|
adante posted:some interesting reading about the state of (and predictions for) go ai http://www.spectrum.ieee.org/oct07/5552 This article is a great read, it's sparked an interest in the game but the article itself was neat as hell on it's own.
|
![]() |
|
![]()
|
# ? Feb 15, 2025 12:57 |
|
Garfie posted:I personally expect Go to be solved by impossible bots within our lifetime. Computing power is one of the main items holding it back. Once Chess is mastered by AI, people will turn their sights to Go for a challenge in computing. It is practically impossible to bruteforce Go. Chess wasn't mastered - Deep Blue simply applied simple rules very rapidly in every possible situation and looked at the result, and then selected the best one. The Spectrum article (best part of being an IEEE member, honestly) goes into great detail on the actual numbers, but it would be impossible to approach Go in the same fashion computationally without some MAJOR advancements in power.
|
![]() |
|
Nodrog posted:I think the most invaluable resource is the Go Teaching Ladder. Download and read the reviews that have been done on the games of people around your strength - chances are, youll be making the same mistakes that theyre making. You can also submit your own games to be reviewed. This is really outstanding. It's just an incredibly efficient way to burn through a bunch of concepts/examples without it seeming like a dry textbook.
|
![]() |
|
Peantoo posted:Anywhere from 20 minutes to 2 hours. It depends on how long you like to think to be honest. I prefer 30 minutes on each side with 5 30 second countdowns (byo-yomis). Others like 5 minutes a side with 5 second byo-yomis. I'd say 20 minutes per side is about average, which equals to about 45 minutes total for a good thinking game. The time flies by too. If you've only got a couple minutes free or you don't feel like playing out an entire game, https://www.goproblems.com is a great resource for Go problems, which are always good to play.
|
![]() |
|
Two Percent posted:As I understand it, the absurd amount of possible plays is the reason there's no Go AI, is this right? Because I'd like to have a computer opponent to train on every once in a while but it seems we'll have to wait for quantum computers for that. The other problem is that positional evaluation is extremely difficult. In order to decide whether one move/sequence is better than another, you need a way of evaluating the board positions they produce. In chess you can use some heuristics for this - like the value of pieces both sides are left with, number of central squares controlled and so on - but evaluating positions in Go tends to be a lot harder than this (especially during the opening). The things which make one position better than another are more abstract than they are in other games, and the reasoning here tends to be more intuitive and difficult to quantify. The quality of a board position in Go is very holistic and depends on global features such as the relationship between different groups and the overall balance of power, and this is probably more difficult to capture in a formula than the quality of a local position. Ashenai posted:There are Go AIs. They are far, far worse comparitively than chess AIs; the best one is 4 kyu, if I remember correctly, which is well shy of Master status. It's still infinitely better than most of us will ever be, though. edit: There was a 9-handicap game played between the European Go champion and the best computer Go program (running on a supercomputer) played a couple of months ago, the sgf is available here if anyone wants to watch it. The program lost, but appparently it was a close game. The pro said after that he thought the computer was approaching dan level. Nodrog fucked around with this message at 01:34 on Jul 8, 2008 |
![]() |
|
double post
|
![]() |
|
Peantoo posted:Yo woss, just letting you know I am still reading your books and didn't give up on Go. good to hear peantoo, drop in sometime
|
![]() |
|
'Quiting' the game for a while is pretty reasonable. I've been playing off and on since 2001 or so and I love the game even though I'm only moderately strong.
|
![]() |
|
Posting to say that I played my first two games tonight and got my poo poo rather rocked. Had a good time and finally understand what hell the game is all about. Yeah, I'm terrible now but I'd love to play more and get better. I didn't know such a binary game could be so drat complicated. Thanks for the games and help tonight. And for the OP.
|
![]() |
|
Picked this game back up a few days ago and I love it as much as always (despite my awfulness). I really just enjoy the floaty push-pull of the pieces and the pleasant conversation it envokes. But wow do I suck. I'm qxan at KGS so if anyone good wants to message me and help, I would be grateful.
|
![]() |
|
Wow this is a pain but one of the KGS admins saw activity in the SA room and we are kinda flagged for breaking some of their terms of service involving language and conduct so go in and say some reasonably polite stuff about getting into the private room and we can get you an invite. Try not to curse or a banhammer stroker named 'BigDoug' will give you a 2-5 day ban from kgs or something.
