Competitive balance measures in us professional sport: an empirical comparison



Download 362.34 Kb.
Page4/5
Date17.08.2017
Size362.34 Kb.
#33659
1   2   3   4   5

Table 11 – Statistical Analysis of the Baseball Models

Measure

Model

SD

SD = 0.067 + 0.018ET* + 0.009YT* - 0.006LT + 0.008RS

NSD

NS = 1.704 + 0.463ET* + 0.219YT* - 0.163LT + 0.191RS

GC

GC = 0.074 + 0.019ET* + 0.009YT* - 0.008LT + 0.010RS

HHI

HH = 0.042 + 0.000ET + 0.003YT + 0.001LT - 0.009RS*

FCCR

FC = 0.241 + 0.008ET + 0.018YT + 0.002LT - 0.044RS*

Table 12 – Baseball Models

Key:


ET – Expansion Team

YT – Young Team

LT – Luxury Tax

RS – Revenue Sharing

Using the measures SD, NSD and GC the existence of the Luxury Tax is the only variable that has a negative sign. This suggests that the Luxury Tax initiative does improve CB in Baseball although the significance level of the Luxury tax variable is low. The other variables all have positive signs suggesting that they have an adverse effect on CB. That is to say that when an expansion occurs or just after an expansion occurs CB is reduced. The anomaly is the Revenue Sharing variable which has a positive sign and therefore appears to reduce CB.

However, using the HHI or the FCCR measures it is the revenue sharing variable that has the negative sign and therefore appears to improve CB whereas the other variables have positive signs, albeit with low absolute values for the HHI model and low absolute values for the FCCR model excluding the existence of a Young Team.

In the models using the measures SD, NSD and GC the constant and the two variables relating to the existence of an expansion team or a young team are all significant. This is in contrast to the models using the HHI or the FCCR where only the constant and the variable relating to the existence of the revenue sharing agreement are significant.

The collinearity of the variables was examined via the tolerance measure statistic as described in section 3.9. In all cases the tolerance values were well above the 0.1 threshold, mainly in the range 0.3-0.5. This was repeated for all models across all sports and as such will not be mentioned again in this study since it is clear that the variables are not collinear.



4.4.2 NBA (Basketball)

Dependent_Variable___Mean___Standard_Deviation'>Dependent Variable

Mean

Standard Deviation

SD

0.152

0.023

NSD

2.752

0.419

GC

0.167

0.026

HHI

0.057

0.027

FCCR

0.350

0.142

Table 13 – Dependent Variables used in Basketball Models

Explanatory Variable

Mean

Expansion Team

0.23

Young Team

0.40

Low Salary Cap

0.23

Medium Salary Cap

0.11

High Salary Cap

0.17

Table 14 – Explanatory Variables used in Basketball Models

Dependent

Explanatory

Coefficient

t-Statistic

Significance

SD

Constant

0.137 *

22.079

.000

Expansion Team

0.012

1.595

.118

Young Team

0.009

1.422

.163

Low Salary Cap

0.017 *

2.118

.040

Medium Salary Cap

0.031 *

2.882

.006

High Salary Cap

0.007

0.749

.458

NSD

Constant

2.473 *

22.193

.000

Expansion Team

0.218

1.615

.114

Young Team

0.173

1.459

.152

Low Salary Cap

0.320 *

2.202

.033

Medium Salary Cap

0.565 *

2.955

.005

High Salary Cap

0.133

0.813

.421

GC

Constant

0.149 *

21.707

.000

Expansion Team

0.011

1.313

.196

Young Team

0.011

1.529

.134

Low Salary Cap

0.023 *

2.544

.015

Medium Salary Cap

0.039 *

3.276

.002

High Salary Cap

0.012

1.181

.244

HHI

Constant

0.074 *

11.798

.000

Expansion Team

0.006

0.752

.456

Young Team

-0.001

-0.195

.847

Low Salary Cap

-0.031 *

-3.787

.000

Medium Salary Cap

-0.036 *

-3.338

.002

High Salary Cap

-0.037 *

-4.084

.000

FCCR

Constant

0.434 *

13.295

.000

Expansion Team

0.031

0.784

.437

Young Team

-0.002

-0.045

.965

Low Salary Cap

-0.159 *

-3.740

.001

Medium Salary Cap

-0.189 *

-3.378

.002

High Salary Cap

-0.198 *

-4.133

.000

Table 15 – Statistical Analysis of Basketball Models

Again there is consistency between the models for the SD, NSD and GC measures. In each the Low and Medium salary cap variables are significant while the other variables are not significant.

