottenberg Theorem
- When there are no transaction costs the
assignment of legal rights have no effect upon
the allocation of resources among economic
enterprises. Stigler (1988) - When there are no impediments to the buying and
selling of playing talent, the assignment of the
rights to this talent will have no effect upon
the allocation of players among Major League
Baseball teams. Rottenberg (1956)
18On an empirical level
- the impact of the reverse-order draft in 1965
- El-Hodiri and Quirk (1971), Demsetz (1972), and
Quirk and Fort (1992) all found no tangible
impact - Daly and Moore (1981) and Daly (1992) both found
an improvement - the impact of the free agency in 1976
- Szymanski (2003) examines (20) empirical studies
on the impact of the introduction of free-agency
in 1976 on competitive balance and finds that (9)
estimate an improvement, (4) document a decline
and (7) found no impact
19If not institutional factors, what?
- Evolutionary Biologist Stephen Jay Gould
- the distribution of athletic talent in a
population should be normally distributed. - At the right-tail of the distribution would lay
those with the greatest level of athletic ability
and assuming that there is a bio-mechanical limit
to potential ability or talent, the athletes in
the far right-tail tend to be relatively equal.
20Goulds Hypothesis (Part 1)
- At the beginning of the 20th century, people
playing Major League Baseball were only white
Northeastern American males. - The population baseball could draw upon was
relatively small and correspondingly, there
existed a large degree of heterogeneity between
players. - As the probability of winning is closely aligned
with playing talent, such diversity may lead to
low levels of competitive balance.
21Goulds Hypothesis (Part 2)
- As the population of players Major League
Baseball has to choose from rises, new players
are added to the population in much the same way,
i.e., normally. - The absolute number of players close to the limit
would rise and, given player demand, as would the
average players talent level or ability. - The Gould hypothesis, therefore, argues that as
the talent pool rises greater player homogeneity
should be observed.
22Goulds Hypothesis (Part 3)
- In which case, one should see a corresponding
rise in the probability of a poor team beating a
stronger team - as the poor team is now stocked with players
closer in talent to those of the stronger team. - Following the theme, increased player demand
through, for example, expansion should decrease
competitive balance. - Such increased homogeneity in talent reduces
individual player differences between competing
teams and therefore would increase the likelihood
of a less talented team beating a more
talented team. - This suggests that, for example, racial
integration would increase competitive balance
because teams were able to choose from a larger
talent pool.
23No .400 Hitter since 1941
Player Year Average
Nap Lajoie 1901 0.4265
Joe Jackson 1911 0.4081
Ty Cobb 1911 0.4196
Ty Cobb 1912 0.4087
George Sisler 1920 0.4073
Ty Cobb 1922 0.4011
Rogers Hornsby 1922 0.4013
George Sisler 1922 0.4198
Harry Heilmann 1923 0.4027
Rogers Hornsby 1924 0.4235
Rogers Hornsby 1925 0.4028
Bill Terry 1930 0.4013
Ted Williams 1941 0.4057
24Cursory Evidence
25Competitive Balance Across Sports
Sport League Years Avg. Level of CB
Basketball NBA 1967-68 to 1975-76 2.59
ABA 1967-68 to 1975-76 2.60
Baseball AL 1901-2000 2.12
NL 1901-2000 2.08
Hockey NHL 1972-73 to 1978-79 2.59
WHA 1972-73 to 1978-79 1.89
Football NFL 1960-1969 1.57
AFL 1960-1969 1.58
Soccer Bundesliga 1964-95 1.32
NASL, MLS 1967-84, 1996-2000 1.34
26Baseballs Labor Pool
- Globalization of baseball is now evident on the
playing fields in the United States. - Players still hail from the traditional areas of
recruitment, such as the United States, Dominican
Republic, Puerto Rico, Venezuela, and Cuba - Many players from Mexico, Australia, Japan, and
Korea also play in the Major Leagues. Even such
countries as Spain, Belgium, the Philippines,
Singapore, Vietnam, Great Britain, Brazil,
Nicaragua, and the Virgin Islands have produced
professional baseball players. - In 2000, the number of foreign-born players on
Major League Baseball rosters was 312,
constituting 26 percent of all players (Levine et
al, 2000).