|
![]() |
|
Keep up the good work goons. Too few people play this game. Myself - I started years ago and never really got better than 5k. I play as Segovia and am willing to help a few n00bs here or there, but my skills are worthless ![]()
|
![]() |
|
Oh segway, you're not worthless, you just need to play more! ![]()
|
![]() |
|
Segway just got ban-hammered for not swearing.
|
![]() |
|
Gimpy Joe posted:Wow this is a pain but one of the KGS admins saw activity in the SA room and we are kinda flagged for breaking some of their terms of service involving language and conduct so go in and say some reasonably polite stuff about getting into the private room and we can get you an invite. Try not to curse or a banhammer stroker named 'BigDoug' will give you a 2-5 day ban from kgs or something. Ha. Goons just can't fit in anywhere ![]()
|
![]() |
|
I hope he doesn't read tonight's backlog and ban me. ![]()
|
![]() |
|
Found this thread today, and seriously, Go is crazy. It's great fun, but I get raped by everyone I play. Looking forward to playing with you guys and getting better. And thanks Plan9 on KGS for helping me out.
|
![]() |
|
Go is a game of devastating reversals (even during end-game recaps)
|
![]() |
|
go is great. come play go...
|
![]() |
|
Go is fantastic. It's very fun and deep, most players are by and large very polite and friendly, and it'll Also the SA Go community has some REALLY good teachers so if you've ever thought of taking up the game you'd pretty much have to be an idiot not to take this opportunity and end up paying some guy $50 an hour for lessons later.
|
![]() |
|
I would have ended up hating Go if it wasn't for the great SA community on the KGS server. Great people. I'm telling you. PS Come play Go.
|
![]() |
Oh, what the hell? I've been posting in that Traditional Games forum thread instead of this one. I'll second qxan's sentiment. It's great having a little room of (some very well skilled) regulars on kgs.
|
|
![]() |
|
Beginners might want to give Igowin a look. It's a free demo version of the commercial software Many Faces of Go, and lets you play 9x9 games against the computer. It's no substitute for larger-board games against actual humans, but it's an easy way to play a lot of quick games (and win some! It adjusts the difficulty dynamically from game to game) and at the very least to help develop an eye for making live groups and get a feel for some common sorts of fights. Other than that, nthing the recommendations for the beginners section on Sensei's Library, and a bunch of the easier problems on goproblems.com (which you should drill yourself on over and over until you can answer them immediately).
|
![]() |
|
Hey guys. I play as 'bojo' on KGS, been awhile since I've stopped into the KGS room. I'll pop in there from now on, now that I'm actively playing again. Slightly related, but there was an article about me in the Kyoto Newspaper here a couple weeks ago. Been playing at a local club in the next town over with a bunch of 70+ year old guys. They peg me as a 2d-3d, but I blame that on localized rank inflation. They don't believe me when I say I am only 4k on the internet. Because of the article I've also gotten invited to play at people's houses too. One guy had an $8k (USD) board with a Meijin sig on the bottom, and showed me his 4d Nihon Ki-in certificate from 1973 (had Kitani's sig on it). Then his buddy came over and we played a game, then later that week gave me a spare folding board, bowls, and stones he had lying around. ![]() My rough translation (it's not perfect, and a couple mistakes): quote:"Western Male, Plays Go" (top title) - "Kyotamba Jr High ALT" (side
|
![]() |
|
Hahaha, bojo. I can't believe you're making an appearance now. The SA room is vacant, but you're more than welcome to the new room man.
|
![]() |
|
So are more people going to join or what? Come on, don't stem the tide now! There are lots of noobs who need other noobs to train with!
|
![]() |
|
Peantoo posted:Hahaha, bojo. I can't believe you're making an appearance now. The SA room is vacant, but you're more than welcome to the new room man.
|
![]() |
I'll probably be on tonight. I've got a couple "game dates," so I might as well make an evening of it.
|
|
![]() |
|
Doomclown posted:What's the new room? ITGO. Go to the SA room under the social tab and ask anyone there (nicely) if you can join the super secret club. We'll add you to the list.
|
![]() |
|
Has anyone in here ever read Mathematical Go? Supposedly it gives an algorithm for perfect late endgames (which I hear from my Go playing friends are not as significant in Go as in Chess). I've studied the math (combinatorial game theory) behind it, and I'd love to hear from a Go player who read the book. Also, I've started to learn go and then forgot the rules multiple times, but maybe I'll get into it again thanks to this thread.
