Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: ICC Green List - Jan 31

Author: Vincent Diepeveen

Date: 17:02:58 01/31/00

Go up one level in this thread


On January 31, 2000 at 15:25:48, Dann Corbit wrote:

>On January 31, 2000 at 15:13:37, Will Singleton wrote:
>[snip]
>>-------------------------------------------------------------------------
>>                       Active Accounts -- Standard
>>-------------------------------------------------------------------------
>>ACCOUNT     STANDARD  R(mhz) |  WIN%  GP/WK  TGP  |  HIGH    LOW   ACTIVE
>>
>>PostModernist   2452    2482 |   59%     70  147  |  2474   2442       14
>>Beadle          2405    2475 |   64%     61  128  |  2418   2378       14
>I am very impressed by these two showings.  Higher than some commerical programs
>on heavy-duty hardware.  Also a fairly large number of games have been played,
>indicating that it is rather unlikely to be a fluke of some kind.
>
>I would like to see the formula for opponents for these machines if possible.
>
>>JudgeTurpin     2401    2471 |   43%     18   39  |  2404   2391       14
>>Talmoves        2427    2427 |   55%      4    9  |  2443   2425       14
>>Galahadx        2381    2422 |   50%      0    2  |  2381   2296        0
>>MissSilicon     2386    2416 |   50%      1    4  |  2397   2386       14
>>gambitmaster    2382    2401 |   20%      4   10  |  2382   2356        0
>>RedBear         2334    2375 |   42%     10   21  |  2364   2334       14
>>GnuChess        2347    2366 |   83%     17   36  |  2377   2298       14
>>GKJunior        2311    2366 |   73%     27   57  |  2311   2177       14
>>Amateur         2305    2324 |   57%     54  114  |  2432   2243       14
>Amateur really took it on the chin as far as ranking in this list.  Do you know
>what happened?  Was it some kind of trojan horse attack, or was it competition
>from a new source that had Ameteur's number or ???

Amateur got milked by several accounts. DiepX for example won a 75 10 game.

Note that searching at icc for amateur doens't work as it's rating is
too low to store it in library idem for diep i think,
but a single loss already is worth 16 points. A few of those losses,
we talk about just needing like 5 of those losses and you lose already
5x16 = 80 rating points which means a lot for any ranking.

ICC ratings are very fluctuating because they use K=32,
where according to FIDE k at 2300 would be ==> (2600-2300) / 20 = 15

So rating win/loss goes more than twice as fast as at fide.
For programs above 2400 rating loss/win goes more than 3.2 times faster
than at FIDE, as K factor is above 2400 at fide: 10.

for those who don't know what a k factor is:
it is the number of points you can at maximum lose.
So playing against an opponent with about the same rating (within 2 points
difference) means you need to score 50% against him,
  0.5 x k = 0.5k

So a 2300 winning
against 2300 loses at ICC  16 points.
2300 at FIDE loses 7.5 points.

a 2400+ against a 2400+ (both same rating) has 5 points at risk,
at ICC it's still 16 points at risk....

Quite a big difference!

>>BugChess        2284    2314 |   92%     48  100  |  2297   2270       14
>>ArasanX         2222    2263 |   33%      2    6  |  2222   2208       14
>>EXchess         2224    2243 |    0%      0    1  |  2262   2224        0
>>Tinker          2222    2202 |   81%     55  115  |  2222   2033       14
>>Hossa           2155    2174 |   37%     40   85  |  2259   2155        7
>>stormx          2065    2065 |   55%      8    9  |  2065   2046        7
>[snip]



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.