predict.dixoncoles: Predict method for Dixon-Coles model fits

In Torvaney/regista: Soccer Analytics

Description

Predicted rates or scorelines based on a Dixon Coles model object

Usage

## S3 method for class 'dixoncoles'
predict(object, newdata = NULL, type = c("rates",
  "scorelines", "outcomes"), up_to = 50,
  threshold = sqrt(.Machine$double.eps), ...)

Arguments

object	Object of class inheriting from 'dixoncoles'.
newdata	A data frame in which to look for variables to predict
type	Type of prediction (rates or scorelines).
up_to	If 'type = "scorelines"', the maximum number of goals for which to calculate the probability of occurring in each match.
threshold	If 'type = "scorelines"', scorelines with a probability below 'threshold' will not be returned.
...	Arguments passed from other methods

Value

A list in which each element is a tibble. The contents of the tibble depends on the value supplied to the 'type' argument. These values are enumerated for each possible value of 'type' below:

'rates'

the side ("home" and "away") and the goalscoring rate of both teams

'scorelines'

the probability ('prob') for each scoreline ('hgoal' and 'agoal')

'outcomes'

the probability ('prob') of each outcome ("home_win", "draw" or "away_win") occurring

Torvaney/regista documentation built on June 7, 2021, 12:21 a.m.