Author: Howard Exner
Date: 12:00:07 04/28/98
Go up one level in this thread
On April 28, 1998 at 11:18:43, Bruce Moreland wrote: > >On April 28, 1998 at 01:25:23, Howard Exner wrote: > >>I believe these ones can be safely dropped from ECM. >> >>They have one of the following wrong with them: >> >>1. Two equally good key moves that make for a problem that is much too >>easy >>2. An outright cook in the key move that is found quickly >>3. A faulty defense (book error) that make other key moves just as valid >>4. The key move leads to nothing really decisive making it not fit the >>traditional understanding of what a combination is (also because of a >>error in the book). >> >>14 16 17 19 28 29 31 43 51 54 63 68 70 71 76 81 84 91 96 99 >>109 119 131 137 144 145 148 153 180 186 190 193 198 >>208 215 227 228 231 259 270 274 292 293 298 >>300 310 312 313 315 321 330 338 345 349 356 366 377 381 390 396 397 >>411 420 425 434 440 447 462 464 467 474 489 494 497 >>501 504 505 510 525 539 557 559 560 561 562 592 593 >>624 627 634 653 677 683 686 691 693 >>701 702 730 735 754 755 758 767 >>811 827 836 846 858 860 864 866 865 867 868 875 879 > >The most likely mistakes in this list are 730 and 846. > >There are some others that I'm not comfortable with, either, but it's >hard to challenge them. > >I think there are also some more cooks in the "correct" ones, but this >has gotten so scattered out that it is hard to make any sense about >this. This wasn't the right place to do this whole project, I think, it >would have been better to do it via email. > >It is hard to know exactly what to do about problems like position 84, >when you score the position as +1.75 on a non-key move, then you find >the key move at +2.2, and over the next hour it rises to +2.3 or >something like that. In reviewing the posts on determining which positions to delete I eventually had to make a call. For #84 the majority of the testers found the other move cxb5 close to the key so that is why I dropped it. The original move I believe to be better but went with the consensus. #730 I agree should not be deleted (my oversight) #846 I felt fit the not decisive enough description. It is though probably the best move. In the end I'm ok with leaving all these in the suite. How do the other particpants in this venture feel about these deletions? I'm thinking if we miss a few that no real harm is done. Fixing chess problems is no small task at any level. Once this "deletion thread" runs its course I'll post the entire modified set to Manfred and ICD. The next step will be for you and the other programmers to agree on a time which defines which ones to delete on the basis of too easy. The past posts on this suggested both 10 seconds and 20 seconds but was never resolved. A simple vote could determine which time to mark as "too easy" to keep. I don't mind posting here as it allows anyone interested to jump in with ideas.
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.