Author: Albert Silver
Date: 05:45:23 01/19/06
Go up one level in this thread
On January 18, 2006 at 23:30:06, ERIQ wrote:
>On January 18, 2006 at 21:23:42, Albert Silver wrote:
>
>>On January 18, 2006 at 20:33:34, Graham Banks wrote:
>>
>>>On January 18, 2006 at 20:02:28, Albert Silver wrote:
>>>
>>>>On January 18, 2006 at 19:18:42, robert flesher wrote:
>>>>
>>>>>Thank-you for the observation, however, unless its a book loss which it is not I
>>>>>think it is fair game. Although I understand that perhaps this can stastically
>>>>>alter the final results, maybe not. I guess book learning would fix this issue,
>>>>>it does give me some thinking to do, cheers.
>>>>
>>>>With all due respect, I think it simply invalidates the results. 1 game in 6 had
>>>>the exact same opening? How can one possibly compare the strength of the engines
>>>>in such a case?
>>>>
>>>> Albert
>>>
>>>
>>>
>>>Hi Albert,
>>>
>>>I don't think you'll find that Robert is the only person who tests like this and
>>>he certainly won't be alone in thinking that it's fine, especially if learning
>>>is activated.
>>>Because I test with learning off and use generic books, I don't allow any
>>>duplicate opening lines, that is the position at which the engines leave the
>>>book.
>>>There are those who criticise this also.
>>>To each his own, but as long as testing conditions and preferences are made
>>>clear, members can make up their own minds about the usefulness or validity of
>>>any testing.
>>>
>>>Regards, Graham.
>>
>>
>>Indeed, and I made up mine and posted it. 31 times the same disastrously scored
>>opening? 13 for another? The same color in all occasions, and this is to go on
>>and on, and prove what? That if one repeats the same losing line that one
>>will... lose? Why not just have it play 200 straight times that same opening
>>with the same color and report the results? However, as you say, one can do as
>>one wishes. The problem is that if the intention is to make comparisons on a
>>program's ability, then you have failed, unless you mean only their competence
>>in that position.
>not really! I think that if a program doesn't know not to repeat a losing line
>then it deserves it's fate.
There's only one problem: it isn't the engine's fault.
Albert
People do the same sometimes, I myself have done
>this stupid thing and have also seen strong players do it too. Once a expert
>player asked me to play the french for money and he said that he would beat me
>every time so I took up the challenge, we played 5min/game and I was black every
>game, after a few games I saw that one line was great for me because he did not
>know it and I kept repeating the line untill he was out of money, well if I
>could figure out my performance rating it would be at least 2400 is that right?
>am I really that strong? no but that was my fate that day :) and his fate was
>that dispite the fact that he was normally around 2100 fide he played at 1900 or
>worst in that line.
> Eriq
This page took 0.01 seconds to execute
Last modified: Thu, 15 Apr 21 08:11:13 -0700
Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.