In general, if one side can lose several moves in a complex and unclear but sharp position to produce a zugzwang, the other side can do that too and since that's the only way to win (hypothetically) it can't be done.
Sorry don't understand that either. The question is about best play, so whether or not the position is complex or unclear is irrelevant.
You're assuming that the only valid method of proof is brute force and ignorance. If you want to prove that the arithmetic mean of two positive real numbers is greater than the geometric mean, a really bad way of attempting it would be to try it out for every pair of positive real numbers.
I didn't say that a chess beginner could solve the game in 100 bytes. I said that he could write a routine to win from any winning king and rook versus king position in less than a couple of hundred bytes. I'll write you a javascript for it if you like.
For this I would have to consider maybe a half dozen generic positions, not the the circa 400,000 positions in the endgame.
You omitted to note that I would estimate less than 1K for the 1MB+ EGTB bishop and knight endgame. I could also write you a javascript for that. Again this would be based on about fifteen properties of the position rather than considering the 11 million individual positions. The compression ratio has jumped from 14 to 1000+ and I expect would continue to increase exponentially for an optimal routine as the number of positions in an endgame increases.
This is not a proof that a conceivably practicable solution of chess on these lines exists, just a conjecture, but @pfren has already pointed out, a solution by EGTBs appears to be completely impracticable simply from a storage point of view. If such a practicable solution does exist, the hard bit is finding it.
Your premise is flawed. It's right there in your first claim: "You have completely solved (Basic Rules) chess if you can produce a routine that guarantees a win from any winning position."
Sure. This would be pretty darn easy to solve if human beings just had a definition of a "winning position" that carries all the way to the starting position in chess ...
--- It's your response that is flawed. We have a definition of what is a winning position (actually two because the FIDE rules define two distinct games). In neither case is it pretty darn easy to solve.
We're already doing that (trying to prove what a winning position is), via working backwards from what humans beings and their understanding of mathematics have already determined to be the best "winning position" understanding that we have. You are extrapolating that you can define in a direct rules-related manner what a "winning position" from some opening or middlegame position.
--- I'm not extrapolating that I could do that. I'm conjecturing only that such routines exist in a form that might fit on computer storage not too remote from what is currently feasible. That such rules exist is obvious, the question is only, "what is the minimum size?". As I said, finding small routines is non-trivial. (Openings and middle games are just types of endgame.)
But that understanding does not exist, nor can you prove it without going through the same traversing we are talking about. The current method is not entirely brute force. It's already taking into account all the basic rules of chess. Brute force would be checking every position regardless of legality.
--- The understanding doesn't currently exist for all endgames. That is not to say that relatively compact rules do not exist.
Where the (human) understanding does exist, it doesn't consist of traversing large numbers of positions.
I'd still regard EGTB generation as a BFI routine. The fact that many (not by any means all) illegal positions are excluded doesn't really change that.
Can there be some ways of cutting down the traversing farther? Perhaps. But even if you do this so well that you eliminate 999,999 out of every million positions, you are still solving for 10^40. It's not even remotely in the range of possible for the reasonably foreseeable future of humanity.
--- As I pointed out, writing a routine for KRK would be very straightforward. It would solve the 400,000 or so positions in that endgame. No specific positions need be included in the routine (not even the draws - the only requirement is that it wins from winning positions).
Go ahead and write your Javascript . You might want to at least switch to assembly language, though. Javascript is hugely inefficient as a high level language, even more so than most compiled languages.
--- Most of the programs I've written have been in IBM 360 Assembler and its descendants and my storage estimates were based on that. I offered Javascript as easy to translate into whatever takes your fancy. I doubt if a user would actually notice any difference in performance.
If it's as easy as you think it is, why not go ahead and do it? You'd be famous overnight all around the world 
.
What you (and GMproposedsolution and Optimmised) are doing here is bringing a layman's understanding to a problem and then insulting the expertise of all the people that have actually taken the time and effort to work on and actually understand the problem over the past several decades.
What you offer is complete conjecture posited from an hour or two of pondering an issue from a 10,000ft perspective. It's remarkably similar to a certain someone stating that they think scientists should look into injecting bleach to combat Covid-19. As if such an obvious and simple solution were to have been completely overlooked by thousands and thousands of scientists working on the issue...
3523 etc are irrelevant since they can't be forced into.
