@13213
"why @tygxc bothers with the 10^17 number" ++ 10^17 is the number of positions relevant to weakly solving chess and by coincidence is also the number the 17 ICCF WC Finalists considered to arrive at the now 112 draws out of 112 games.
"The branching factor for the latest engines is 1.5 to 2."
++ Chess is easier to prune. As 10^38 = 3^80, and as perfect ICCF WC finals draws last average 39 moves, the branching factor without transposition is 3.
"chess is easier to prune" tygxc says, despite the fact that the opposite is the case.
chess is more complex at every aspect, it therefore takes more computing power to calculate each position and "pruning". important to also note that tygxc claims to be able to prune branches of the proof tree itself, which by definition is not allowed.
"As 10^38 = 3^80, and as perfect ICCF WC finals draws last average 39 moves"
several errors here. tygxc again falsely claims the "ICCF WC finals draws" as perfect games. there is nothing but circumstantial evidence for that claim, and we need mathematical proof.
the 10^38 number is total positions without promotions, as tygxc cites the paper "On the number of positions in chess without promotion" for this number. a weak solution must take promotions into account.
tygxc then tries to calculate a branching factor of positions that dont include promotions... into a statement about positions that do include promotions.
In addition, the games that tygxc talks about are not completed games. they just reached a position where a draw was agreed upon. there are still many more branches to be made.
it is important for all viewers to note that Tygxc likely knows that he is wrong in some far reaching part of his brain, but is too prideful to begin the intellectual healing process.
hello viewers, its also greatly important to note how i JUST EXPLAINED HERE:
how tygxc's description of the 10^17 number is objectively incorrect, but literally 2 posts later tygxc continues to posts the same claim:
without even bothering to acknowledge a counter argument. tygxc has been doing this for two years.
you might be wondering, surely tygxc is ignoring me because he's blocked me? that would be incorrect. you see, tygxc is the one downvoting posts that disagree with him, and that includes mine.
@13213
"why @tygxc bothers with the 10^17 number" ++ 10^17 is the number of positions relevant to weakly solving chess and by coincidence is also the number the 17 ICCF WC Finalists considered to arrive at the now 112 draws out of 112 games.
"The branching factor for the latest engines is 1.5 to 2."
++ Chess is easier to prune.
Whether it is easier to bodge a solution is not relevant.
As 10^38 = 3^80, and as perfect ICCF WC finals draws last average 39 moves, the branching factor without transposition is 3.
Firstly, that is a nonsense calculation, combining numbers in a way that makes no sense (as well as repeating multiple other blunders).
Secondly, practical chess fails to achieve rigor by a million miles. This is as true for every engine game as it is for every human game. Ignoring 34 out of 36 legal moves entirely is not solving chess, it is playing chess.
If ICCF games were rigorous, there would be no point in playing more than one of them.
The branching factor of chess remains about 36, and the definition of a weak solution requires dealing with this fully only for each opponent move. Even allowing for transpositions, the cost of 40 moves is impractical (unlike the 10^14 cost of the proof tree of checkers)
I know @tygxc has been having problems with the necessity of rigour, so perhaps a picture will work. Here is a diagram of the operation of a strategy in a game much simpler than chess, where each position in the analysis tree generates 4 new positions (plus others that happen to have been reached by other paths). Note that on the diagram only the opponent moves are shown - proponent moves are assumed to be defined by the strategy, so they have a branching factor of 1 and can be ignored in the tree.
Now assume that all but one of those branches have been rigorously analysed to a result of a draw (or better) for white, but the move labelled 16 has not been analysed any further. All we have is a zero ply evaluation of the position which is x.xx (some number of "pawns" output by a neutral network that has been trained on a large set of positions, say, like for Stockfish 16 NNUE).
Can we be certain of the result of the strategy?
The answer should be simple to everyone. But I am sure one person will have trouble with it.
@13217
"If ICCF games were rigorous, there would be no point in playing more than one of them."
++ They try to win, but fail.
"The branching factor of chess remains about 36" "the cost of 40 moves"
++ No, 36^80 = 3*10^124, much more than the number of legal positions.
So that does not take transpositions into account.
"All we have is a zero ply evaluation of the position which is x.xx"
++ No, just human logic is enough. 1 a4 cannot be better than 1 e4.
1 Nh3 cannot be better than 1 Nf3. 1 e4 e5 2 Ba6? cannot be better than 2 Nf3.
1 e4 e5 2 Nf3 Nc6 3 Ng1 cannot be better than 3 Bb5 or 3 Bc4.
3 Nd4?, 3 Nxe5?, 3 Ng5?, 3 Nh4? can be dismissed right away.
@13217
"If ICCF games were rigorous, there would be no point in playing more than one of them."
++ They try to win, but fail.
"The branching factor of chess remains about 36" "the cost of 40 moves"
++ No, 36^80 = 3*10^124, much more than the number of legal positions.
So that does not take transpositions into account.
This is trivially correct, and is one reason why solving chess would very likely have higher complexity than the square root of the number of legal positions.