The coefficients of all the variables are positive suggesting that they have an adverse effect on CB. This is intuitively correct for the Expansion and Young team variables but counter intuitive for the variables that relate to league measures introduced to improve CB.

There is also consistency between the models for the HHI and FFCR measures. All three salary cap variables are significant and have negative signs. The Young team variable has a negative sign in both but is not statistically significant. The Expansion team variable is the only one with a positive sign but that isn’t significant either.



As can be seen there are significant differences between the HHI, FCCR measures and the other three. Different variables are significant and have different signs.

4.4.3 NFL (American Football)

Dependent Variable

Mean

Standard Deviation

SD

0.197

0.023

NSD

1.552

0.161

GC

0.218

0.027

HHI

0.040

0.003

FCCR

0.269

0.022

Table 16 – Dependent Variables in American Football Models

Explanatory Variable

Mean

Expansion Team

0.11

Young Team

0.24

Low Salary Cap

0.32

High Salary Cap

0.08

Free Agency Plan B

0.11

Free Agency Plan A

0.43

Balanced Schedule

0.30

Table 17 – Explanatory Variables used in American Football Models

Dependent

Explanatory

Coefficient

t-Statistic

Significance

SD

Constant

0.203 *

32.741

.000

Expansion Team

-0.003

-0.186

.854

Young Team

0.002

0.143

.887

Low Salary Cap

0.030

1.091

.284

High Salary Cap

0.042

1.356

.185

Free Agency Plan B

-0.001

-0.100

.921

Free Agency Plan A

-0.043

-1.753

.090

Balanced Schedule

-0.004

-0.261

.796

NSD

Constant

1.581 *

36.513

.000

Expansion Team

-0.049

-0.483

.633

Young Team

0.010

0.132

.896

Low Salary Cap

0.248

1.293

.206

High Salary Cap

0.335

1.556

.130

Free Agency Plan B

0.035

0.377

.709

Free Agency Plan A

-0.296

-1.743

.092

Balanced Schedule

-0.030

-0.299

.767

GC

Constant

0.225 *

31.033

.000

Expansion Team

-0.002

-0.144

.886

Young Team

0.003

0.264

.793

Low Salary Cap

0.030

0.941

.354

High Salary Cap

0.041

1.145

.261

Free Agency Plan B

-0.001

-0.093

.927

Free Agency Plan A

-0.046

-1.606

.119

Balanced Schedule

-0.003

-0.162

.873

HHI

Constant

0.043 *

74.408

.000

Expansion Team

-0.001

-0.682

.501

Young Team

-0.001

-0.596

.556

Low Salary Cap

-0.001

-0.376

.709

High Salary Cap

-0.002

-0.753

.457

Free Agency Plan B

-0.001

-1.175

.250

Free Agency Plan A

-0.004

-1.569

.128

Balanced Schedule

-0.001

-0.796

.433

FCCR

Constant

0.285 *

62.589

.000

Expansion Team

-0.008

-0.788

.437

Young Team

-0.003

-0.314

.756

Low Salary Cap

-0.000

-0.011

.991

High Salary Cap

-0.006

-0.282

.780

Free Agency Plan B

-0.010

-1.019

.316

Free Agency Plan A

-0.026

-1.432

.163

Balanced Schedule

-0.004

-0.365

.718

Table 18 – Statistical Analysis of American Football Models

There are no significant variables in any model. This suggests that all the initiatives and effects examined do not significantly influence the CB of the NFL. The strongest variable throughout is the Free Agency Plan A variable.



The HHI and FCCR models are very similar in that all coefficients are negative. The SD, NSD and GC models have the same signs for all their variables with the exception of Free Agency Plan B which is positive for the NSD model and negative (albeit very small) for the other two models.

4.4.4 NHL (Ice Hockey)

Dependent Variable

Mean

Standard Deviation

SD

0.115

0.026

NSD

2.062

0.451

GC

0.124

0.028

HHI

0.052

0.016

FCCR

0.313

0.094


Download 362.34 Kb.

Share with your friends:
1   2   3   4   5




The database is protected by copyright ©ininet.org 2024
send message

    Main page