What I was saying is that if either of those diagrams replaced the diagram in FIDE Art. 2.3 your argument couldn't be correct, because both are demonstrably won for White with best play. Therefore your argument is either invalid or makes some (unstated) assumption which is true of the standard starting position but not of the positions in posts #3523 and #3532.
If the latter, what would that assumption be? Would it apply equally to all FRC positions?
You're assuming that the only valid method of proof is brute force and ignorance. If you want to prove that the arithmetic mean of two positive real numbers is greater than the geometric mean, a really bad way of attempting it would be to try it out for every pair of positive real numbers.
I didn't say that a chess beginner could solve the game in 100 bytes. I said that he could write a routine to win from any winning king and rook versus king position in less than a couple of hundred bytes. I'll write you a javascript for it if you like.
For this I would have to consider maybe a half dozen generic positions, not the the circa 400,000 positions in the endgame.
You omitted to note that I would estimate less than 1K for the 1MB+ EGTB bishop and knight endgame. I could also write you a javascript for that. Again this would be based on about fifteen properties of the position rather than considering the 11 million individual positions. The compression ratio has jumped from 14 to 1000+ and I expect would continue to increase exponentially for an optimal routine as the number of positions in an endgame increases.
This is not a proof that a conceivably practicable solution of chess on these lines exists, just a conjecture, but @pfren has already pointed out, a solution by EGTBs appears to be completely impracticable simply from a storage point of view. If such a practicable solution does exist, the hard bit is finding it.
Your premise is flawed. It's right there in your first claim: "You have completely solved (Basic Rules) chess if you can produce a routine that guarantees a win from any winning position."
Sure. This would be pretty darn easy to solve if human beings just had a definition of a "winning position" that carries all the way to the starting position in chess ...
--- It's your response that is flawed. We have a definition of what is a winning position (actually two because the FIDE rules define two distinct games). In neither case is it pretty darn easy to solve.
We're already doing that (trying to prove what a winning position is), via working backwards from what humans beings and their understanding of mathematics have already determined to be the best "winning position" understanding that we have. You are extrapolating that you can define in a direct rules-related manner what a "winning position" from some opening or middlegame position.
--- I'm not extrapolating that I could do that. I'm conjecturing only that such routines exist in a form that might fit on computer storage not too remote from what is currently feasible. That such rules exist is obvious, the question is only, "what is the minimum size?". As I said, finding small routines is non-trivial. (Openings and middle games are just types of endgame.)
But that understanding does not exist, nor can you prove it without going through the same traversing we are talking about. The current method is not entirely brute force. It's already taking into account all the basic rules of chess. Brute force would be checking every position regardless of legality.
--- The understanding doesn't currently exist for all endgames. That is not to say that relatively compact rules do not exist.
Where the (human) understanding does exist, it doesn't consist of traversing large numbers of positions.
I'd still regard EGTB generation as a BFI routine. The fact that many (not by any means all) illegal positions are excluded doesn't really change that.
Can there be some ways of cutting down the traversing farther? Perhaps. But even if you do this so well that you eliminate 999,999 out of every million positions, you are still solving for 10^40. It's not even remotely in the range of possible for the reasonably foreseeable future of humanity.
--- As I pointed out, writing a routine for KRK would be very straightforward. It would solve the 400,000 or so positions in that endgame. No specific positions need be included in the routine (not even the draws - the only requirement is that it wins from winning positions).
Go ahead and write your Javascript . You might want to at least switch to assembly language, though. Javascript is hugely inefficient as a high level language, even more so than most compiled languages.
--- Most of the programs I've written have been in IBM 360 Assembler and its descendants and my storage estimates were based on that. I offered Javascript as easy to translate into whatever takes your fancy. I doubt if a user would actually notice any difference in performance.
If it's as easy as you think it is, why not go ahead and do it? You'd be famous overnight all around the world .
What you (and GMproposedsolution and Optimmised) are doing here is bringing a layman's understanding to a problem and then insulting the expertise of all the people that have actually taken the time and effort to work on and actually understand the problem over the past several decades.
What you offer is complete conjecture posited from an hour or two of pondering an issue from a 10,000ft perspective. It's remarkably similar to a certain someone stating that they think scientists should look into injecting bleach to combat Covid-19. As if such an obvious and simple solution were to have been completely overlooked by thousands and thousands of scientists working on the issue...
Where did I say it would be easy? Where, for that matter did I say that it was anything but complete conjecture?
