Author: leonid
Date: 09:11:46 10/28/01
Go up one level in this thread
On October 28, 2001 at 11:18:46, Paul wrote: >On October 28, 2001 at 10:25:51, leonid wrote: > >>On October 28, 2001 at 08:37:47, Paul wrote: >> >>>On October 28, 2001 at 07:29:06, leonid wrote: >>> >>>>On October 27, 2001 at 17:36:53, leonid wrote: >>>> >>>>>On October 27, 2001 at 16:52:40, Paul wrote: >>>>> >>>>>>On October 27, 2001 at 13:06:43, leonid wrote: >>>>>> >>>>>>>Hello! >>>>>>> >>>>>>>I think that the next mate is just in the middle. It is not very hard to solve >>>>>>>and still it is not extremely easy. The best one when you want to be a sure >>>>>>>winner but not at big price. >>>>>>> >>>>>>>[D]bBQQ1nrq/b2NNq1k/K6n/QQqq1q1B/3Q1Q2/3r1qR1/3q2q1/3RQQq1 w - - >>>>>>> >>>>>>>Please indicate your result. >>>>>>> >>>>>>>Thanks, >>>>>>>Leonid. >>>>>> >>>>>>Hi Leonid, >>>> >>>>Hi, Paul, >>>> >>>>>>since Pretz is in the garage at the moment for a revision of the engine, I >>>>>>thought I'd go over and comfort her with your new position. But I think the >>>>>>engineer broke something, because it doesn't find a mate (in 12) very quick: >>>> >>>>This position is mate in 12. Verified it at night by searching 11 moves by brute >>>>force. Took slightly more time that expected, since branching factor jumped >>>>after 10 moves. Three hours and 45 min. 12 was found already by selective. >>>> >>>>Cheers, >>>>Leonid. >>> >>>Ok, so a mate in 12 it is! I only checked until depth 9 with Chest and estimated >>>it would take about an hour to prove that there is no mate in 11, so I stopped. >>> >>>I haven't been able to come up with a faster time ... almost 8 minutes for mine >>>to find the mate in 12. >> >>The most important is that your program find mate in 12. This signify that there >>is no bug. Some time selective will not reach minimal depth and sometime it do. >>At least, it is what I could see with mine. It found in few positions mate in >>much higher level that other programs, including your. > >That's because you search more selective than I do ... from what I understand >from your program, if you would search more moves than just checks, you also >would find shorter mates ... you could try? It is true. Actually, my selective (and I expect that this is the same for every other program) search only promissing moves. Promissing moves are those moves that give quick response in most of cases. Not necesserily the shortest response. Checking moves are the most promissing. Moves that affect vital king's lines are the next, and so like. Between many variations of "promissing moves" I left only checking because they provide high simplicity of code and no need for big look-up tables. After having the chance to write my 64 bits program (if it will ever happened) I, probably, will look once again into my code to see if old ideas still have sense. With each big speeding old trick for "speed against depth" become useless. >>If you will have some free time, try this position on some other good program >>that search well with selective. Say me what time you found. I have only Mate >>2.0 inside of Hiarcs but this engin do brute force only. > >I have no other programs installed on my comps anymore, just Pretz & Chest, so >can't try anything, sorry! I'm also curious how fast other programs found this >mate in 12, to see where I am with my Pretz ... For now I see only your result here. Maybe others don't have that much competitive numbers to offer. I still hope that somebody will come with Chess Master. Its results in some positions are impressive. I only have somewhere my old Chess Master 4000. Program is just too old to try, even if I still can find it. Last version (after few responses that came here) is more expedient in selective mate. Cheers, Leonid. >>Your result in today position is excellent! You found minimal depth with >>selective. I did the same and at almost identical time with you. Mine is 2.8 >>sec. >> >>Cheers, >>Leonid. > >Groetjes, >Paul
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.