Will computers ever solve chess?

Nothing has really changed for the outcome of this thread.  Computers are still not going to solve chess in our lifetimes, and 10^46.7 is still a much larger number than people can fathom ...

Solve means create an algorithm that generates moves in a finite and reasonable time that can be proven to always be optimal. The age of the universe is finite but not a reasonable time for a chess game move.

I know more about how they work then you do ...that's why I have said the outlook for solving chess has not changed yet.

P.S. for the other poster...qubits are not the issue.  Instruction sets, decoherence, and the inability to store intermediate analysis data are the main issues.  Quantum computers have yet to even make a serious dent in the problems they are best suited to solve (cryptography for example), so it's wildly premature to say anything about solving chess.

There's already a thread on this topic, and the quantum computer side was roundly trounced.

#6864

The 10^45.888 is an upper bound for the number of chess positions, not an estimate. estimate < upper bound. This includes many illegal or irrelevant positions. A more reasonable estimate is 10^20. To solve chess not all possible positions need to be visited, only a small fraction of it needs evaluation to calculate from the initial position towards a table base. Calculation itself needs only position generation, no evaluation. The evaluation is to calculate until the reduced table base of legal, relevant, drawn positions is reached and then to look up if the position is in it or not. Only a quick evaluation is needed to prune candidate moves.

I already pointed out to you where the author of the 10^45.888 number made a disclaimer that his own number cannot be verified.  The 10^46.7 number is more widely accepted.

You are also making poor assumptions.  The numbers in the 10^40s ranges eliminate illegal positions for the most part.  Your notion that 25 orders of magnitude can be summarily tossed out is ridiculous, and you have zero basis for your 10^20 number, as shown by your lack of support every time you have been asked about it...

P.S. Tablebases work backwards from mate.  A moment's reflection will reveal to you that tablebases therefore don't evaluate nor store drawn positions.  Try again . 

#6872
Look up in a dictionary the meaning of 'upper bound'.

The vast majority of possible positions are either illegal or irrelevant.
The Syzygy table base e.g. contains illegal positions like this:

The Syzygy table base contains irrelevant positions like this:

Table bases are generated backwards: from 3 men to 4 men, from 4 men to 5 men, from 5 men to 6 men, from 6 men to 7 men and now in progress from 7 men to 8 men. They can be pruned to eliminate all but drawn, legal and relevant positions for the purpose of solving chess: calculating from the initial position towards a table base. This would drastically reduce the storage capacity needed to store the table base as well as the time to access it. An 8 men table base with only drawn, legal and relevant positions can be generated from the existing 7 men table base by pruning all won, lost, illegal, and irrelevant positions.

Solving chess then is proving that for all reasonable white moves there exists at least one black move that leads to a table base draw position.

In the future, all chess matches will begin with both players sitting down, standing up again and shake hands on the draw. It will be crowdy on the top podium step.

I know the definition of an upper bound.  But it's not an actual upper bound worth anything if the creator of the upper bound disclaims his own work, now is it?

I eagerly await your groundbreaking work in pruning a 7 man tablebase using your brilliant plan and thereby creating a smaller 8-man tablebase that is down to only drawn games (!?) plus legal and relevant positions.  I'm sure your code for determining "relevant" positions will be ruthlessly efficient and concise.

What's your plan?  Python?  Javascript?

#6876
There is no need for your condescending tone.
There are opinions that the number of sensible chess games is only 10^40 and not Shannon's 10^120. If that is true, then the number of sensible positions must be << 10^40.
https://www.youtube.com/watch?v=Km024eldY1A 

Shannon's number was never supposed to be anything like "sensible chess games".  It's an estimate of all possible chess games (using an estimated average game length of 40 moves), which logically must include all the not-so-sensible games as well.  So your premise fails right out of the gate.

I don't need to watch a YouTube video (and a particularly insipid one at that, I could have done a better one), I have actually read Shannon's paper...have you?

Also, you've done it again...your own video, after the 10:00 mark says there are 10^40 sensible moves, directly contradicting your own statements of less than an hour ago.  Now, given that you can't seem to make arguments *with* your own chosen supporting evidence without making a mistake (you've failed this test twice now), how would anyone begin to believe your 10^20 claim when you *dont* support it with evidence or even a simple back-of-the-napkin estimate?

#6878
It is clear that number of possible chess games >> number of possible chess positions
10^120 >> 10^46
However
upper bound > estimate > lower bound
number of sensible games << number of possible games
number of sensible positions << number of possible positions
10^40 sensible games >> 10^20 sensible positions
You are most wellcome to present your own better video or paper of how many sensible positions there are in your estimated opinion.

"10^40 sensible games >> 10^20 sensible positions"

10^40, 10^45.88, and 10^46.7 all refer to positions, not games.  So, *again*, this statement is meaningless.  Just admit 10^20 is a number you pulled out of nowhere.  

10^40 viable/evaluation-worthy positions is a very conservative estimate.  It's not a hard estimate to figure out.  There's 10^46.7 unique positions possible, and if only 1 out of every million positions is viable, that's still 10^40 right there.

For your 10^20 claim to stand up, only 1 in every 100,000,000,000,000,000,000,000,000 (100 septillion) chess positions would be considered "sensible".  I think we can all have a good laugh pondering that notion, and then move on to some more reasonable discussion.

#6880
It matters. Many players quit checkers after it was solved to be a draw.

It's positions that count for a precise strategy. It doesn't generally matter how you got to a position if it is your list of positions that mate in 75 (or whatever). Likewise in the openings, the route to a position is of no interest once you get there.

Thanks, I pulled an interesting paper from that StackExchange thread:

https://www.cs.virginia.edu/~robins/The_Limits_of_Quantum_Computers.pdf

It's an older article but describes the limits of quantum computing better than most articles I have seen.