|
![]() |
|
Does anyone play on IGS? I made an account there because I much prefer Goban to KGS's client, but if everyone's on KGS I might have to make a KGS account too.
|
![]() |
|
Some of the SA members go to IGS and a few other accounts for various reasons but KGS is the place we all hang. I personally don't use IGS so I don't know what else to tell you except that KGS is cooler and better for conversations and community.
|
![]() |
|
helopticor posted:Has anyone in here ever read Mathematical Go? Supposedly it gives an algorithm for perfect late endgames (which I hear from my Go playing friends are not as significant in Go as in Chess). I've studied the math (combinatorial game theory) behind it, and I'd love to hear from a Go player who read the book. Go and chess are backwards, although in both cases the emptier the board is, the more complex it is. It takes about as much effort to study go end-games as chess openings, which is to say it's not really that much effort compared to everything else.
|
![]() |
|
helopticor posted:Has anyone in here ever read Mathematical Go? Supposedly it gives an algorithm for perfect late endgames (which I hear from my Go playing friends are not as significant in Go as in Chess). I've studied the math (combinatorial game theory) behind it, and I'd love to hear from a Go player who read the book. I haven't read it, but I did flip through it in a university library and I thought it was really dry. If you're interested in mathematical endgame there's some interesting content on it at sensei's library mostly written by Bill Spight.
|
![]() |
|
Does the endgame vary by rule set? I know that in China, people quite often play right until the VERY end (you cannot pass), and that part always really annoyed me. Yes, it does take SOME skill, but it seems trivial. I just played a couple of games on the server though and it seems that that rule is at least not part of the Japanese rule set. I've also played that passing costs 1 point. Explain?
|
![]() |
|
Under Chinese rules, the final score is the number of stones you have alive on the board plus the areas they enclose, however prisoners and dead groups do not count towards the result. Every neutral point in these rules is a point, although players often will stop filling in once its clear that the final result wont change due to neutral points being filled in. Sometimes filling them in will cause the other player to have to play within their territory to defend their group, allowing the other player to snatch two neutral points in a row and gain. In Japanese rules, the final score is determined by the areas you've enclosed plus the dead stones you have captured on and off the board. The main difference is really that stones in seki are worth points under Chinese rules. (A seki involves two groups with no eyes and two common liberties, so neither player can move first to capture without becoming captured themselves - so both groups are considered alive)
|
![]() |
|
Is there a version of Go for the iphone? I looked through the app store and the only iphone Go I could find online (iShudan) wasn't there (yet?). edit: Apparently there are more, but I don't see any in the app store and searching for "Go" brings Facebook up as the second hit. ![]() kaptainkaffeine fucked around with this message at 21:37 on Jul 21, 2008 |
![]() |
|
Man, I haven't played in like year. I should get back into Go, even though I was only ever like KGS 12k, tops.
|
![]() |
|
helopticor posted:Has anyone in here ever read Mathematical Go? Supposedly it gives an algorithm for perfect late endgames (which I hear from my Go playing friends are not as significant in Go as in Chess). I've studied the math (combinatorial game theory) behind it, and I'd love to hear from a Go player who read the book. The approach they take is quite cool - they treat Go as being a multi-armed bandit problem, which is a widely used framework in reenforcement learning. The idea is that youre confonted with a slot machine which has several arms, where each arm has a different reward function. You can pull one arm each turn, and when you do this you get a reward corresponding to the arm you pulled. Your task is to maximise your reward over time, which means that you have to estimate the reward function for each bandit-arm and identify which ones are 'best'. When applied to Go, they essentially treat each board position as being a 'bandit', with every possible move corresponding to an 'arm' (more technically each node of the search tree is a seperate bandit). They then use results from the literature on bandits to decide which moves are most profitable to explore (because computation time is limited, you only want to explore the moves which are likely to be good). In order to assess how 'good' each move is, they combine the UCB1 algorithm which has been studied in the context of bandits, with a monte-carlo approach where moves are evaluated by the program playing lots of games of Go against itself where that move is made, and seeing the percentage in which that move leads of victory. The result is they they can cut off a lot more unprofitable branches from the search-tree than they could with standard alpha-beta pruning, and achieve decent results. Nodrog fucked around with this message at 00:51 on Jul 22, 2008 |
![]() |
|
![]()
|
# ? Feb 15, 2025 12:57 |
|
Just jumping in to say that a seki totally can have an eye and that stones inside an eye in a seki are a reasonably major contention point for counting styles.
|
![]() |