Slashdot Mirror
Computer Cracks 5x5 Go
gustgr writes "The American Go Association is reporting that Go for the 5x5 board has been solved by the computer program MIGOS, reports the program's creator, Erik Van Der Werk, a professor at the University of Maastricht in Holland. At about a quarter of the full-board version, 5x5 go is miniscule, similar in scale to "solving" 2X2 chess. The fact that a programmer would even consider this a noteworthy challenge is itself a remarkable testament to the game's complexity. Van Der Werk's approach is described in detail in an
article at the Netherlands Organization for Scientific Research (NOSR)."
From the friendly article:
Subject: computer-go: 5x5 Go is solved
Date: Sun, 20 Oct 2002 15:27:04 -0100
From: Erik van der Werf
To: COMPUTER GO MAILING LIST
The fact that an editor would even consider this a newsworthy article is itself a remarkable testament to the site's simplicity.
Funny how the stock market crashed the day before 5X5 Go is solved.
Rock that crushes, Paper & Scissors that don't matter.
Slashdot has a longstanding joke that with every chess article, some wide-eyed enthusiast will blurt out a quick description of Go like he's first to discover it in all the West. Speed is essential! There may be some pasty white guy who does not know the wonder that is Go.
I fully expect someone to breathlessly explain the Great Goodness that is Chess.
Chess is fun. Go is fun. People have generally heard of both. That is all.
Transcend Humanity. Please.
What is this, Classic Slashdot? Next do we get a story on the impending end of the dot-com bubble?
Warning: Apple/Nintendo fangirl. Likes her electronics cute & cuddly. May be rabid.
Sony has released a new devixe, tentatively dubbed the "CD Burner," capable of burning the first second of any of your music CDs! Programmers hope some day to move to the entire first track.
"Kaylee, that's the buffet bar." "But how can we be sure unless we question it?"
5x5 is 1/4 the size of 19x19??? More like 1/14th.
The way that chess games work is they check n ammount of moves into the future. With each iteration into the next move it splits off into a massive tree of moves. As an example, the first iteration has 10 potential moves, the next has 100 and the next has 1,000 With Go as an example there may be 100 potential moves on the first iteration and then 10,000 and then 10,000,000 The number of potential moves grows way faster then in chess.
Check out this for a decent comparison between chess and go for those of you who have been missing out.
Also, dig my sig biotches.
Direct away from face when opening.
In the past couple days, people have been talking about "cracking" an 80 bit hash with a 69 bit effort. It's logarithmic, people. 69 bits is not three-quarters of 80 bits, it's a factor of 0.000488 in terms of the workload to crack it.
SHA-1 is now 0.000488 (4.88*10-4) as strong as it was. And by my calculator, 5x5 go is 4.866*10-161 as hard as a brute-force solution as a 19x19 board would be.
[
Because Go is incalcuably more complex to design a computer program for, there are only two pieces, but they can go anywhere at any time (Ok, not *anywhere at any time* but pretty much), and the number of combinations there are to a simple move is much more difficult than the moves are in chess.
Or so I would assume, I've never actually tried to make a program for either, but it would appear so to anyone who has played more than a few games of each.
I've always believed within my lifetime, chess would be solved. In other words, a computer would come up with the perfect solution to chess so that no matter what moves you possible make, out of the, i dunno, billions, trillions, or higher number of possible moves, the computer knows how to beat you. The simplest comparison I can think of is tic-tac-toe. If you've played tic-tac-toe enough, you've learned that no matter who goes first, someone can always force a cat (tie game). I wonder, is it possible to always force a draw in chess or might it be that whoever goes first can always win? Sure the computing power to figure this out is beyond anything we have now, but with quantum computing and other advancements, I expect to see chess solved in my lifetime.
I heard rumours that there was a solution for "Tic-Tac-Toe" very close to being announced. The only hold up is finding a large enough
distributed network to explore all paths in real-time.
Vintage computer adverts: http://www.vintageadbrowser.com/computers-and-software-ads
For another thing, go is spectacularly more complex than chess. The very best go programs are competition only for weak amateurs. There's an archived NYT article that summarizes the problems reasonably well.
Although the standard go board is 19x19 intersections, the game scales, unlike chess. Things you learn on a small board are sometimes applicable to larger ones. A 5x5 is usually not interesting for human play; most consider 9x9 the minimum size for a worthwhile game. This means that a computer has been programmed to force a guaranteed win at a smaller size, and hopefully paves the way for further development and understanding.
This isn't as much "normalization" as it is "don't take so many drugs when you're designing tables."
But what if it's 2x2 chess with all knights?
Want to improve your Karma? Instead of "Post Anonymously", try the "Post Humously" option.
