Like the article in Science News Jaynes makes it clear that the manner in which the coin is tossed is important. To accentuate the effect of bias---and to make the experiment easier to do---he reports how he tossed the 'metal lid of a small pickle jar' in three different ways: in the first he can control whether he gets heads or tails; in the second he always gets tails; in the last he gets 54 heads from 100 trials.
Slashdot Mirror
There's a reasonable discussion of coin tossing in section 10.3 of Ed Jayne's book "Probability Theory: The Language of Science" published by CUP.
Like the article in Science News Jaynes makes it clear that the manner in which the coin is tossed is important. To accentuate the effect of bias---and to make the experiment easier to do---he reports how he tossed the 'metal lid of a small pickle jar' in three different ways: in the first he can control whether he gets heads or tails; in the second he always gets tails; in the last he gets 54 heads from 100 trials.