In the fabric of the Cosmos by Brian Greene there is an experiment using the example of Mulder and Scully and boxes. It is an analogy to Bell's experiment determining whether articles have definite spin before you measure them. It shows that if they have definite spin before you measure them two entangled particles must have the same (or more precisely opposite) directions of spin more than 50% of the time. However it seems to me that you would get the exact same result either way. If you test three axes and each can be clockwise or counterclockwise, as a book shows there are nine possibilities. If the spin is randomly determined whenever you measure it and is always the same for both particles if you measure the same axis then it seems that in three out of the six possibilities the spins would be the same (or more precisely opposite). The other six spins would be the same 50% of the time. Therefore it seems than either case the spins would be the same more than 50%. Where am I going wrong here?