I have a question regarding simulations that test balanced counts. When a simulation is used to test the expectation of a balanced count, is the true count conversion estimated to the nearest half deck (quarter deck, full deck)? the way a human player would estimate. Or is the true count conversion precise? Thanks