Skip to contents

Replays the history one ply at a time and returns the method-specific win estimate for each player after every ply. See mxo_win_prob() for the interpretation of heuristic and MCTS estimates.

Usage

mxo_win_prob_curve(game, method = c("heuristic", "mcts"), ...)

Arguments

game

An mxo_game object whose history will be replayed.

method

Win-probability method, forwarded to mxo_win_prob(). Default "heuristic" for speed.

...

Additional arguments forwarded to mxo_win_prob().

Value

A type-stable tibble with columns ply (int), player (int), win_prob (dbl).

Examples

g <- mxo_new_game(n = 3L, k = 3L)
g <- mxo_move(g, "present", 0L, 0L, 0L)
g <- mxo_move(g, "present", 0L, 1L, 26L)
mxo_win_prob_curve(g)
#> # A tibble: 4 × 3
#>     ply player win_prob
#>   <int>  <int>    <dbl>
#> 1     1      1    0.543
#> 2     1      2    0.543
#> 3     2      1    0.543
#> 4     2      2    0.543