Match outcomes from repeated table tennis matches

Match outcomes from repeated singles table tennis matches

data(table_tennis)

Format

A likelihood function corresponding to the match outcomes listed below.

Details

There are four players, A, B, and C, who play singles table tennis matches with the following results:

• A vs B, A serves, 5-1

• A vs B, B serves, 1-3

• A vs C, A serves, 4-1

• A vs C, C serves, 1-2

Object table_tennis is a hyper2 log-likelihood function for these observations; S is the strength of the serving ghost. Probability vector table_tennis_maxp is the maximum likelihood estimate.

These objects can be generated by running script inst/table_tennis_serve.Rmd, which includes some further discussion and technical documentation and creates file table_tennnis.rda which resides in the data/ directory.

Examples

data(table_tennis)
dotchart(table_tennis_maxp)


specificp.gt.test(table_tennis, "S", 0)
#> 
#> 	Constrained support maximization
#> 
#> data:  table_tennis
#> null hypothesis: sum p_i=1, S <= 0 (notional)
#> null estimate:
#>            S            a            b            c 
#> 9.999902e-06 3.700757e-01 3.120442e-01 3.178700e-01 
#> (argmax, constrained optimization)
#> Support for null:  -12.2133 + K
#> 
#> alternative hypothesis:  sum p_i=1 
#> alternative estimate:
#>         S         a         b         c 
#> 0.4504171 0.2410390 0.1445159 0.1640281 
#> (argmax, free optimization)
#> Support for alternative:  -9.399965 + K
#> 
#> degrees of freedom: 1
#> support difference = 2.81334
#> p-value: 0.01768911 (one-sided)
#>