I wanted to do an AFL final series simulation with elo rating. I have gathered all the team’s elo rating for the simulation. But I have no clue how to do the simulation, am thinking will have to use rbinom to simulate game results?
I would like the simulation to store the result of each game so that I can compare the result of the actual final in 2019. Am thinking a simulation of 1000 times and it can show the win percentage for the team?
I look up a lot of simulation online, but can’t figure out how to do it. Some would do a 1 ,0.5,0. For the win, draw or lose. But I can’t understand the concept behind it.
> dput(ratings(1:8,)) structure(list(team = c("Geelong", "Collingwood", "West Coast", "Essendon", "GWS", "Footscray", "Brisbane Lions", "Richmond"), elo_rateing = c(1908L, 1945L, 1934L, 1849L, 1821L, 1907L, 1971L, 2083L)), row.names = c(NA, 8L), class = "data.frame") > head(ratings, n = 8L) team elo_rateing 1 Geelong 1908 2 Collingwood 1945 3 West Coast 1934 4 Essendon 1849 5 GWS 1821 6 Footscray 1907 7 Brisbane Lions 1971 8 Richmond 2083