24A Page 113
-------------------------
24 - RECTANGLING
-------------------------
| 24A | Introduction |
| 24B | Making and entering rectangles |
| 24C | Crude convergence |
| 24D | Starts for converging rectangles |
| 24E | Rectangle significance tests |
| 24F | Conditional rectangle |
| 24G | Some generalised rectangles |
| 24W | Theory of convergence |
| 24X | Theory of significance tests |
| 24Y | Other theory of rectangles |
24A INTRODUCTORY
(a) General remarks on Chi-breaking.
The ultimate criterion in chi-breaking, as in chi-setting, in the ΔD count.
As in setting, and for like reasons, runs are limited to:
1-wheel runs, known as short wheel-breaking runs;
2-wheel runs known as rectangles.
Even these are impracticable to run by actually trying all possible wheels, involving millions of trials [25X].
Instead methods are used which, in effect, count ΔD against each character supposing it to be a dot: a good count is evidence that it is a dot; a bad count that it is a cross.
This applies equally to short wheel-breaking runs [25 A] and to rectangles: a rectangle could be treated as a short wheel-breaking run whose wheel is composite, e.g. in a 1+2 rectangle the "wheel" is ΔΧ1 + ΔΧ2 which is 41 x 31 = 1271 long.
Back to General Report on Tunny. Contents.