Re: Testable Predictions by HdB
*** T. Winter wrote:
5-9. HdB 5-point, 9-point, 17-point, 33-point and 65-point.
Duuhh. Everything else than 25-point is pointless :-(
[ ... ] Also in normal precision increasing HdB's method to 127 points
did not help (but that may be different on a system, like Intel, with
a more extended precision for the evaluation).
No. Increasing HdB's method to 127 points is pointless.
4. With HdB's method you may increase the number of points without
ill-effect to the precision, but there will be no improvement when
you cross a threshhold.
That threshold is 25 points. As I have "predicted" repeatedly. More news
later on ...
Han de Bruijn
.
