Author: James Swafford
Date: 09:39:50 10/14/02
Go up one level in this thread
On October 14, 2002 at 12:09:25, Uri Blass wrote: >On October 14, 2002 at 10:06:39, James Swafford wrote: > >>On October 14, 2002 at 09:44:06, Vincent Diepeveen wrote: >> >>>On October 14, 2002 at 07:05:42, Dave Gomboc wrote: >>> >>>>On October 13, 2002 at 15:35:12, Vincent Diepeveen wrote: >>>> >>>>>On October 13, 2002 at 14:40:12, James Swafford wrote: >>>>> >>>>>I wonder why. there is just one person ever in this whole >>>>>planet who said 12(6) = 18 ply, and that's robert hyatt. >>>> >>>>Bullshit -- I said this years ago, after asking a member of the DB team. >>> >>>>Hsu just said it himself too: >>> >>>>EeEk(* DM) kibitzes: next question: Seems like people wants to know >>>>what the exact meaning of ""12(6)" in the Deep Blue log files, can you >>>>explain this? >>>>CrazyBird(DM) kibitzes: 12(6) means 12 plies of brute force (not >>>>counting the search extensions & quiescence). >>> >>>by definition this means: it was 12 ply not counting search extensions >>>and qsearch. of course hardware search is part of that. >>> >>>>CrazyBird(DM) kibitzes: 6 means the maximum hardware search depth >>>>allowed. >>> >>>that's part of that 12 ply. If you interpret it different then that's >>>your problem and wrong. >>> >>>'brute force search' is everything including hardware. In fact hsu >>>even specified it for the technical gifted that it is just >>>EXCLUDING qsearch and excluding extensions. >>> >>>>CrazyBird(DM) kibitzes: this means that the PV could be up to 6 plies >>>>deeper before quiescence. >>> >>>*possibly* if it prints (6), the logs clearly show that 5(6) doesn't >>>show a 5 ply mainline. but < 5 >>> >>>>Apparently that's not clear enough for you? The PV can be up to 6 plies >>>>_deeper_. Deeper than what? Than the software search of 12 plies, of course! >>> >>>Which country do you get from, his statement is very clear. >>> >>>Later it was repeated even to show the full logfile here not leaving >>>away a single line after when the question is asked: >>> >>>---- >>>EeEk(DM) kibitzes: kib question from ardee: Does "12(6)" mean 12 total ply or >>>12+6=18 total ply? This has the been source of huge arguments for years! >>>aics% >>>CrazyBird(DM) kibitzes: 12 total in terms of brute force. 6 is just the max >>>partition in hardware. >>>aics% >>>EeEk(DM) kibitzes: question from parabola444: You mentioned Deep Blue searched >>>about 12 plies brute force + extensions, which is similar to what pc programs >>>these days get on a fast pc - since Deep Blue hardware was much faster, how come >>>it didn't search significantly deeper ? >>>aics% >>>CrazyBird(DM) kibitzes: to all the book readers, if you do like the book, please >>>tell your friends would might be interested. thanks. >>>aics% >>>---- >>>So it is clear his answer was very clear : it is 12 ply. >>> >>>So is it clear now? >>> >>>Apart from that. For 18 ply you would need to see squareroot(40^18) >>>nodes to search it using fullwidth. That's a minimal tree, excluding >>>qsearch and other overhead. >>> >>>You really believed all that time that >>>deep blue searched 40^9 = 262144000000000 nodes ? >> >>Where do you get 40? Is that something they published, or you came >>up with? I think 35-36 would be more accurate. > >35^9 is also something that they could not search. >I assume it is 35 if you do not like 40. >> >>Also, you're assuming they aren't doing any decent move ordering. >>You know as well as anyone it's possible to bring your effective >>branching factor to something less than sqrt(bf). > >No > >sqrt(35^18)=35^9 is after you have perfect move ordering >and it is known that nobody has perfect move ordering. You're right... 35^9 is the minimal tree, _assuming_ you're not hashing or pruning. -- James > >Hash tables can help but it seems that they do not use it effectively and >even if they could be 10 times faster thanks to hash tables >35^9/10 is still impossible. > >Uri
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.