gd_simulate_matches: Goal Difference Model Simulate Matches
In neilcuz/panenkar: Panenkar

Simulate matches using the basic goal difference rating model

gd_simulate_matches(
  training_set,
  test_set,
  xi,
  max_gd = NA,
  min_matches = 10,
  market = "result"
)

`training_set,`	columns of which must be "home_team", "away_team", ("result" and/or "over_under_2_5" or similar depending on the market specified) and "match_date"
`test_set,`	same column restrictions as training_set, the set to get outcome probabilities for
`xi`	the time weight parameter for weighting the historic goal difference to derive the rating. This is just using the same Poisson regression time weighting approach but xi may be different
`max_gd`	maximum goal difference to consider, if > than this value it is trimmed, Default: NA results in no trimming. Trimming may be helpful for removing outliers.
`min_matches`	minimum number of matches a team must have played before a rating is assigned. Default: 10.
`markets`	which markets to calculate probabilities for, Default: c("result", "over_under")
`over_under_goals`	used if '"over_under" present in markets, Default: 2.5
`bind_to_test`	if FALSE just the probabilties will be outputted. THis is good for optimising parameters, Default: TRUE