Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: The King at Leiden

Author: Aaron Tay

Date: 05:28:05 05/22/03

Go up one level in this thread


On May 22, 2003 at 03:55:59, Uri Blass wrote:

>On May 22, 2003 at 00:53:37, Stephen A. Boak wrote:
>
>>On May 21, 2003 at 08:50:15, Uri Blass wrote:
>>>
>>>That tells us that testers were simply lazy at that time and did not try all
>>>combination of changing one parameter.
>>>
>>>Ed found that reducing the chess knowledge of Rebel from 100 to 25 helps but no
>>>tester was able to find it.
>>>
>>>Uri
>>
>>Let's think.
>>
>>How many combinations of single parameters & values are there?
>>How many single parameters can be adjusted on CM9000?
>>How many possible numerical settings on each parameter can be used?
>>Multiply the figures.  Call this result 'A'.
>
>I do not know and I guess 300 possibilities(I do not think that changing
>parameter to all the possible options is a good idea and if you see that
>increasing parameter from 100 to 150 is counter productive then you do not need
>to test changing it from 100 to 175).
>
>>
>>How many games will it take, at each single parameter & value combination, to
>>determine if that setting is an improvement?
>>
>>What will the test be?  Test Suite, Actual Game play?
>
>I think that test suite can be a good indication if the tester was not lazy to
>spend years to build the correct test suite that determines if a change is an
>improvement.
>
>I have an idea how to do it but I was too lazy to spend some years of
>computer time to build the test suite.
>I also cannot be sure that my idea is right because I did not test it.
>
>
>>
>>How long does it take to 'be sure' if using Actual Game play?
>
>I do not know.
>maybe 1000 hours but in many cases it is possible to decide after 100 hours
>that the change is bad.
>>
>>If using Test Suite, how many minutes or hours for running the Suite one time.
>>Call this answer 'B'.
>
>I think that in case of test suite 100 hours may be enough when in many case 10
>hours in solving part of thetest suite are enough to decide that there is no
>improvement.
>>
>>Multiply 'A' times 'B'.  What is the total minutes or hours?
>
>something near 100,000 hours(if games are used) or 10,000 hours(if test suites
>are used) .
>
>200 testers when everyone of them use 500 hours and cooperate may be enough
>even in case of games.
>>
>>Now let's think again.  Is it laziness, or a lack of time (Game Play or Test
>>Suites) or something else?
>
>lack of cooperation.
>>
>>It is easier say 'lazy' than to be not lazy.
>>
>>Regards,
>>--Steve
>
>I agree that it is easier to say lazy than to be not lazy.
>I did not criticize a specific tester in my post.
>
>I only think that testers could get better result in the same time of testing.


They could, I dont think there's a logical proof anyway saying they couldnt..
Evidence for the case of Rebel indidcates that's it's not easy.

>Uri



This page took 0 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.