Solved?

Forum rules
Please DO NOT POST BUGS on this forum. Please report (and vote) bugs on : https://boardgamearena.com/#!bugs
User avatar
Jellby
Posts: 452
Joined: 31 December 2013, 12:22

Solved?

Post by Jellby »

Apparently Quarto is a solved game, and one can find computer programs that give you the best move. I don't know if people use that, but shouldn't the mere availability be enough to disable Arena and rankings?
User avatar
aesche
Posts: 381
Joined: 06 April 2020, 02:31

Re: Solved?

Post by aesche »

You‘ll get open ears on that from someone responsible, but I‘m not a huge fan of that direction…
User avatar
tchobello
Posts: 405
Joined: 18 March 2012, 13:19

Re: Solved?

Post by tchobello »

Jellby wrote: 20 February 2022, 19:09 Apparently Quarto is a solved game, and one can find computer programs that give you the best move. I don't know if people use that, but shouldn't the mere availability be enough to disable Arena and rankings?
Are you talking about basic or advanced rules of Quarto ?
User avatar
Jellby
Posts: 452
Joined: 31 December 2013, 12:22

Re: Solved?

Post by Jellby »

That would be basic, I guess. I haven't checked, but I'd imagine that if solvers are not available for the advanced variety, they'll probably be there soon.
User avatar
RicardoRix
Posts: 1756
Joined: 29 April 2012, 23:43

Re: Solved?

Post by RicardoRix »

kalah and connect4 are solved games too.
Doesn't necessarily mean that people can't still play the game.

It's perhaps not that surprising since there are only 16 pieces and 16 locations to start with and it narrows down very quickly after each move. Number crunch the whole thing then it's not really very dependent upon standard/advanced game, if anything there might be less variations with advanced because the game ends quicker.

There is a far wider abundance of engines for go and chess, which all essentially mean the same thing: if you use a computer engine - you win.

I think it's very much a take it or leave it situation. BGA can't realistically get involved with this kind of thing, it might be close to impossible to tell the difference between an engine and a highly skilled player, let alone have the time and resources to apply the checks.
User avatar
Jellby
Posts: 452
Joined: 31 December 2013, 12:22

Re: Solved?

Post by Jellby »

I'm not saying the game should be removed, just the "competitive" aspects, as it has already been done for chess and other games (https://boardgamearena.com/faq?anchor=f ... edrankings)
User avatar
megamau
Posts: 62
Joined: 26 March 2017, 20:52

Re: Solved?

Post by megamau »

There will be "solvers" or very strong engines for every game which is worth playing competitively.
I don't think we should be removing the competition, as it will eventually engulf all games.
Parcly Taxel
Posts: 5
Joined: 16 March 2022, 10:58

Strongly solved!

Post by Parcly Taxel »

Jellby wrote: 20 February 2022, 20:29 That would be basic, I guess. I haven't checked, but I'd imagine that if solvers are not available for the advanced variety, they'll probably be there soon.
There is a perfect player for both Quarto variants, from any position, that has been around since 2018. It calculates optimal moves in less than a second, and apparently without any fancy neural networks (i.e. pure alpha-beta pruning).

So the game is not only solved, it's strongly solved. I strongly recommend that competitive modes be disabled for Quarto – and further down the line Quantik, Soluna and Squadro too; these games (1) are perfect information and (2) have low enough game tree complexities that a perfect engine running on an ordinary computer is very much possible.
Last edited by Parcly Taxel on 18 March 2022, 07:53, edited 2 times in total.
User avatar
RicardoRix
Posts: 1756
Joined: 29 April 2012, 23:43

Re: Strongly solved!

Post by RicardoRix »

Parcly Taxel wrote: 16 March 2022, 11:14
Jellby wrote: 20 February 2022, 20:29 So the game is not only solved, it's strongly solved. I strongly recommend that competitive modes be disabled for Quarto – and further down the line Quantik, Soluna and Squadro too; these games (1) are perfect information and (2) have low enough game tree complexities that a perfect engine running on an ordinary computer is very much possible.
How does publishing the source of the solver help?
Solvable, strongly solved (which is no different to solved), non-solvable. All these are mildy irrelevant, this next statement says it all:
You need to move the argument forward from here:
megamau wrote: 22 February 2022, 05:55 There will be "solvers" or very strong engines for every game which is worth playing competitively.
I don't think we should be removing the competition, as it will eventually engulf all games.
Parcly Taxel
Posts: 5
Joined: 16 March 2022, 10:58

Re: Strongly solved!

Post by Parcly Taxel »

RicardoRix wrote: 16 March 2022, 11:57 You need to move the argument forward from here:
megamau wrote: 22 February 2022, 05:55 There will be "solvers" or very strong engines for every game which is worth playing competitively.
I don't think we should be removing the competition, as it will eventually engulf all games.
I do not think that the competition should be removed from many games like Carcassonne, Race for the Galaxy, Terra Mystica, Coup and so on – all those games do have chance or hidden information, to the point where even the best computer can still lose. I have a Tantrix set, and the instruction booklet for the game notes that while with chess "the winner is not in doubt", with that game "the best player wins, but not always".

I consider that last phrase a succinct summary of the spirit of boardgaming: with enough experience anyone should have a realistic chance of winning over anyone else, yet they should always have a realistic chance of losing too. For the specific games I have picked out, a human's probability of winning over an engine or someone using it would be zero or negligible, and furthermore by the Church–Turing thesis it is impossible to perfectly tell a human and computer apart when they play (Lichess and some online backgammon servers use computers themselves to catch obvious cheaters, but the theoretical result still stands).

In short, engines will not "engulf" competitions for the majority of games here, but a majority is not all.
Post Reply

Return to “Quarto”