Here's a diagram outlining the bottom five strips of my 1969 or 1970 CalComp 565 method of testing randomness of bit sequences from the IBM 1130 FORTRAN (pseudo)random number generator. See article I'm posting to comp.programming and sci.math the evening of 2005.Jul.05 for explanation of the method. Note that this diagram is greatly expanded to allow room for the explanations in each rectangle within each strip. For the actual CalComp 565 plots, each strip was appx. 1.5 or 2 inches from left to right, and one quarter inch tall per strip, total of about ten stacked strips. +-------------------------------+-------------------------------+ | | | | | | | (5 coins) |1| 5 | 10 | 10 | 5 |1| | | | | | | |/32 +-------------------------------+-------------------------------+ | 4T| 3H + T | 2H + 2T | H + 3T |4H | (4 coins) |1/ | 4/ | 6/ | 4/ |1/ | | 16| 16 | 16 | 16 | 16| +-------------------------------+-------------------------------+ | | | | | (3 coins) | 3c=TTT| 3 coins = H + 2T | 3 coins = 2H + T |3c=HHH | | p=1/8 | p = 2/8 | p = 2/8 | p=1/8 | +-------------------------------+-------------------------------+ | | | | (2 coins) | twoCoins = 2T | two coins = H+T | twoCoins = 2H | | p = 1/4 | p = 2/4 = 1/2 | p = 1/4 | +-------------------------------+-------------------------------+ | | | (1 coin) | one coin = T p = 1/2 | one coin = H p = 1/2 | | | | +-------------------------------+-------------------------------+ (Naturally, to accomplish this AsciiArt diagram, I started at the bottom, describing each rectangle clearly, then worked my way upward abbreviating more and more as the crowding got more severe.)