Consider a lookup table or decision matrix.
Imagine this table, for Rock-Paper-Scissors
R P S
R - L W
P W - L
S L W -
And where it is interpreted as so, reading ROW then COL
ROW(R) at COL(R) -> R *ties* R
ROW(S) at COL(P) -> S *wins* P
(The two diagonal halves of the matrix are inverted mirrors in this case.)
