Author: Peter Herttrich
Date: 03:44:13 03/04/98
Go up one level in this thread
On March 04, 1998 at 05:16:44, Kai Lübke wrote:
>On March 03, 1998 at 04:55:26, Peter Herttrich wrote:
>
>>Hello
>>
>>How about the following stupid idea:
>>
>>Clean the list. Throw out old programs. If Rebel 9 is in test,
>>throw out Rebel 8/7/6. Same with all other program.
>>Be true, no one buys an old program.
>
>Disadvantage: we don't get a real comparison between version N and
>version N+1 of a program.
In any way you will not get a real comparison, the new versions
will be testet with the new hardware (200MMX). And games between
P90 and P200MMX are without worth anyway. And i repeat: No one will buy
an old version of a program.
>
>>Split the list in classes. 200Mhz/90Mhz or so.
>>In every class you have a top-rider.
>
>Disadvantage: we don't get the (for some people important) info on how
>changing the speed influences the rating - this saves us lenghty
>discussions about "how many (ELO) points do I gain when going from 90 to
>200 MHz".
Then you have to test *all* versions on *all* hardware. Sigh...
Do you really want this?
And i guess this leads us into the old statistic discussion.
Look at the published SSDF-result
Fritz 5.0 Pentium MMX 200 MHz, 2589
Opponent
Result
Nimzo98 P200X
9½-10½
Rebel 9 P200X
14-6
MCP 7 P200MMX
26-19
Genius5 P200X
9-9
Shred 2 P200X
6½-3½
Rebel 8.0 P90
28-7
Junior 4 P90
28½-6½
SPARC 20 MHz
11½-3½
What will say us this result? Nothing, because the hardware is not
comparable in every case. Only the single-results against the same
hardware are relevant.
>
>>Split the list in commercial available programs and free programs.
>
>Disadvantage: we lose all data about how free proggies measure up
>against commercial ones (and get even more "What would be Crafty's
>position on the [commercial] list?" questions).
>
Partially agreed. Perhaps i have to think about this point again.
>>Throw out not available programs.
>
>Might be a good idea (if you mean "not available anymore" and not "not
>available yet").
Right, so i thought.
>
>>At last split the list between programs and 'Hard'-Boardcomputer.
>
>See "commercial/free split".
>
Not agreed :-) I say (IMHO), the difference between this two classes
is so wide, that there is no discussion.
>>I think the list will be shorter
>
>ALong with length, it will also lose some meaning.
>
Meaning?! Hm... i say, this list has a much less meaning then
the most of us think.
>---
>Shep
Cheerio
Peter
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.