Go is better than chess because

Not quite sure Go is more complicated than chess. Ok, sure on a purely mathematical standpoint yea but realistically no one play by mathematical standpoint, no one here is a computer (well…I hope).

Realistically, both are played by following a bunch of concept and rules of thumb… And both have quite a few layers of complexity. One could say than spotting the good move in Go by having good fundamental is basically the same as spotting good move in chess by having good eyes for tactics.

And realistically, it can take years to develop a similar level of skill in both game. Both do require quite heavy capacity to focus as well. It’s kinda like comparing banana to orange…both are fruit and both are liked by lot of people, both get good side and cons but it’s hard to say one fruit is better than the other from any objective standpoint :stuck_out_tongue:

Now, I could say than I love about Go than there is no draw…and than I like about chess than game are generally faster (especially fun when you get opponent dragging lost game forever). But it’s more byproduct of the game design than anything’s else and hard to say it’s really an objective advantage or making game better or worst.

2 Likes

This is really the only downside go has over chess to me. :wink: NZ rules + 7.0 komi forever <3

5 Likes

It’s not only about trees and branches. Humans are more disturbed with movement and different pieces. This is difficult to quantify.

1 Like

In my opinion good way to judge the complexity of a game in a human relevant way is the spread of the Elo ratings. If I’ve just learned the rules of go I can find a player that can beat me 9 out of 10 times and they can find a player that beats them 9 out of 10 times etc. For a more complex game there should be more of those steps until we reach top human play (or alternatively the standard deviation of the Elo distribution to avoid dependence on the number of players).
This of course also depends on the player base and intensity of professional study, and draws might distort this metric at the top levels of chess (a sufficiently good player can essentially force a draw even against a stronger player), but I think it’s still a good baseline for complexity according to which Go > Shogi > Chess > Gomuko > Checkers > … > Tic Tac Toe (with very quick and careless research into the Elo distributions of those games).

3 Likes

Such a metric also would rate “best-out-of-5 Chess” as being significantly more “humanly-complex” than Chess, as it would also have many more steps of skill as measured by Elo or similar ratings spread. (“best-out-of-5 Chess” is the game where you play 5 subgames of regular Chess in a row alternating colors, the winner is the one who wins more subgames).

Generally, this kind of metric is fairly correlated to just the average length of the portion of a game that allows for meaningful decisions. One might be happy with this, or not, depending on the intended purpose of such a metric.

4 Likes

Go on 8x8 board not necessarily better than chess on 8x8 board.
Go on 19x19 board surely better than chess on 8x8 board.

3 Likes

The question of complexity from a human point of view seems to come down to the question how good a game is at measuring the difference in people’s ability to perform complex reasoning.

After all, if you ask how complex it is for humans to play, people playing any game to the highest humanly possible level will presumably be developing mental structures of roughly equal complexity, if their brains and training are comparable, and if you insist on measuring complexity by how hard it is for us to achieve. On this basis, play in all games too hard for perfect play by humans would rate as equally complex. This is why I feel it is more interesting to look at how well the game results measure the complexity of thought. That seems to come back to looking at how many levels of skill the ranking can distinguish; which seems to favour Go. Taking @hexahedron’s point about a 5-game series, one could all how many games of chess are equivalent to one of Go.

I suspect I am making a few assumptions which would be interesting to examine.

3 Likes

Chess is better than Go, because…

At the end of the game, there’s an algorithm to determine who won :wink:

1 Like

Didn’t read the whole thread. Did someone already reply with …

… apples are better than oranges?

(BTW in German it is apples vs. pears.)


The probably most true (for me) and probably most incendiary could be …

Go is better than Chess like macOS is better than Windows.

<g,d&r>

NOTE: Although, as said, this bears truth FOR ME, it was meant as a joke (meaning that I ROTFL’d), NOT to start a platform war I don’t need that: I run Windows and Linux in virtual machines, where they belong :sunglasses:

<edit>

Just to make it clear: Whatever rocks your boat!

</edit>

2 Likes

…on the other hand, knowing when the game is finished under FIDE rules requires figuring out whether any sequence of legal moves leads to checkmate, which is quite a challenge to do algorithmically :upside_down_face:

image

5 Likes

It required a significant amount of compute, but in 2016 the number of legal go positions (reachable by only legal moves from the starting position) on 19x19 was determined to be exactly 208168199381979984699478633344862770286522453884530548425639456820927419612738015378525648451698519643907259916015628128546089888314427129715319317557736620397247064840935 (~2.1E170). (Counting Legal Positions in Go)

With chess, this number has not been determined yet. The problem with chess is that it’s far from trivial to determine if a given chess position can be reached by a sequence of only legal moves from the starting position (this is trivial with go, because any position where all groups have at least 1 liberty can be reached by only legal moves from the starting position).
As of yet, only an upper bound of 8726713169886222032347729969256422370854716254 (~8.7E45) has been determined by John Tromp, while his estimate for the actual number is ~7E44. (On the number of chess positions - Chess Forums - Chess.com)

5 Likes

Interesting. Actually it’s easy to define that algorithm, right? Just examine all the legal sequences of moves for checkmate. Whether that’s practical or not is another question.

Indeed, it is theoretically well-defined, just impractical. But as the paper linked above shows, the problem is surprisingly tractable for “most” positions, see the subheading “Experimental results” here for their results on the whole Lichess database:

2 Likes

Go is more complicated.
chess is not complicated enough even compared to Shogi.
before you compare two board games, you should try all other board games. you have to be open minded, not fanatic or blind faith.
face the truth.