Isn't this the best solution to counter tournament level cheating whether it be OTB or online?
Plus it would remove the chess by numbers approach of the best player being the one who remembers which line is best to take, going to the path of whatever opening.
Anyone know Hans' 960 rating?!
No. 960 only solves the non-existing problem of opening theory. It doesn't prevent or solve the possibilities for cheating.
@MidiChlorianCount said in #1:
> Isn't this the best solution to counter tournament level cheating whether it be OTB or online?
>
> Plus it would remove the chess by numbers approach of the best player being the one who remembers which line is best to take, going to the path of whatever opening.
>
> Anyone know Hans' 960 rating?!
Yes! Yes it absolutely should!
Capablanca wanted randomization-- Fischer wanted it as well (Fischerrandom is the old name for Chess 960 as I'm sure about 80-90% reading this know.)
So on and so forth. This is the best creative step and very nice catch, dear friend.
@Molurus said in #2:
> No. 960 only solves the non-existing problem of opening theory. It doesn't prevent or solve the possibilities for cheating.
I wouldn't say opening theory is a "non-existing problem", but overall you're right. Someone could still plug a 960 position into an engine, of course.
@tryan82 said in #4:
> I wouldn't say opening theory is a "non-existing problem", but overall you're right. Someone could still plug a 960 position into an engine, of course.
Stockfish not only plays 960, it even plays all available chess variants on Lichess. Including crazyhouse and stuff like that. :)
The question if opening theory is a problem is an entirely different problem. For now let's just say it has nothing to do with (preventing) cheating.
@MidiChlorianCount said in #1:
> Isn't this the best solution to counter tournament level cheating whether it be OTB or online?
What makes you think that? Are you assuming there's no computer program which can play chess 960 on a high level? In that case, you are mistaken.
"My, oh what a good day for a walk outside"
There are a lot of reasons Chess 960, a.k.a., Fischer Random is better than chess, or at least a good progression from chess; and there's quite a lot of reasons that Capablanca and Fischer both wanted and enjoyed the idea of a randomized starting position. Let's say for one: you have a computer in your pocket, and you have to input the starting position from your fingers, to get the beginning scenario. For one you will have to place each piece on its square, to have the accurate starting position, or for two, you will have to know the exact number of the 960 possible positions, and then input that. Now that is a big deterrent in itself, as it will slow someone down immensely.
Yes, the universe is an open universe; and we can go - yes but, yes but, yes but -- ah naturally. But there is no denying that is easier to walk on soft grass or compacted earth rather than through quicksand or deep mud or slush or deep snow -- And this, this is the equivalent of deep snow, slush or deep mud for someone walking. Absolutely you can push through the mud or the slush or the snow; but the point here is the immensity of difficulty involved as a measure of hindrance.
Sorry that came so late, Abigail. ~ 6 minutes.
@Abigail-III said in #6:
> What makes you think that? Are you assuming there's no computer program which can play chess 960 on a high level? In that case, you are mistaken.
Really? I'd be interested in finding how they performed in relation to regular chess. I'm assuming that they use machine learning to "teach" optimal moves from each opening. However, given the magnitude of potential positions / opening paths from each of the 959 other variants, the training set surely cannot be as good?
Although maybe it is reasonable - although the bulk of data would come from lower level players, so maybe not optimal.
Stockfish can absolutely win any game against any human, every time. In Chess960. Absolutely the initial positions are not mapped out; and there a lot of factors, and Stockfish does play with a "Book" or can play with a book. But Stockfish without an opening book, is still magnificently strong. Let us say that Stockfish without an opening book is about 3700 -- someone correct me if I am wrong in that. But then against it should be taken into account that Stockfish's opening book will be perfect -- in any case, Stockfish without an opening book or Stockfish on the initial position of randomized Chess 960 should be ~ 3500-3700 in my view.
I mean but that's not all that important I suppose. The primary thing is that it will, well, as said.
@MidiChlorianCount said in #8:
> Really? I'd be interested in finding how they performed in relation to regular chess. I'm assuming that they use machine learning to "teach" optimal moves from each opening. However, given the magnitude of potential positions / opening paths from each of the 959 other variants, the training set surely cannot be as good?
Computers can play millions of games against each other and against themselves. They don't need to train on a given training set. Remember AlphaZero wiping Stockfish's ass +28=78-0 (and later +155=839-6)? That was only after self-training for no more than 9 hours.
With no opening book for either humans and computers, humans are even more at a disadvantage.
