Only Chess
19 Dec 10
23 Dec 10
1 edit
Hi V
May dig out a CC game see what the box thinks.
This bit:
" but that doesn't stop me looking at the solution given by the computer
and then working backwards, filling in some human reasoning."
Do you not have this in the wrong order. Human first - then computer.
I look at what the players played. If happy, no need for a box.
23 Dec 10
Originally posted by greenpawn34Don't take this out of context - this is in the context of your "extremely complex" games that have been "ripped to pieces over days/weeks ". So the assumption is that the annotator can't easily reproduce the player's thoughts. Let's face it, if it takes Umansky 2 weeks and 10 pages of detailed analysis to decide that one move is better than another, no annotator is going to explain that after 5 mins of shuffling the pieces around on their own. We're not talking here about moves based on general principles or simple tactics, but rather extensive and deep calculations.
" but that doesn't stop me looking at the solution given by the computer
and then working backwards, filling in some human reasoning."
Do you not have this in the wrong order. Human first - then computer.
I look at what the players played. If happy, no need for a box.
23 Dec 10
Originally posted by VarenkaThose extensive and deep calculations are only going to happen occasionally, maybe once or twice in any game. You reach a point where such calculation is needed and then spend several days trying to find the truth in the position. Once done, it likely doesn't need to be repeated unless the opponent finds something that you missed.
Don't take this out of context - this is in the context of your "extremely complex" games that have been "ripped to pieces over days/weeks ". So the assumption is that the annotator can't easily reproduce the player's thoughts. Let's face it, if it takes Umansky 2 weeks and 10 pages of detailed analysis to decide that one move is better than another, no an ...[text shortened]... ed on general principles or simple tactics, but rather extensive and deep calculations.
23 Dec 10
Originally posted by DiophantusThe top CC games can take years to play. So if they only have to spend several days "once or twice in any game", can you tell me what is taking the time? I realise they play multiple games in parallel, but it still doesn't come close to your assessment of how much calculation is needed. Where is the time spent?
Those extensive and deep calculations are only going to happen occasionally, maybe once or twice in any game. You reach a point where such calculation is needed and then spend several days trying to find the truth in the position. Once done, it likely doesn't need to be repeated unless the opponent finds something that you missed.
23 Dec 10
Originally posted by VarenkaThe time is often spent doing the other things in life, working for example. It has to be remembered that CC is still essentially amateur and even the top guys have jobs to go to. Even the "several days" will not be several days solid work on that one position, it might well be a couple of hours a day spent looking at that game over several days. Sometimes the time is spent looking at other games or even doing nothing! For more on this sort of thing you could do worse than have a look at Tim Harding's Winning at Correspondence Chess, especially the chapter entitled "The Correspondence Player's Armoury". Much of that chapter deals with time and how to use it.
The top CC games can take years to play. So if they only have to spend several days "once or twice in any game", can you tell me what is taking the time? I realise they play multiple games in parallel, but it still doesn't come close to your assessment of how much calculation is needed. Where is the time spent?
23 Dec 10
Originally posted by DiophantusI think you grossly underestimate what it takes to win the CC world title. There are quite a number of retired people who devote large amounts of time to playing CC. Van Oosterom even hired GM Jeroen Piket as his personal secretary! I’ve also chatted to Kenneth Frey who made it into the ICCF top 10 at one point; again, retired and devoting large amounts of time to CC.
23 Dec 10
Originally posted by VarenkaSo ask them what they do with their time! I've told you what I do with mine which seems to accord reasonably well with how others use their time and the advice given by Harding in his book.
I think you grossly underestimate what it takes to win the CC world title. There are quite a number of retired people who devote large amounts of time to playing CC. Van Oosterom even hired GM Jeroen Piket as his personal secretary! I’ve also chatted to Kenneth Frey who made it into the ICCF top 10 at one point; again, retired and devoting large amounts of time to CC.
23 Dec 10
Originally posted by DiophantusI know what they do with their time: lots of extensive and deep analysis. Harding wasn't writing a book for World CC champions, just as what RHP players do isn't exactly representative of what Umansky did either.
So ask them what they do with their time! I've told you what I do with mine which seems to accord reasonably well with how others use their time and the advice given by Harding in his book.
- Joined
- 20 Jan 07
- Moves
- 24091
24 Dec 10
1 edit
Originally posted by greenpawn34Q: 0.89. (what does that mean?)
Hi V
"It’s highly likely that all of the top ICCF players make heavy use of
computers in order to decide on their moves."
Most likely they do. But this lad was World Champion in 1996.
Computers have advanced incredibly since then.
An anotator should try to explain what were the players up to.
(Not what a modern computer considers is the best the pole and dump
you in the duck pond with an evalutaion on 0.89. (what does that mean?)
A:That's very nearly a ninth of a pawn advantage GP.
24 Dec 10
So its' 8/9ths of a pawn. So that means somewhere on the board
the computer has seen 1/9th of pawn as compensation.
I wonder what 1/9th of pawn compensation looks like?
I've always disliked how a box classes all positions with numbers.
It should just use smiley faces.
😏 (Mate in x number of moves)
😀 (winning easily)
🙂 (winning)
😉 (better but no clear advantage)
🙁 (losing)
😠 (Getting Hammered)
😳 (Totally lost - resign now)
😕 (unclear)
😲 (Look out there is a trap about)
😴 (postion is drawn)
26 Dec 10
Originally posted by greenpawn34Good job Greenman
So its' 8/9ths of a pawn. So that means somewhere on the board
the computer has seen 1/9th of pawn as compensation.
I wonder what 1/9th of pawn compensation looks like?
I've always disliked how a box classes all positions with numbers.
It should just use smiley faces.
😏 (Mate in x number of moves)
😀 (winning easily)
🙂 (winning)
😉 ...[text shortened]... - resign now)
😕 (unclear)
😲 (Look out there is a trap about)
😴 (postion is drawn)
- Joined
- 18 Jan 07
- Moves
- 12466
26 Dec 10
Originally posted by greenpawn34So, instead of having it grade positions in tenths of a pawn, you'd have it grade them in elevenths of a smiley? Granted, it fits in perfectly with the Imperial measurement system, but I fail to see how this is less of a numerical measure...
Forgot one
😛 (Only Chandler would play this play this crap.)
Richard
