incotel@calva.netI learned the game when I was 12. I quickly became the strongest player of my family. My strategy : get the maximum of edges. I wrote at this time a computer opponent following this strategy in BASIC. Of course this opponent was not as good as possible and presumely because it can't search more than one ply. So I decided to learn the assembly language of the Intel 8080. My goal make an alpha-beta Othello program using mobility this time since this strategy appears to me much more useful. At 14, I named, modestly, this assembly Othello program Othello Master. It was not perfect at all, but reaching depth 5, it was sometimes able to beat me... The journal l'Ordinateur Individuel organized some tournaments for Computer Othello programs. There was 3 categories namely "Assembly/Compiled" for assembly or compiled programs, "Interpreted" for Interpreted programs and "Pocket Computer". I entered several programs doing Othello 6x6 in these category for the Sharp PC1211, Casio FX702P, Sharp PC1500, ... All this programs were written in Basic and the Opening book was computed by the assembly program doing depth 6! (the computer running the assembly program was far from being portable for the tournament). In 1983, I owned a personal computer, a SANYO PHC25, if I remember well. I ported my assembly program for the Z80 coding it in hexadecimal manually... This program eventually finished second of the tournament in the Assembly category... Not that bad... During 1983 and 1989, although I did made some programs, I do not remember very good record. In 1989, I was owning of course a Compatible PC and for the purpose of my master thesis, I begin the coding of a brand new program using bit boards for representating the board. This program appears on the 1989 Computer Olympiad but did not do a too much good results. It was taking too much edges and the opening book was also faulty (as usual :) )... In 1994, I discovered the Internet Othello Server (IOS) and I started to play on it by myself. When for experimenting some parallel minimax algorithms, I did a small game program. I took one of my C program I wrote in 1984 (In fact my first C program) and started to play with it on IOS. Well, looks like the evaluation function was not too good :). I wrote an other one from scratch, add hash table and I had a new program to experiment my parallel things. The parallel programs made the sensation when running on a 32 processors Connection Machine 5, it was able to beat Logistello in a tournament... (In the meanwhile I had also improved the endgame search, looks for jcwend.c on IOS). The sequential version of the program appears to be names Thumper on IOS. There was after some deviations from this program, Baloo , using consistency search and latter Bugs with some small changes in the search and also in the opening book learning technic. Bugs had some good records, but is still a bad program from my point of view, having a human coded evaluation function, it is very unstable in the choice of the moves and this instability also gave a slow search. (If you are stable enough, you will be able to sort the moves and so to have a faster search...).