World Draughts Forum

Also known as 16377718018836900735 for those who can use printf properly. :) Perft 28 confirmed.

-paul

Hi Rein,
Really great stuff! So basically you are pre-computing the whole transposition table and then running shallow perft() searches on each entry? I'm interested to learn how you generated the unique positions. Assuming you keep them in memory in an array, how do you find/insert new positions ...

I am using almost the same technique for my distributed perft() implementation, but one that also allows computing divide() in an easy way! I am actually writing a small article on this, which I will be happy to share once I get clearance.

Hello Aart,
I was guessing you must be doing something ...

A counterpart statistic to perft N, the number of lines of a given length N, is uniq N,
the number of unique positions reached after N moves. This corresponds with
http://oeis.org/A133047 where Jonathan Schaeffer provides numbers for the first
15 ply. I thought I'd add a few more:

Depth Uniq Ratio ...

move divide(27)
-------------------------------
12-16 = 516399283859880203
11-16 = 519502096014967805
11-15 = 476666239516455180
10-15 = 468705060101275533
10-14 = 400425747281243848
9-14 = 486493422418651579
9-13 = 631652334435528457
-------------------------------
perft(27 ...