Chess will never be solved, here's why

Sort:
mrhjornevik
JhadTrump9202ks wrote:

StOp SPAM!!!!!!!!!!!!!

What weird definition of spam do you use ?

MARattigan
tygxc wrote:

@10703

I go with the number 10^38 positions without underpromotions to pieces not previously taken, but you can take the whole number of 10^44 legal positions if you want.

Can you at some stage decide which game you're trying to solve and what you mean by a position? You've just been talking about ICCF chess which includes the 50 move rule up to the point that a position is reached that can be adjudicated by a 7 man tablebase and also the triple repetition rule, so 10^44 is certainly not the whole number of positions in any sense in that game; 4.8x10^44 is our best estimate of the number of collections of those attributes of a position that are sufficient to determine forward play under basic rules chess.

You conceded that point already a few pages into the thread, but forgot to update your cut and paste file.

MARattigan
mrhjornevik wrote:
...

Now atleast we speak the same language.

Strongly solved means we have a mateix of moves to follow for a result.

Weakly solved means we have a strategy to follow for a result.

...

That's not quite correct. You need only a strategy in either case and following a matrix counts as a strategy. The difference between the two is that a weak solution of a position guarantees the best forcible result from that position if it's followed, but a strong solution of a position guarantees the best forcible result from any subsequent position if it's followed from the subsequent position (but not necessarily from the initial position).

Don't try to talk the same language as @tygxc on that subject too much. He quotes the definitions a lot but doesn't understand them and feels free to completely ignore them. He's been insisting since the start of the thread that the Syzygy tables are a strong solution of the positions they cover under competition rules but trying to convince him they're not is rather like trying to hammer nails into two short tungsten planks.

MARattigan
mrhjornevik wrote:
MARattigan wrote:
tygxc wrote:

@10696

...

Chess has been strongly solved for 7 men or less.

...

No it hasn't. Do you not read the posts?

It atleast have been weakly solved. Imagne a king and rook endgame. We know the strategy for how to win. I dont know if 7. But with fewer pices it become relativly easly for a computer to calculate

It's not generally been even weakly solved for positions with castling rights but for all other positions with 7 or fewer men it's been strongly solved under basic rules and under competition rules weakly solved if the ply count under the 50 move rule is less than 4.

You can look up solutions here. You need to do some arithmetic with the ply count and DTZ if the former is > 0.

You may have to wait some time for 8 men.

MARattigan
mrhjornevik wrote:
tygxc wrote:

@10707

Some trolls come up with castling rights and nearly triggered 3-fold repetition or 50-moves rules all besides the question.

By definition If ypu dont have the matrix its not weakly solved? It is ultra weakly solved?

Lets say white has the rook in a rook king endgame. I agree we dont need a matrix for every move white can do, but we need it for the shortest possible win and every move black can do to stall. Now we already have the strategy so I dont think writing the matrix is benifitial, but without it the endgame remains soft solved.

It's still strongly solved under basic rules according to the definition .

DiogenesDue
JhadTrump9202ks wrote:

StOp SPAM!!!!!!!!!!!!!

I agree...your post should be reported.

MARattigan
MARattigan wrote:
mrhjornevik wrote:
tygxc wrote:

@10707

Some trolls come up with castling rights and nearly triggered 3-fold repetition or 50-moves rules all besides the question.

By definition If ypu dont have the matrix its not weakly solved? It is ultra weakly solved?

Lets say white has the rook in a rook king endgame. I agree we dont need a matrix for every move white can do, but we need it for the shortest possible win and every move black can do to stall. Now we already have the strategy so I dont think writing the matrix is benifitial, but without it the endgame remains soft solved.

It's still strongly solved under basic rules according to the definition . But not under competition rules. For a simple example:

 
 

The above is a rather incompetent attempt to mate myself with a rook. The FEN for that final position is 1R6/8/8/8/8/1K6/8/1k6 w - - 96 49

If I ask Syzygy what to do next:

https://syzygy-tables.info/?fen=1R6/8/8/8/8/1K6/8/1k6_w_-_-_96_49

It tells me White is winning with DTZ 3. That's in ply and adding it to the 96 ply count still falls within the magic 100 ply count that represents 50 moves. Nevertheless White cannot actually win under competition rules because if he plays 49.Kc8 he draws by repetition and anything else draws under the 50 move rule. So the Syzygy tablebase can't handle that position.

mrhjornevik
MARattigan wrote:
mrhjornevik wrote:
...

Now atleast we speak the same language.

Strongly solved means we have a mateix of moves to follow for a result.

Weakly solved means we have a strategy to follow for a result.

...

That's not quite correct. You need only a strategy in either case and following a matrix counts as a strategy. The difference between the two is that a weak solution of a position guarantees the best forcible result from that position if it's followed, but a strong solution of a position guarantees the best forcible result from any subsequent position if it's followed from the subsequent position (but not necessarily from the initial position).

Can ypu explain that again?

I understand that weak is from that position.

But how is that different from a strong solve?

The best result from any subssequent position if its followed from a subsequent position?

Does that mean that if you have a weak solve for a position, after one move (the subsequent position) you now have a strong solve ?