I tend to agree with @pfren that the EGTB approach for the full game of chess is impracticable from a storage point of view, so if there is a solution that could be practicably stored, what I suggested is the only alternative I can see. I would guess that all the people I'm apparently insulting would agree with that.
It corresponds roughly with the approach adopted by experts prior to the invention of the computer over the past several centuries.
And, further, where did I suggest injecting bleach to combat Covid-19?
Where did I say it would be easy? Where, for that matter did I say that it was anything but complete conjecture?
I tend to agree with @pfren that the EGTB approach for the full game of chess is impracticable from a storage point of view, so if there is a solution that could be practicably stored, what I suggested is the only alternative I can see.
It corresponds roughly with the approach adopted by experts prior to the invention of the computer over the past several centuries.
And, further, where did I suggest injecting bleach to combat Covid-19?
The bleach thing is from current events and was a comparison, I did not say that you are "a certain someone"...it been all over the news, etc.
New evidence proves Black wins with best play on both sides!?
New evidence proves Black wins with best play on both sides!?
Is this a draw? What if the Black pawn on f7 is on g7 instead?
New evidence proves Black wins with best play on both sides!?
Is this a draw? What if the Black pawn on f7 is on g7 instead?
Not if you move the f7 pawn one file to the right, but then neither is it the starting position.
A symmetric position like the starting position is either a draw or a zugzwang or zwang (first player wins). This is not true of positions in general, so the asymmetric "starting" positions are not as relevant to the question.
New evidence proves Black wins with best play on both sides!?
Is this a draw? What if the Black pawn on f7 is on g7 instead?
Not if you move the f7 pawn one file to the right, but then neither is it the starting position.
A symmetric position like the starting position is either a draw or a zugzwang or zwang (first player wins). This is not true of positions in general, so the asymmetric "starting" positions are not as relevant to the question.
There are 180 pages in this thread. No one has offered the game that Ponz requested as refutation in his first several posts.
In the meantime, other questions have been posed.
Since there is currently a solution for every 6 (or maybe 7) piece game of chess do you know if every "starting" 6 piece game is a draw? For example both sides have king, pawn, and bishop in their original positions.
My post addresses these.
PATRIOT If both sides start with king, pawn, and bishop in their original position--it certainly is a draw..
And of course every starting position of 6 or 7 pieces is Not a draw. For example White could have a lone king and Black could have K and 4 queens [assuming no instant stalemate] and then Black would have an easy win.
@Ziryab
Yes, you're right, my apologies.
I think probably the symmetry of the starting position across the board is connected with the idea that the starting position is drawn, so I would count these as more relevant indicators of the situation for the full starting position
If, for example, the fraction of winning symmetric partial starting positions began to look like
1-ℯ⁻ⁿ as a function of the number n of pieces this might convert some people to the idea that the initial position is actually won for one side or other. Unfortunately we currently only have relevant EGTBs for 2, 4 and 6 men which is a bit underwhelming. A fraction of 0 for the two man position is obviously a good start.
@Ziryab
Yes, you're right, my apologies.
I think probably the symmetry of the starting position across the board is connected with the idea that the starting position is drawn, so I would count these as more relevant indicators of the situation for the full starting position
If, for example, the fraction of winning symmetric partial starting positions began to look like
1-ℯ⁻ⁿ as a function of the number n of pieces this might convert some people to the idea that the initial position is actually won for one side or other. Unfortunately we currently only have relevant EGTBs for 2, 4 and 6 men which is a bit underwhelming. A fraction of 0 for the two man position is obviously a good start.
Eight pieces. Known outcome. Gioachino Greco knew the outcome in the early seventeenth century, but there were errors in his analysis. Jozsef Szen worked out the key ideas in the 1830s. See http://chessskill.blogspot.com/2020/06/three-pawns-problem.html
i just googled 'shannon's number' and i got...360.479.5104
anyway i utubed it and got...https://www.youtube.com/watch?v=R5Wpn3dFrEs
(there's a rumor that this songs abt IM Pfrens goldfish that died last week)
@Ziryab
Yes, you're right, my apologies.
I think probably the symmetry of the starting position across the board is connected with the idea that the starting position is drawn, so I would count these as more relevant indicators of the situation for the full starting position
If, for example, the fraction of winning symmetric partial starting positions began to look like
1-ℯ⁻ⁿ as a function of the number n of pieces this might convert some people to the idea that the initial position is actually won for one side or other. Unfortunately we currently only have relevant EGTBs for 2, 4 and 6 men which is a bit underwhelming. A fraction of 0 for the two man position is obviously a good start.
