From: John Joyce <dangerwillrobinsondanger@...>
Date: 2007-10-30T00:58:39+09:00
Subject: Re: Checkers library


On Oct 29, 2007, at 1:05 AM, M. Edward (Ed) Borasky wrote:

> Rick DeNatale wrote:
>> On 10/28/07, M. Edward (Ed) Borasky <znmeb@cesmail.net> wrote:
>>> James Edward Gray II wrote:
>>>> On Oct 28, 2007, at 10:11 PM, M. Edward (Ed) Borasky wrote:
>>>>> Yeah, especially in light of the fact that there is a "perfect"
>>>>> checkers program now -- it can't be beaten, only tied.
>>>> I bet that took more than a weekend to build.  ;)
>>> The original Samuels checker program was probably built in a  
>>> couple of
>>> weeks and probably would play perfect checkers on an infinitely  
>>> fast IBM
>>> 704. :) I haven't paid much attention to what's inside the current
>>> champion -- Texas Hold 'Em seems like a lot more fun.
>> It might be my senility, but I seem to remember the Samuel's program,
>> or a close descendant of the original beat the current world  
>> champion,
>> at least in a single game, quite early, like in the early 1960s.
>
> I don't think it was that good. It was good enough, however, to  
> remove enough of the challenge from programming the game that  
> nobody put in the effort to make better programs. Instead, they  
> moved on to more difficult games.
>
> Curiously enough, it looks like Go is nowhere near being even  
> competently played by a computer, while chess and poker have  
> credible computer players and checkers has gone the way of Tic-Tac- 
> Toe and Blackjack.
>
Apparently, you've never played some of the video game versions of Go  
available in Japan. They're competent enough!
Go is like Pente, but harder. Simple to learn, impossible to master.  
Fatigue is a major factor in Go.