"All we have is a zero ply evaluation of the position which is x.xx"
++ No, just human logic is enough. 1 a4 cannot be better than 1 e4.
That is not logic, that is someone misusing the word "logic" without any relationship to it's meaning.
If you disagree, specify where you used modus ponens, modus tollens or contraposition.
It's also a really dumb argument, because no human is looking at all of the millions of potentially crucial, unanalysed nodes from each move of an ICCF game!
the proper term would be "human assumptions". because logic would be derived directly from axioms.
isnt it funny how tygxc completely ignores the plaintext demonstration of his folly? he cant answer that simple question, can you be certain of the result of 16?
@13087
"ICCF is a joke. I wouldn't lose a single game either."
++ Prove yourself wrong and sign up for an ICCF WC preliminary then. Note there are decisive games in preliminaries, in semifinals, in candidates, and until this year even in finals.
If you think you are better than all these ICCF players, then prove it.
I don't believe ICCF games are a joke - they're obviously top-level games, as they are played with the use of extensive, multi-engine analyses.
Though I don't really find the games to be too impressive, in terms of skill level. I'd be far more impressed if ICCF players used their own brains to battle each other, and didn't use any external analysis tools at all ... the way correspondence chess used to be, in the pre-engine days.
Just sit there with a physical board and ponder the moves yourself, trying different variations and squinting at the board, until you find the next move that you want to send to your opponent ... That's "pure" correspondence chess, IMO. And that would test the actual playing level and chess understanding of all those involved.
But of course, with all the analysis tools available these days, this kind of expectation isn't reasonable anymore.
But ... anyone can play chess at a superhuman level, if allowed to use the assistance of superhuman engines - and I believe that's the point that llama was making ...
I assume it drew itself then and future ones will draw themselves any engine has drawn itself in opening position since forever and once you do reach the table base its got lots of wiggle room with equal material for draws and even more so with the 50 move rule. So its not brute force solved without a full table base but it seems obvious. There is not a single line that doesn't currently go to 0.00 on modern engines and the stronger they get the faster they seem to get to 0.00. So I think its a draw and we probs wont get hit by comet and by 30XX if we still exist our tech will have a way to solve this in a more efficient way than any of us can imagine today. I cannot imagine a forced win line and the ideas would need to be pretty damn deep because tons of people with engines tablebases databases haven't found a forced winning idea it always goes to 0.00. So not solved yet but would be shocked if there was some forced mate in 500 from the opening esp given the 50 move rule lol.
Evaluations by current engines are not reliable, much like evaluations from previous engines...that's the point. The 0.00 is not conclusive unless it is using the tablebase.
Yes thus the it’s not solved until we have a full table base but I don’t see how that’s relevant at all. The table base plays better than an engine and one using one wins where is the one where they both see the table based positions with szygy probe tho right. No matter what elo what strength they all draw themselves from opening positon under same conditions since forever. However if we had a full table base right now I would expect it to say draw TBD on move one and I don’t think that eval would change wether white plays e4 d4 c4 e3 whatever. It’s not proven till it’s proven but intuitively I’m sure it’s draw. But hey I agree with you doesn’t count until we have one even if of sure of the outcome already.
Have humans walked on Mars? NoCan humans walk on Mars? YesWill humans walk on Mars? Maybe
lol...yee u !
chess will never be solved beacuse for example, we strong players think its a blunder but what about in a million moves?
sooo true ks92. the 50-move rule is to concede to a prac tickle tie. should solving chess be subject to ?...NO !
mathematical objectivity - he appears to have contempt for it too.
i know one thing. if abuncha egghead jurists ganged up on me (like they have on Ty) id grow lotsa contempt too. not for the math but for the court ! [tuesday growl]
now get notified every time a post is added
really ??...plz...dont do a 'X' off...just uncheck 'Follow' (lower RH corner...look dn)
where they have free will and choice
u gave those two up. {burp...serve him another ?}
@13217
"If ICCF games were rigorous, there would be no point in playing more than one of them."
++ They try to win, but fail.
...
Hey - I'm as good as an ICCF grandmaster!
Have humans walked on Mars? NoCan humans walk on Mars? YesWill humans walk on Mars? Maybe
lol...yee u !
Lola has been on Mars for years, so...
I don't know why @tygxc bothers with the (inadequate) 10^17 number anymore when the c. 5000 positions in a selection of unco-ordinated games amounts to a proof to him.
Interesting fact. The branching factor for the latest engines is 1.5 to 2. This is how they achieve such impressive depth (if you think about it, it's the only way without massive increases in nodes per second). This is rather effective for playing chess as it achieves the best balance between thoroughness and practical results (I.e. a 100% weighting on the latter). Anyone who thinks you can solve a game while ignoring 95% of the legal moves for the opponent is more deluded than one who thinks that you can get by with ignoring 90% of them.
Another interesting fact. A 300,000,000 nodes per second engine with a given branching factor gets only 58% deeper at the same branching factor in 4 days as it does in 4 seconds! It's because it's the log of the number of nodes that determines search depth.