Author: Bruce Moreland
Date: 21:03:07 02/04/98
Go up one level in this thread
On February 04, 1998 at 13:00:13, Bruce Moreland wrote: > >On February 04, 1998 at 06:01:52, Amir Ban wrote: > >>On February 04, 1998 at 03:12:35, Howard Exner wrote: >> >>>On February 03, 1998 at 15:52:30, Amir Ban wrote: >>> >>>> >>>>Yet another chapter in the ECM saga: >>>> >>>> >>>>No. 411 (Qxd6): >>>> >>>>I get the simple gxh7. Trying Qxd6, it gets about the same score (+3.5). >>>>The position is already very unbalanced anyway, so there may be more. >>>> >>>>Recommend: Scrap it, or add gxh7. >>> >>>Rebel 8 gets 3.16 for Qxd6 and 2.26 for gxh7. My subjective view >>>is that Qxd6 is more decisive. >>>> >> >>Perhaps so. Qxd6 ends up an exchange and a pawn up with the queens >>traded. gxh7 is two pawns up, pawn on seventh and black king stripped. >>Rebel is conservative. > >Mine takes Qxd6, but I don't want to get involved in arguments about >positional chess. > >>>> >>>>No. 422 (f4): >>>> >>>>I find Qh4 (+92). f4 gets about the same, but I can't find a convincing >>>>continuation after f4 exf4 e5 (or Rb6). >>>> >>>>Recommend: Vague. Drop it. >>> >>>Book is 1.f4 exf4 2.e5 fxe5 3.bxg6 Bc8 4.Rb6 1:0 Is there better choices >>>for black? >>>> >> >>2... f5 looks better, though white is happier in this position. The >>a-pawn gives black counterplay. > >I get 1. f4 in a couple of minutes, at +0.97, going up to +1.37 after 42 >minutes. > >My line is just a strong attack. > >I'm worried that interesting positions like this may be cut even if they >turn out to be sound, just because we can't figure out how to win them. > >>>>No. 425 (Ng4+): >>>> >>>>The Ng4+ idea works, but I prefer the intermediate Bf5 Qg2 Ng4+. >>>> >>>>Recommend: Add Bf5. >>> >>>Agree > >This is a transposition a tempo later, right? > >>>>No. 427 (Ne2+): >>>> >>>>Perhaps I have the position wrong. Black is already a queen down and it >>>>looks hopeless. I find Rxg4 for the exciting score of -1209. >>>> >>>>Recommend: Find what the right position is. >>> >>>The Black Queen is on e6 but the key move might want to be changed >>>to Rxg4. > >I get Rxg4 with a nice jump here, but I didn't run it longer than 20 >seconds. Howard may very well be right though. At the least it's >cooked. > >>>>No. 434 (Ncxd6): >>>> >>>>This is fairly difficult. Nbxd6 is a transposition. >>>> >>>>Recommend: Add Nbxd6. >>> >>>Agree > >They are both +4, so it's either cooked or transposed or both. > >>>>No. 440 (Rf3): >>>> >>>>It looks wrong: Rf3 Bd1 and black has egg on his face. >>>> >>>>Recommend: Drop it. >>> >>>Agree as the book missed your Bd1 defense. > >I don't find this in an hour on a P2/300. > >>>>No. 462 (Nxg6+): >>>> >>>>Wrong position again. I get Qxh7 mate. Mate-in-one problems are rather >>>>easy, for me at least. >>>> >>>>Recommend: Find what the right position is. >>> >>>Remove the white rook on h6. However the move Nxg6 looks too easy. >>>Rebel 8 sees a mate in 0 seconds. > >Easy problem, yes. > >>>>No. 464 (e5): >>>> >>>>I cannot find the justification for e5. If the idea is e5 dxe5 Ne6 then >>>>counting material actually black gains by it. >>>> >>>>Recommend: Investigate. >>> >>>after black plays dxe5 the book continues with Nf3. Black's defense is >>>not right and I agree that this is a vague position. Drop it as white >>>has >>>other good first moves. > >Not found in 4 hours on an alpha. > >>>>No. 467 (Rf6): >>>> >>>>I don't see a winning line after Rf6 Qf8 or Ng5. Perhaps it's the best >>>>move, but the assignment is "white wins" and I don't see it. >>>> >>>>Recommend: Investigate. >>> >>>The main book line is the Ng5 defense and loses to Rxg5 but Qf8 seems >>>to hold. > >Maybe, but in my line white wins back the d5 pawn, and black is still >pretty cut up. I only ran it for an hour though, and got +1.23. > >>>>No. 470 (Nc3+): >>>> >>>>I don't think this works: 1...Nc3+ 2.Nxc3 bxc3 3.b3 Ra8 >>> >>>The book gives Qa6 here and continues 4.Rd4 Ra8 5.Ra4 Qxa4! 