Author: Rolf Tueschen
Date: 18:37:01 05/25/02
Go up one level in this thread
Part 1
On May 25, 2002 at 20:38:05, Christophe Theron wrote:
>On May 25, 2002 at 17:19:19, Rolf Tueschen wrote:
>
>>On May 25, 2002 at 16:00:01, Peter Fendrich wrote:
>>
>>>You and some other guys keep saying this ("Nothing") but that is not a good
>>>interpretation of the list.
>>>The ratings are the best predictions of each engines strength. Fritz is still
>>>the best prediction of who is the best even if it's a quite unsafe prediction.
>>
>>This is absolutely false. And at the first glance it's both true and false.
>>
>>- it's true because now, at the moment of the cut of the information stream
>>FRITZ was 8 points above the second (but the moment of the cut is important;
>>without further information we cannot judge whether the cut was good or bad)
>
>
>The moment of the cut is not important in statistics.
>
>The later the better, that's all. More data is better than less data.
>
>You want us to believe that the list can be manipulated by chosing the right
>moment to publish it.
Thanks for not saying that I in fact said it. Because I do not say this!
>
>It's wrong. It is mathematically not possible to know if future results will go
>in the direction you wish, so it is not possible to know if NOW is the right
>moment to stop in order to get the result you wish to have or if it is better to
>wait.
>
>They stop from time to time and publish the results, and that's perfectly OK.
>
Let me please cut the topic in different parts because it becomes clearer what I
mean. This is the first part. Interesting for me that you supported my theory
although you wrote as if you could oppose me with your description. Let me at
first repeat the SSDF itself is not claiming to do science. So, I will skip your
thesis that it is mathematically not possible to know future results. Because
that is possible of course in SSDF, but not necessarily mathematically. But from
the design of the practice I would know it if I were a staff member. And that is
exactly may point of critic!
Another point. They do stop at a certain moment. And this time here they stopped
when Fritz 7 was 8 points ahead!! Don't you have a problem with it?
Let me say something to your mathematical argument. I don't need maths if I know
from chess that Fritz will dominate XY by a percentage of NM. Or other
knowledge. We know that a lot of pre-testing has been done.
What I wanted to discuss is the point that the fact of these _8_ points alone,
is normally sufficient that I could _not_ stop the tests, in combination with
the big margin of error of course. 30 or some is too big. I write stopping the
tests, but this is not what happens, they _publish_ the results of that moment.
IMO this is wrong! The results make no sense if we talk about a ranking list.
Ranks. We simply can't differentiate ranks with 8 points difference and 30
points error.
Now you could argue, that they stopped when it must have stopped because of the
dates. But then I want to know the details when they started. You see? Nowhere a
real control of the factors. Now I know, it's unfair to beg for more
exactitudiness. Simply because they are amateurs in SSDF. I know all these
arguments.
============
IMO we make the wrong conclusions. It's all or nothing. If I make critics, I am
told then try to make better testings. For me as a neutral critic this sounds as
if SSDF could say, if that sort of critic can not be stopped we won't continue.
Because no matter where I tried to discuss that in peace, I was attacked this
way. With threats and unsensitive proposals. Because it's simply not true that a
good critic must be ready to do the work he criticized.
When I write a critic against a certain point, I do this in the hope that this
point could be changed to the better, here in the system of SSDF. But you could
see how someone who had constructed with others the basics of SSDF has already
reacted. It's a pity. If you would say that such critics are harmful for them
and depressing, and that I should better stop the critics, I would stop of
course. But you can't expect me to take such arguments as successful against my
critic. We should call it what it is, excuses.
Rolf Tueschen
This page took 0.04 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.