Conversation

This Tweet was deleted by the Tweet author. Learn more
Replying to
something something bayes P(future betrayal | past betrayal) > P(f. betrayal | p. betrayal and good apology) > P(f. betrayal | p. betrayal and recent loyalty) > P(f. betrayal | p. betrayal and good apology and recent loyalty) > P(f. betrayal | history of only loyalty)
Replying to and
so, "enemies for life" in that the comparisons are strict (past betrayal never fully forgotten) but, "unilateral letting-go" in that over time trust can increase even if they don't explicitly apologize, and can even be a better signal than an apology (b/c talk is cheap)
1
Show replies