Eight pieces. Known outcome. Gioachino Greco knew the outcome in the early seventeenth century, but there were errors in his analysis. Jozsef Szen worked out the key ideas in the 1830s. See http://chessskill.blogspot.com/2020/06/three-pawns-problem.html
This is more what I was wondering about. If I understand this right if one side is to move first it's a win, if the other moves first it's a draw. But if it's the Szen position then either side moving first is a win.
I realize a starting position like 2 rooks and a king in their original positions is very easy, but I was wondering if positions similar to the Szen have a first move advantage that results in a forced win. It seem like there are probably some positions where it's not so obvious and some could be forced wins and some are forced draws.
PATRIOT I already showed that Black wins from the starting position. Have found one game i n trillions of games played where Black forces a win from the starting position. [then I found more]
You were right all along--there were so many draws at the higher levels because players who had Black had believed in the Fake News that they had a disadvantage and could only play for a draw. Once I realized Black wins with perfect play I was winning game after game with experts and masters and grand masters with Black.
@Ziryab
Yes, you're right, my apologies.
I think probably the symmetry of the starting position across the board is connected with the idea that the starting position is drawn, so I would count these as more relevant indicators of the situation for the full starting position
If, for example, the fraction of winning symmetric partial starting positions began to look like
1-ℯ⁻ⁿ as a function of the number n of pieces this might convert some people to the idea that the initial position is actually won for one side or other. Unfortunately we currently only have relevant EGTBs for 2, 4 and 6 men which is a bit underwhelming. A fraction of 0 for the two man position is obviously a good start.
Eight pieces. Known outcome. Gioachino Greco knew the outcome in the early seventeenth century, but there were errors in his analysis. Jozsef Szen worked out the key ideas in the 1830s. See http://chessskill.blogspot.com/2020/06/three-pawns-problem.html
This is more what I was wondering about. If I understand this right if one side is to move first it's a win, if the other moves first it's a draw. But if it's the Szen position then either side moving first is a win.
I realize a starting position like 2 rooks and a king in their original positions is very easy, but I was wondering if positions similar to the Szen have a first move advantage that results in a forced win. It seem like there are probably some positions where it's not so obvious and some could be forced wins and some are forced draws.
Follow the link, and maybe another link from the article to see Szen's position. In the position that I posted this morning, both king's are in the starting position. In Szen's, the kings are both equidistant from their own pawns. In Szen's, the player on move wins with perfect play.
we know that the player who moves first CAN and does lose in certain conditions at its most basic level, right ? K+p vs K (opposition). this proves that at least a real sample does exist.
we know that the player who moves first CAN and does lose in certain conditions at its most basic level, right ? K+p vs K (opposition). this proves that at least a real sample does exist.
But that isn't with equal material (and obviously can work only if Black moves first). With KPKP and all pieces in their starting positions all positions are drawn.
Can you think of any symmetrical position with the pieces (what pieces there are) in their starting position that is won by the second player? (If you post the FIDE starting position you'll need to prove it.)
THIS JUST IN: PONZ PROVES CHESS IS A WIN FOR BLACK!
Forget all the propaganda and Fake News from the grand masters and strongest players!! The C D and E players were correct. After all in their games Black won often!
It is well known that the game of 1-3-5-7 Nim is a forced win for the 2nd players. In poker it is often an advantage to move 2nd or later.
White by moving first breaks the symmetrically and Black can pounce on that mistake!!
In Ponz's own correspondence games--Black almost always won. In a national correspondence championship Ponz won ALL 7 games in the Finals,
So what happened?? PATRIOT was right--the players were not trying to win and thus the very low number of wins. If only the players at top correspondence chess decided to forget the Fake News that chess was a draw--we would see more and more wins.!
So Ponz has put the theory of trying hard with Black to the test!
Ponz put out a couple of challenges on chess,com where PONZ would Only Play Black vs experts and masters and grand masters. Ponz's opponents would play White and were allowed to use chess engines and data bases. Also all opponents were allowed 3 days per move. Ponz was determined to try and win every game and he actually scored very well vs his opponents.
Here is one of the games where Ponz played Black against a grand master. Please note that in all of the trillions of games played--this is the first game played where neither side made a mistake and yet on e side won [Black] Nobody can point out a losing mistake made by the grand master![after Black made his 20th move White resigned]
Grand master vs Ponz