MARattigan
mrhjornevik wrote:
MARattigan wrote:
mrhjornevik wrote:
...

Now atleast we speak the same language.

Strongly solved means we have a mateix of moves to follow for a result.

Weakly solved means we have a strategy to follow for a result.

...

That's not quite correct. You need only a strategy in either case and following a matrix counts as a strategy. The difference between the two is that a weak solution of a position guarantees the best forcible result from that position if it's followed, but a strong solution of a position guarantees the best forcible result from any subsequent position if it's followed from the subsequent position (but not necessarily from the initial position).

Can ypu explain that again?

I understand that weak is from that position.

But how is that different from a strong solve?

The best result from any subssequent position if its followed from a subsequent position?

Does that mean that if you have a weak solve for a position, after one move (the subsequent position) you now have a strong solve ?

No.

Assume we're considering basic rules chess and consider this mate in 20 position.

 
 
White to play
 

Many sources will give you a weak solution of the position. Many are reasonably accurate so no positions occurring in the analysis will have a mate depth greater than 20.

This is a position that could be reached from the above position if you don't follow the weak solution (you can make up your own moves).

 
White to play
 

It's a mate in 30 so it doesn't occur in the weak solutions you're given. However a strong solution of the initial position will not only give you a strategy for winning from the first position, but also from the second and in general getting the best result from any mess you could get into from the first, though that could be only a draw (difficult to lose in this case).

stancco

Just my two cents

MARattigan

I don't need a link. A brain works better.

stancco
MARattigan wrote:

I don't need a link. A brain works better.

I report MARattigan here for having a missing link

MARattigan
tygxc wrote:

@10707

"I dont know if 7"
++ Yes we have 7-men endgame table bases. Work on 8-men table bases is in progress.
Some trolls come up with castling rights and nearly triggered 3-fold repetition or 50-moves rules all besides the question. They cannot come up with a reasonable game that would lead to such a pathologic exception.

Here's a position that wouldn't be solved in the 32 man tablebase you refer to in the previous post if castling rights were excluded.

And here's one that won't be solved in the 8 man tablebases under construction.

But whether you regard those as reasonable positions is known only to yourself.

The point is what you think is a reasonable game has sweet FA to do with the definition of "solved", whether ultra-weakly, weakly or strongly.

Why do you keep quoting the definitions and pontificating over them when in practice you just totally ignore them?

MARattigan

Took a while to think of that, considerin'.

MARattigan
Optimissed wrote:
MARattigan wrote:
MARattigan wrote:
mrhjornevik wrote:
tygxc wrote:

@10707

Some trolls come up with castling rights and nearly triggered 3-fold repetition or 50-moves rules all besides the question.

By definition If ypu dont have the matrix its not weakly solved? It is ultra weakly solved?

Lets say white has the rook in a rook king endgame. I agree we dont need a matrix for every move white can do, but we need it for the shortest possible win and every move black can do to stall. Now we already have the strategy so I dont think writing the matrix is benifitial, but without it the endgame remains soft solved.

It's still strongly solved under basic rules according to the definition . But not under competition rules. For a simple example:

 
 

The above is a rather incompetent attempt to mate myself with a rook. The FEN for that final position is 1R6/8/8/8/8/1K6/8/1k6 w - - 96 49

If I ask Syzygy what to do next:

https://syzygy-tables.info/?fen=1R6/8/8/8/8/1K6/8/1k6_w_-_-_96_49

It tells me White is winning with DTZ 3. That's in ply and adding it to the 96 ply count still falls within the magic 100 ply count that represents 50 moves. Nevertheless White cannot actually win under competition rules because if he plays 49.Kc8 he draws by repetition and anything else draws under the 50 move rule. So the Syzygy tablebase can't handle that position.

I'm wondering why you expect anyone to understand this. It's clearly nonsense since it seems to be linked to a game where the winning side was deliberately playing for a draw under the 50 move rule or whatever.

I wasn't expecting you to understand it @Optimissed, just everybody else except @tygxc.

nov04-inactive

even if chess is solved, no one would have to worry cuz there is no way someone even Magnus could memorize that much moves

MARattigan
Optimissed wrote:
...

I didn't try hard to understand it because I assumed there was nothing to understand. ...

I think that's the main cause of your problem.

Elroch
Optimissed wrote:
mrhjornevik wrote:

The question if chess can be solved is about mathematicics, not about computers. Someone Said the amount of possible moves is infinite, its not. Infinity is actualy infinitly much bigger. Therfore, with enought time (not infinitly Long time

For practical purposes we can regard the permutations in chess as infinite. That's a very different thing from saying that the number of permutations is actually infinite. It isn't something I would expect the gang to understand.

That's because it's nonsense. Finite numbers are all TINY compared to aleph-null (the first infinite number). An infinite number is not just "very big", it has a different character to every finite number.

MARattigan

Finite numbers are all TINY compared to almost every other finite number (but not as tiny).

MEGACHE3SE
tygxc wrote:

@10678

See Figure 2 from 1 s/move to 1 min/move reduces errors from 11.8% to 2.1%.

that doesnt mean anything, thats two points on a graph, and you dont even have the right y values.