I would agree with that statement of btickler's... I'm just leaving room to be surprised. A new method of processing could be invented any time, but it's going to take new technology to crack chess.
I also hope it doesn't happen any time soon.
Will computers ever solve chess?
Sort:
"there is always just one option for the moves for one side, the entire range for the other"
This statement is clearly incorrect for positions where there is a forced move (that is to say 'only move' positions) and many other forcing combinations where making a second choice move loses outright to an easily seeable continuation. So, saying that in every chess position there are (insert number over 25) possible moves and then extrapolating from that is nonsense and facile as I said many posts ago.
Well btickler's statement is an opinion and not supported by any math. It appears he (and others) have confused the complexity of the game (often expressed with the Shannon number) with the amount of work required to solve the game. There is no known connection between these numbers for solving chess (other than to say brute force, which is a naïve analytical method, can't be used).
An example is tic-tac-toe. It has a large amount of games that can be played and almost nobody ever writes all down. But young children learn to play it perfectly without memorizing or writing down all the possible games. They use deductive logic.
Tic tac roe or noughts and crosses as it is commonly called in England falls rather neatly to a magic square formula which whilst it has nothing to do with the game itself can be used to construct a program to play the game. Every possible combination of tries in tic-tac-toe does not need to be put into the computer program, In fact, the program using this method bears little resemblance to the game were it not necessary to have an output routine and user interface.
There may well in the future be more complex but similar, apparently unrelated ideas which can be applied to chess.
Feb 3, 2018
0
#2826
chessspy1 wrote:
"there is always just one option for the moves for one side, the entire range for the other"
This statement is clearly incorrect for positions where there is a forced move (that is to say 'only move' positions) and many other forcing combinations where making a second choice move loses outright to an easily seeable continuation. So, saying that in every chess position there are (insert number over 25) possible moves and then extrapolating from that is nonsense and facile as I said many posts ago.
No, it is not. Firstly because when there is one legal move that is the entire range of legal moves for that position. This is a rare extreme circumstance in (critical) balanced positions. What matters to the statistics is that it is more typical to have many legal moves and (very likely) typical to have several moves that achieve the best result. Secondly, and rather importantly, it is essential to analyse all responses, even the ones that appear bad, to solve chess properly.
[Note that the claim about there being usually several moves to achieve the best result is based on tablebase statistics. Experiment at http://www.k4it.de/index.php?lang=en&topic=egtb to get a feel for this. It is the norm regardless of the result to have several moves that achieve the optimal result. (Of course when a side is losing, all moves achieve the best result! This is not as crucial as what happens when the theoretical result is a draw)].
Regarding the impact on the statistics, the crucial thing is how much a very carefully chosen optimal strategy can reduce the total number of distinct positions that can be reached by an opponent. Even ignoring those where the second player has made an error (i.e. is losing, if the true result is a draw), it only takes the second player to typically have 2 non-losing moves to mean that the number of positions that has to be dealt with is impractically large. (This relies on perfect games where neither player gives up and the second player tries not to accept a quick draw being rather long and an empirical reduction for transpositions.
Note that the number of positions being the crucial thing to complexity, the possible games provide an vastly larger number of ways to reach a fraction of those positions. This compensates a lot for the fact that games are shorter if bad moves are played by one side.
Hi Scott,
I agree, for example, the Scandinavian defence for black may well fall to computer analysis. Maybe even the Kings Gambit.
The question is, should we then give up playing them?
vickalan wrote:
Well btickler's statement is an opinion and not supported by any math. It appears he (and others) have confused the complexity of the game (often expressed with the Shannon number) with the amount of work required to solve the game. There is no known connection between these numbers for solving chess (other than to say brute force, which is a naïve analytical method, can't be used).
An example is tic-tac-toe. It has a large amount of games that can be played and almost nobody ever writes all down. But young children learn to play it perfectly without memorizing or writing down all the possible games. They use deductive logic.
Did a parrot just squawk "18 years!"...? All your regurgitated arguments have been addressed before numerous times by numerous parties.
chessspy1 wrote:
Hi Scott,
I agree, for example, the Scandinavian defence for black may well fall to computer analysis. Maybe even the Kings Gambit.
The question is, should we then give up playing them?
The Scandinavian Defense for Black may well fall to computer analysis. [i wrote a book on the Scandinavia/Center Counter and won from a whole lot of masters with that opening but am now saying that opening is dubious]
Regarding the King's Gambit--this line:
btickler wrote:
...All your regurgitated arguments have been addressed before numerous times by numerous parties.
But never refuted.
OTOH, your claim "Chess cannot be solved with current technology, nor with any reasonably foreseeable technology" was found to be completely baseless.
chessspy1 wrote:
...There may well in the future be more complex but similar, apparently unrelated ideas which can be applied to chess.
Well said. It reminds me of the history of flight - a lot of weird things were tried and people who lacked imagination and creativity said "it'll never be done". Some even said humans never flying was itself proof that it could never be done. Now humans fly.
I agree with Vic,
If we were meant to fly God would have strapped a Boeing 747 to our asses
Feb 3, 2018
0
#2834
vickalan wrote:
btickler wrote:
...All your regurgitated arguments have been addressed before numerous times by numerous parties.
But never refuted.
OTOH, your claim "Chess cannot be solved with current technology, nor with any reasonably foreseeable technology" was found to be completely baseless.
Why would you say that? There has been no even half convincing case to the contrary proposed (I think the nearest is the idea that quantum computing might some day (a) have a way to be used and (b) be developed to a high enough degree to actually do it. [As it happens, I believe both of these are likely, but a belief is not sufficient!]
When I described the most efficient possible approach to solving chess, I thought it went without saying after previous discussion that the computational demands were utterly out of reach for any feasible conventional computing.
Note that the application of "brute force" as efficiently as possible has been how all comparable game solutions have been achieved (eg connect4 and checkers).
Elroch wrote:
...When I described the most efficient possible approach to solving chess, I thought it went without saying after previous discussion that the computational demands were utterly out of reach for any feasible conventional computing..
Where is this shown, and who did the calculation?
Thanks for the info, but I haven't seen 1% of total positions related to the number of mathematical operations that are required to solve chess. Is that shown anywhere?
Yes, White has 20 possible first moves, and then Black follows with 20 possible moves. That's 400 possibilities so far.
That's the start of a brute for analysis. Claude Shannon in 1949 determined that brute force won't work to solve chess because there are too many possible games. After 1949 most mathematicians have moved on to other analytical methods. But it's still interesting. It shows that chess has lots of games with very poor play - almost like moving pieces randomly.
vickalan wrote:
btickler wrote:
...All your regurgitated arguments have been addressed before numerous times by numerous parties.
But never refuted.
OTOH, your claim "Chess cannot be solved with current technology, nor with any reasonably foreseeable technology" was found to be completely baseless.
Keep living in your dream world 
...
Let me know if you are aware of an academic or scholarly paper that has concluded that chess cannot be solved. Or if you did the calculations I'd be interested in seeing them.
Let's be clear:
My position: Chess cannot be solved with current technology, nor with any reasonably foreseeable technology.
There is no mathematician or chess-theorist who has determined the number of mathematical operations required to solve chess, so therefore the amount of time required to solve chess is unknown. It could be within a decade, or it may never happen.
Your statement above is not supported by any academic material or mathematical analysis. If you hold that view then you might want to show how you arrived at it. Of course it has no basis so I'm sure it will be quickly refuted.