Summit Supercomputer: the strongest in the world ?
Ah ... ctrl V. And Windows key + V to access clipboard history. Good.
From: https://www.rankred.com/fastest-supercomputers-in-the-world/ 
at one meg per position - going to take that guy 50 million years ! 
Even if you eat shark cartilage and watch Jimmy Fallon ...
you're not going to make it ...
Chess will never be solved, here's why
Sort:
Jan 26, 2022
0
#262
Jan 26, 2022
0
#263
tygxc wrote(#292 and many others):
A scientific paper shows there are 3.81*10^37 legal chess positions. ...
Er, No it doesn't. This has already been pointed out to you.
If as you say you plan to solve the competition rules game it's about 250 (American) billion times that number.
These bad figures tend to spread themselves around the internet. Could you stop repeating them please?
You should say, "A scientific paper shows there are 3.81*10^39 legal chess positions under competition rules ignoring the triple repetition rule and about 10⁴⁸ positions I don't reckon to be important".
Jan 26, 2022
0
#264
@MARattigan
I'm wondering if the image in post #297 came out well ...
A small favor if you will ...
does it look good on your screen ?
Jan 26, 2022
0
#265
@playerafar
Looks great.
But I liked the 360/50 with the flashing lights.
Jan 26, 2022
0
#266
MARattigan wrote:
@playerafar
Looks great.
But I liked the 360/50 with the flashing lights.
Lol ! Thanks !
That 360 was from 1967 !
Jan 26, 2022
0
#268
playerafar wrote:
e4 f5 is the Fromm gambit I believe. Its considered unsound.
You misremember. 1. f4 e5 is the From gambit. It is considered suboptimal, but may (probably is) sound. 1. e4 f5 is MUCH worse for black and may be dead losing, but I do not believe this has been proven.
Whereas the King's Gambit is considered 'sound' by GM's if I've got that right.
Theory is pretty even, but engines seem to quite strongly prefer black.
But many GM's would prefer the black side of it?
A lot more allow it as black than play it as white.
I'm getting that impression too.
But that preference could depend in turn on who the opponent is ...
Elroch wrote:
playerafar wrote:
e4 f5 is the Fromm gambit I believe. Its considered unsound.
You misremember. 1. f4 e5 is the From gambit. It is considered suboptimal, but may (probably is) sound. 1. e4 f5 is MUCH worse for black and may be dead losing, but I do not believe this has been proven.
Whereas the King's Gambit is considered 'sound' by GM's if I've got that right.
Pretty equal, I believe.
But many GM's would prefer the black side of it?
Preference is partly subjective.
I'm getting that impression too.
But that preference could depend in turn on who the opponent is ...
May you unblock me?
Jan 26, 2022
0
#272
Elroch wrote:
playerafar wrote:
e4 f5 is the Fromm gambit I believe. Its considered unsound.
You misremember. 1. f4 e5 is the From gambit. It is considered suboptimal, but may (probably is) sound. 1. e4 f5 is MUCH worse for black and may be dead losing, but I do not believe this has been proven.
Whereas the King's Gambit is considered 'sound' by GM's if I've got that right.
Theory is pretty even, but engines seem to quite strongly prefer black.
But many GM's would prefer the black side of it?
A lot more allow it as black than play it as white.
I'm getting that impression too.
But that preference could depend in turn on who the opponent is ...
You're right on that one Elroch. About the Fromm.
Good thing I said 'I believe' ... I wasn't sure.
I inverted it with the other position.
And - I didn't bother to check first Its been a lot of years because neither one gets my attention.
e4 f5 hardly even an opening - or maybe it has a name.
Update: Yes ! Duras Gambit.
There are all kinds of silly ways for black to respond to various white moves.
1) b4 c5 ?? but maybe that has a name.
or
1) c4 d5 ...cxd. Has a name? Yes I'll check.
but ... if now Nf6 e4 Nxe4 ??
then how about Qa4ch ... Oops! Death of knight ...
yeah I'll 'check' it. It is after all - a check.
Never gonna give you up, never gonna let you down NeVeR gOnNa SaY i LiKe LiChEsS
Jan 26, 2022
0
#274
Black loses his knight.
does this mean c4 d5 should not be included in 'positions to be 'solved' ' ?
I would say no. Black does better if he recovers his pawn right away with Qxd5 ... in which case white can come out on the queen anytime with Nc3 right then or soon.
But black does much better after c4 with c6 (imho) first ... with d5 to follow next move.
But that doesn't mean that c4 d5 should be dismissed as 'solved'.
Dun dun
Dun dun dun duuun duuuun
Jan 27, 2022
0
#277
@Elroch
e4 f5 does have a name. Its called the Duras Gambit.
A particular analysis board on this site initially says it gives white an advantage of about 2 points if white immediately takes the pawn.
And c4 d5 has a name too.
Its called the Anglo-Scandinavian gambit.
After cxd black could continue with c6 for example.
That's considered better than Qxd5.
Point: none of these positions should be considered 'solved' by the computers.
#312
We know that 1 e4 f5? loses for black, although it is not worked out till checkmate.
We know things without a full proof.
Jan 27, 2022
0
#279
tygxc wrote:
#312
We know that 1 e4 f5? loses for black, although it is not worked out till checkmate.
We know things without a full proof.
Its still not 'solved'. There's a lot of material on the board - many ways for white to go wrong.
Remember - a GM can spot a 'C' player a rook and win. He has a methodology the other player does not have.
Its chess - not checkers.
Jan 27, 2022
0
#280
You'd do much better to analyze from the simplest endgames - where indeed positons are completely 'solved'.
Very thoroughly known - even without computers.
I believe computers recently found a win for a Queen against two bishops.
But at the human level - its hardly 'solved'.
Anyway - even if you could somehow 'cutdown' to 10 to the 27th power possible positions ... how abou 50 million years for the world's strongest computer - to give one meg's attention to each of those ?
The game was designed with mysteries and unknowns and skill levels in mind. And as a pastime.
It survives as such.
Decades ago - many players were fearful that computers would destroy chess.
Instead - ironically - chess has done Better not Worse
#294
Slice the whole of chess A00 to E99 into 500 chunks, e.g. C67.
Then solve the relevant chunks one by one.