27Schmidt Berri (2003)
- Competitive Balance and Major League Baseballs
Labor Pool are cointegrated - MLBs Labor Pool is weakly exogenous
- Other factors, captured through various dummy
variables, appear not to be relevant.
28Back to Perfect Competitive Balance The work of
Stefan Szymanski (from the Western Economic
Association meetings in 2005)Tilting the Playing
Field (less)
29Empirical specification for model designed to
explain attendance in MLB
- Data- club attendance, wins, new ballparks,
strikes, 1978-2003 - Not capacity constrained
- Panel regression
- ? (Attendance)it ? ?i ?t ?i ?
(winpercentage)it ?Xit ? - issues price, alternative functional forms, lags
30Estimated increase in attendance due a unit
increase in win percentage (e.g. from 50 to
51) American League
31Estimated increase in attendance due a unit
increase in win percentage (e.g. from 50 to
51) National League
32Redistributing wins would have big effects
33Attendance maximising distribution of wins in the
2003 National League
34Empirical issues
- Linear model fits best but is not plausible at
the extremes - Almost all observed win percentages lie between
0.33 and 0.66 - Capacity constraints may bind at very high win
percentages - Alternative is to fit a quadratic model to the
linear estimates - This approach favours the conventional
competitive balance justification for restraints
35Fitting a quadratic model from the linear
estimates
36American League 2003
37National League 2003
38Optimal distribution of wins, American League
2003, quadratic model
39Impact of optimal win distribution 2001-2003
40Szymanski Conclusions
- The impact of wins on attendance varies across
the league. - In other words, fans ask more (or less) of their
teams in different markets. - If Major League Baseball wished to maximize
league attendance, it would not want perfect
competitive balance. - Attendance would increase with less competitive
balance (relative to what we see)
41How does competitive balance impact league
demand? According to Schmidt-Berri (2001)
- Competitive balance has a larger impact (i.e.
more competition leads to more demand) the longer
the time period one considers. - In other words, balance across three years has a
smaller impact than balance over five years. - from the article.These results suggest that
fans may not prefer competitive balance in a
given season, but over time, the slogan wait
till next year must remain valid for the teams
located toward the bottom of the standings
42more Schmidt and Berri (2001)
- from The Wages of Wins
- The largest impact was seen when we looked at
improvements in our five-year measure. A movement
from the least competitive five-year measure to
the highest level of balance would result in
about 3,500 additional fans per game for each
team. This works out to a 14 increase in
attendance. Again, this was the largest impact we
found. So a very large change in competitive
balance across the longest time period we
considered, and 14 is all the oomph we uncover.
43How does competitive balance impact league
demand? According to Humphreys (2002)
- from The Wages of Wins
- Humphreys looked at three measures of competitive
balance, the Noll-Scully, the Herfindahl-Hirschman
Index, and Humphreys Competitive Balance Ratio
(CBR). - Humphreys (2002) defines the CBR as follows The
CBR scales the average time variation in won-loss
percentage for teams in the league by the average
variation in won-loss percentages across seasons
it indicates the relative magnitude of each type
of variation across a number of seasons (p.
137). - In a model designed to explain league attendance
Humphreys found that only the CBR was
statistically significant. - Although Humphreys does not address this issue,
his results suggest that a movement from the
lowest observed level of balance to the highest
would increase a teams attendance by 4,000 fans
per game. Again, that is quite similar to what we
found with the Gini Coefficient.
44Competitive Balance Lessons from The Wages of
Wins
- The measurement of competitive balance should
focus on outcome in the regular season, not the
post-season. - Relative to its history, baseball is more
balanced today. - Relative to sports like football and soccer,
baseball is not as balanced. Compared to
basketball, though, baseball does not have a
competitive balance problem. - Competitive balance appears to be dictated
primarily by the underlying population of talent,
not league policy. - Although we can find a statistical relationship
between competitive balance and attendance, the
estimated economic significance seems quite
small. Consequently, it is not clear whether the
fans truly care about the level of balance in a
league. - The relationship between team revenue and wins
suggests that perfect competitive balance would
actually lower league revenues.