GO

I love the stategy involved in playing chess and the analogous similarities of chess to real life.  However, for several years, I've been wondering if the game of Go was actually superior in these two and other areas.  Knowledgeable input requested.

In terms of mathematical complexity, of course go has more positions, more complexity.  But as limited as humans are we'll never master either in terms of perfect play, so in practical terms both are "infinitely" complex.

In terms of ratings some try to show that Go ranking (say 10kyu through 9-10 dan) number about 20 while chess classes (200 point jumps) go only from 1000-2800 which give only 10 levels.  Therefore go is twice as complex in practical terms... but this is a pretty flawed comparison.  For one thing in go there are no draws, I believe white starts with 8.5 moku (correct me if I'm wrong) for being second.  Because the half a moku there are no ties.

Chess has no such corresponding feature.  In fact it's draw margin is quite notable.  This means to go up a class in chess (win 75% of games vs 1 class lower) represents a large increase in skill, whereas in go a rise from 6 kyu to 5 kyu does not represent such a large leap.

I find go fascinating, but had to make a choice between the two.  I chose chess because of where I live I'm able to easily go to chess clubs and tournaments whereas with go I would be learning largely on my own with little guidance and little OTB competition.  I've played maybe 50 games (7-8 years ago) and know only the rules and very most basic ideas. 

My brother tried getting a bit better at it, but only kept it up for about 6 months.  He was able to find some great games of go played in the past by the equivalent of a super GM (9- dans) and even the basic explanations (of course that's all I could understand anyway) made it obvious the strategy goes very very deep.  A great game for sure.

I think it's 4.5 moku.

I'm have no doubt that Go is very deep, a game to be respected. I've never heard anything different.

After perusing some English language Go forums I saw a few posters dismissing chess, calling it a "cookbook game." I can guess what that means but I'm pretty sure it's a put down of chess as shallow--to which I can only say the best geniuses are still finding it a nice deep pool to swim in after a few centuries.

Chess is complex (and unmasterable) enough for me, which is the same reason I don't go into chess variants too much, that sometimes try to make the game more complex.

Hehe, a cookbook game.  I'm sure elitist attitudes like that would hate to know what eastern go players think of their quaint little get togethers.

Chess is more similar to human life because there are knights and kings and pawns.

Go is like two bacterial colonies fighting each other.

 Ah, but the OP asked about relevance to "real life." Therefore Go wins out because bacteria are more prevalent than humans. ; )

I played a fair bit of Go in college, but fell out of practice for lack of opponents. It was big in the computer science department because of the problems it presents for artificial intelligence. Which is an advantage in the digital age. There's not much point in cheating with a Go program, as a decent amateur can beat one.

I played a great deal of Go beginning in High School and I still play occassionally.  I wonder is it still true that Go programs are weak. It was true once, but does it remain true?  To me, there seems nothing intrinsic to go that would make it so un-programmable. But I don't understand the challenges. High level broken-pattern pattern recognition is essential to high level go. Those anti-spam images where you identify a few distorted letters superimposed over a grid... that kind of pattern deciphering is involved in high level Go. I don't understand why computers suck so grievously at it. Much like Nalimov tablebases... Go on smaller grids can be solved _completely_. But the 19x19 grid is well beyond crunchability as yet. Still, I'd be surprised if someone hasn't found some way to combine number-crunching small sections of the board with pattern recognition to crack Go.

 According to wikipedia things have certainly progressed since I was in college. However, the best programs can still only beat low level expert players on a full board. There are other problems besides search tree complexity in Go programming. For example, it's much harder to make a position evaluation function in Go. Efficient tree searching depends on knowing which positions are good, so the search tree is not only bigger but it's harder to search.

Uh oh, this thread is dangerously close to a "which game is better", which of course, can only be answered on an individual-by-individual basis.

Why is go hard to program? the number of possible positions is the obvious answer. That doesn't necessarily make a more interesting game. (AKA, size isn't everything). What's interesting is the difference between how a human plays versus how a computer plays. As it stands, the game is exceedingly difficult to brute force (though, this won't last forever). However, top human players are said to be able to read out to around 65 moves ahead.

What does this mean? well, it's not like reading 65 moves ahead in Chess. Playing go is a bit more like playing a simul, (with the slight caveat that the different boards can interact.), and so reading 65 moves ahead is closer to saying, they can read 10-12 moves ahead on each "board", and then come up with the proper ordering for each of these sequences (timing is important, just like in chess).

Playing go is a lot more like playing a very positional, closed game in chess (although there's loads of a local tactical play, you can't keep tempo with all local tactical plays, your opponent will simply ignore you and cut a large swathe somewhere else, a lot of setup is usually required to create a situation where the entire game hinges on a combination), and it is that kind of chess that, as far as I remember humans held off the computers for the longest amount of time.

Why? because humans could "see" the formation that would solve the position, yes there was calculation involved, but their intuition guided them to the solution, and then they calculated backwards from that. Go is similar in that sense, except that that kind of calculation that computers have trouble with is omnipresent. 

I'm not reading that sentiment here.

I started my post from the first post:

"I've been wondering if the game of Go was actually superior in these two and other areas.  Knowledgeable input requested."

To which my comment was primarily directed. The limit of most strategy games is in the minds of the people playing it, not in the game itself.  It's impossible to judge if go has deeper or shallower strategy, since human beings will play either chess or go as deep as they possibly can, and both have their share of geniuses.

Okay, but I don't see anything in the thread that is dangerously close to posing one game better than the other.

The OP asked a sincere question, no more, as I see it. Less than insinuating.

Haha, well, maybe I'm just sensitive, I've seen many a thread on go forums degrade from an innocent comment like that down into a freakish morass of flaming and game/idol-worship :)

Thanks for the thread clarification, goldendog.