Author: Dave Gomboc
Date: 16:46:25 06/10/99
Go up one level in this thread
On June 10, 1999 at 19:36:38, KarinsDad wrote:
>On June 10, 1999 at 18:43:50, Dave Gomboc wrote:
>
>[snip]
>>>
>>>You are twisting my words Dave (and yes I saw the wink).
>>
>>I'm not twisting your words. You said to go for the shortest draw. After the
>>moves 1.Nf3 Nf6 2.Ng1, if 2...d5 (and all other moves) lead to at best a draw,
======================================================
>>then 2...Ng8 will be the preferred move, to achieve the draw ASAP.
=========================================
>
>I do not remember ever saying that you should go for the quickest draw in a
>position which may be won (with perfect play). Please tell me the message in
>which I said that.
I did not say you said that. Above, where I said "you said to go for the
shortest draw", the "when the position is drawn" was intended to be implied.
>>Personally, I would think that trying for the longest draw (hoping that the
>>opponent would make a mistake) would be better. But neither criteria is as good
>>as imitating a fallible opponent and heading for positions that they will blow
>>it in, so it's a moot point.
>>
>>>I said that the quickest draw is best in a drawn position (and only due to
>>>reasons of decreasing the size of a perfect tablebase). Since I do not think the
>>>opening position is drawn with perfect play and I obviously think that 1. Nf3
>>>Nf6 2. Ng1?? d5!! is a win for black with perfect play, then I think that 1. Nf3
>>>Nf6 2. Ng1?? Ng8?! in an attempt to go for a draw is dubious and nowhere near
>>>perfect play for black (why attempt to equalize the position by removing
>>>development when you can improve the position by developing?).
>>>
>>>KarinsDad ;)
>>
>>Your view that one tempo from the start position gives a won game for White is
>>very extreme, particularly when consulting modern opening manuals, in which
>>Black is reported as having many routes to equality in several different opening
>>systems.
>>
>>Dave
>
>My view is that this MAY occur given perfect play. I do not say that it WILL
>occur. I do not know any more than you do.
>
>I think that just the original extra tempi for white of having first move MAY
>give white a win with perfect play.
>
>I feel stronger that giving one side 2 extra tempi (such as black after 1. Nf3
>Nf6 2. Ng1) has an even greater chance of resulting in a win for that side with
>perfect play.
>
>Attempting to go for a draw by repetition from the first move is ONLY good for
>white if perfect play results in a win for black and is ONLY good for black if
>perfect play results in a win for white. However, if either of these 2 cases was
>in fact what would happen with perfect play, then there would be NO chance to
>get the draw since perfect play by both sides would result in a win for one of
>them.
No, it's also good if perfect play results in a draw, neither side has a winning
advantage after 1.Nf3 Nf6 2.Ng1, and you optimize for the shortest draw (which
is what I originally asked, if you look back far enough.)
>Hence, it only makes sense to attempt to go for a draw if the starting position
>is a drawn position with perfect play. However, using the route 1. Nf3 Nf6 2.
>Ng1 appears to fairly weak for white since black gains two tempi. Therefore, the
>most likely forced drawn position probably takes many ply to get to. It seems
>likely that white would make it's best move, and then black would make it's best
>move, etc., until the draw occurs many many many moves later. Neither side would
>make inferior moves such as Ng1. The game might require 600 ply or more. Who
>knows? But, it seems apparent that a side would only go for the quickest draw
>ONCE it gets to a position where a quick draw can be forced. The opening
>position is not that type of position.
Ng1 isn't an inferior move if neither player can win and they both seek the
shortest draw, which is precisely the condition I was describing. Then it
deserves a double exclam.
>However, I do agree with you that a program may go for a slow forced draw or a
>complicated draw if there is plenty of time remaining, just in an attempt to get
>the other side to mess up.
>
>If one side or the other had a forced win with perfect play from the opening
>move, then that side would never go for a draw. So, white would never play 2.
>Ng1 if it had a forced win and black would never play 2 ... Ng8 if it had a
>forced win.
Yes, I agree.
>I will now drop out of this discussion since it cannot be proved one way or the
>other and it seems that neither you nor I will budge from our position. Nice
>talking to you about it though.
Well, you're free to do as you wish.
>KarinsDad :)
Dave
This page took 0.01 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.