Author: A. Steen
Date: 00:07:08 11/21/05
Go up one level in this thread
On November 21, 2005 at 01:56:36, Aaron Gordon wrote: >On November 21, 2005 at 01:26:29, A. Steen wrote: > >>>>I am on the side of EGTBs. >>>> >>>>Best, >>>> >>>>A.S. > >If this is the case, why not just agree and be done with it? Now this is bedevilment indeed! Say I believe that most gorks are deeks. I hear someone saying to the interested world: "Most gorks are deeks." Good so far! But then she says: "Here, as an example, is a gork I randomly chose who is unsurprisingly a deek." And when I look, I see the claimed gork is not a gork, or worse, it is a gork but is not a deek. Should I stay schtumb? >You seem to be >overly aggressive here for apparently being a supporter of EGTBs. Follow the thread in chronological=posting sequence, and you will see I am (if anything) the victim. >I believe in >his main post he was saying EGTBs are helpful, which IS the case. You can twist >this any which way you'd like it, but when you brush off the BS it all comes >down to the fact of EGTBs being beneficial. Often, very probably usually - but certainly not always. Referring back to my nomenclature, Type (b) finds are often superior as they are cheaper in time terms. Types (c) and (d) can be dangerous, but only if they result in shuffling cycles as demonstrated in the subsequent example the OP provided. >Also, to your comment about the move being as fast as the EGTB access itself. >What you left out is that the EGTB move would have been played instantly, You must be one of these "troll" types the OP referred to, right? A quick glance at my EGTB folders tells me that the average file size for an EMD tablebase is about 100Mb (counting multiple files for the same tablebase as one file), and I haven't yet got far enough into the sixes, each one of which raises the average. Cacheing is thus only a small benefit, as successor EGTBs (sometimes, when post-promotion as opposed to post-capture, about as large) also have to be stored. What do these evidently non-mechanical big HDDs you use cost? How many others use them? Can I have some? >where >the engine without EGTBs will end up thinking and wasting valuable time. In his >case with Fruit, not only did it waste time thinking, it also shuffled pieces >around for a period of time. With EGTBs this would have been avoided. It didn't shuffle in the example provided of them shuffling. Here is the example the OP provided and to which I objected: [D]8/2k5/8/3R4/8/8/5K2/8 w - - 0 1 Best, A.S.
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.