The "guide hole" of ASJ would be put over the 1, 2, 3 .... 25 of ASL successively, this corresponding to position ASJ = ASL + 1, 2, 3 .... 25. The repeats at each position were immediately seen because a repeat implied a hole in each banbury at the same place and the dark background could then be seen standing out against the white of the sheets. The scores at each position were recorded, scores past the "blue line" being separately noted and - if there was more than 50% chance of one or other message being dummy the portion past the blue line was ignored entirely. When comparisons had been made with ASJ on top - they were then made with ASL on top giving the results for ASL = ASJ + 1, 2, 3 .... 25.
25. Having made the comparisons the next stage was the "decibanning". This was a process of evaluating the scores in terms of convenient units. It was done from one of a series of charts. The chart used depending on the % chance of the comparison being "bogus" owing to one or other message being a dummy. Repeats of more than one letter long were allowed for by a system of bonuses. An example should make this method clear. Score:- 0/30 + 11 xx/171 between two messages dummy %'s 15 and 35 respectively - the 0/30 before the blue line, the remainder after the blue line. The first part of the score is evaluated from a "1" chart (for use for comparisons that are certainly genuine); looking opposite "0" and under "30" we see "5" entered in black - this is a score of -5 units. For the rest we take a "3" chart (used for comparisons with about a 50% chance of being genuine); for the in the score we get a bonus of 4 repeats and for the we get a bonus of 1 repeat making the score 11 + 4 + 1 = 16/171. Looking opposite 161 and under 171 we see 19 entered in red - this is score of +19 units. Total score -5 + 19 = +14. If the comparison was between ASJ and ASL at position ASL = ASJ + 5, an entry of [or 14 in red] would be made on the J and L deciban sheets in the same sort of way as illustrated in Para. 16.
26. The other main job done by the Big Room was tetra counting. All tetras from the catalogue in which the message trigrams involved