0-1 >>> >> >>Right, and if 4.Bd4 Bb4. So it's correct. >>>>4.Bd4 and if the >>>>c-pawn falls, what is black's threat ? If now 4...Bc5 5.Qg3. >>>> >>>>Recommend: Investigate. > >+1.23 in 22 seconds, +1.87 in 33 minutes. > >>>>No. 475 (Rh5): >>>> >>>>Probably correct, but I'm not sure. I followed 1.Rh5 Ne2+ 2.Kh1 Ng3+ >>>>3.hxg3 Qxf1+ 4.Kh2 now I see that 4...f6 loses but perhaps 4...h6 5.Bxh6 >>>>g6 holds. >>> >>>Be3 will force a mate. >>> >> >>Right. Keep it. >> >> >>>>If it's correct, it should be extremely difficult. >>> >>>Yes this will be a tough one. >>>> >>>>Recommend: Investigate. > >I ran this for four hours on the Alpha. After 34 minutes it had a draw >score. Next iteration took a huge amount of time to resolve, it ended >up terminating in a fail-high condition, failing high from the draw >score, so the score is still pretty low. I started this up on a P2/300 this morning and let it go all day. The fail high resolved at +1.15. Mine finds 6. Be3, but didn't see a mate in this search, it just won an exchange and a pawn back. Maybe it will do better in later iterations, I'll let it run until I need the machine for something else. >>>>No. 482 (Bxb5+): >>>> >>>>I get h4. Different plan, similar result (+115). >>>> >>>>Recommend: Add h4. >>> >>>Bxb5+ seems more forceful. >>>> >> >>I don't know. > >I get 1. h4 for a while and then switch to 1. Bxb5 at 90 seconds at >about +1.2. > >>>>No. 488 (Bh3): >>>> >>>>I think it's wrong: Bh3 Ne1 and black is worse than ever. >>>> >>>>Recommend: Drop it. > >I get Bh3 at +0.20 after three and a half minutes, +1.64 after 45 >minutes. The line I get is 1. ... g3, so I don't know what happens >after 1. ... Ne1. > >>>>No. 489 (Ne4): >>>> >>>>It's wrong: 1...Ne4 2.Nxe4 Rg6 3.Rf1! and if 3...Qe8 4.Rf8+! teaches >>>>black a lesson. >>>> >>>>Recommend: Drop it. >>> >>>Agree > >Not found in 4 hours on the alpha. > >>>>No. 497 (f5): >>>> >>>>After f5 exd4 I can't find anything to justify a piece (fxg6 hxg6 Bxg6 >>>>Nf8). >>>> >>>>Recommend: Investigate. >>> >>>Nf8 is the book defense and continues 4.Bxf7+ Kxf7 5.g6+ Kg7 6.Rdf >>>Bf6 7.Rxf6 1:0 >>>The side defense of 3... fxg6 4.Qxe6+ is simply denoted as +- but >>>this line might hold on. >>>> >> >>You mean 4.Qe6+. Seems so. Drop it. > >I'm running this now. Not found in 4 hours. >>>>No. 500 (f3): >>>> >>>>Looks wrong: 1...f3 2.Nxf3 Qg3 3.Be3 and where is the continuation ? >>>> >>>>Recommend: Drop it. >>> >>>The book continues with 3 ... Bg4 4.Qd6 Be5 0-1. Looks correct. Is >>>there something better than Qd6 though? >> >>No. Rd2 and Qc5 fail. It's correct. > >I'm running this now, too. This one found 1. ... f3 in approximately two hours on the Alpha. The fail-high resolved to +4.41. 1. ... Nxf2 was at +1.42 in the same iteration, so this seems to be good but either not as good or not as direct as 1. ... f3. This may be the coolest problem I have ever seen. It is solvable but the complications are fantastic. bruce
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.