In my opinion, for it to be chess, it would have to have two kings otherwise no one could win. Therefore 2x2 chess would start with checkmate and is absurd.
$2B OR NOT $2B = $FF
Note that a liberty is an empty spot on the board that is either next to your stone or can be reached by moving across your stones horizontally or vertically. This is great for computer scientists who don't know the game yet, http://brooklyngoclub.org/jc/rulesgo.html:
The Alternating Rule:
Two players, called Black and white, keep alternating moves till the end of the game. Black plays first. A move by a player begins by his placing a stone on an empty intersection of the go board. The first player who cannot put down a stone without breaking a rule loses the game.
The Rule of Capture:
After a stone is placed on the board, all enemy stones which have no liberties are removed. A player's move is not finished until this phase has been completed.
The Rule for Suicide:
Suicide is illegal. Precisely, after a stone has been played, and after the rule of capture has been applied to his enemy stones, if the stone has no liberty, then the move was illegal.
The SuperKo Rule:
A player is not allowed to place down a stone if, after the rule of capture has been applied, the resulting Board position has appeared previously in the game.
Transcend Humanity. Please.
I think it will. We still have weightlifting competitions even though we have forklifts at our disposal.
Transcend Humanity. Please.
AFAIK, the current state of the art of Go on computers is Goemate and Go4++.
GNU Go is actively developed, but it still does not match commercial Go software, ranking 1-2 stones weaker. It is rated from 8 to 9 kru, which is a weak amateur.
Computers have thus far not been too great at cracking go via the usual searching algorithms, as it has a high branching factor - starting at 361, much higher than chess! It is only recently that Go programs have even begun to achieve low levels of competence. Besides the limited searching and pattern recognition of current software, future programs may improve by decomposing Go into 'subgames', allowing it to be more readily attacked.
To have a program which has solved Go (unlike the best chess programs, which are merely at the strength of Grandmasters)
It should be noted that even on a 9x9 board (let alone 19x19), competent amateurs can beat any computer program.
19x19, 13x13, and 9x9 (the "standard" sizes, though 7x7 is fun sometimes), require totally different strategies. 9x9 is pure life and death, 13x13 is mostly fighting, and 19x19 requires a good understanding of balancing influence for defined territory (don't spread your stones too thin while not letting them get bunched up).
For all who don't play go or are new to go, the biggest problem with the 19x19 and even 9x9 computer programs is that the computer can't see the dual threat someone might play with a sequence of moves. For example, you can start to attack a specific section of the board, and use what you played to grab hold of an even larger section of territory, or even kill a large portion of their stones. It's easy to fool the computer in Go.
Actually, the fact that Go only has two pieces is why it's so much more complex. Chess pieces individual behavior is what usually limits the number of moves in Chess. Also, since Chess doesn't easily scale down, 2x2 chess doesn't work: QK QK Neither side can move, since any move they make still leaves their king in check. I guess that means that White looses by default, since White goes first and can't make a legal move. Unless you play by rules like with blitz and don't count check, only actually capturing the king, in which case White always wins (unless he's REALLY dumb).
A 5x5 go board has only 847,288,609,443 possible game states, even including impossible boards. Assuming the relatively tame pace of scoring 100,000 boards per second towards completion, which on a board of that size is trivial, this solution takes a simple brute-force time of 98 days. That solution space can be cut down by almost two orders of magnitude with simple reflection and rotation tricks, implying a realtime tree search space of about a day and a half.
Given that my full board scorer moves faster than that, and given that the university probably has more than one PC to work with, I wonder how it is that anyone can justify this as something larger than a publicity stunt, especially given that none of go's emergent structures even fit onto a 5x5 board.
This is horseshit, in short. Mod story down.
StoneCypher is Full of BS
Since computers have started to beat strong chess players, it *is* taken for granted by many that computers can beat reasonably strong people with today's processing power.
Presently, if a typical geek started playing Go, they would get their ass kicked by the weakest computer for a week or two.
After a month, they would be winning the odd game, if the computer gave them a 3-stone headstart. (Like 3 free moves to start in chess).
After three months, they would win some games in an even match against the weaker programs (Turbo-go)
After six months, they would be winning against a 3-stone or higher handicap for the computer.
Then they find a stronger Go program.
They start to lose every match again.
After another month or so, they start to win on the weaker levels.
Take it six months ahead, and they are smashing the computer in an even match with no handicap, playing white (white moves second) or at lower levels against a 3 or 4 stone handicap.
The only thing that makes the game playable against a computer is that Go has an incredible handicapping system that lets uneven players play against each other.
So what makes this story interesting? Aside from the brute strength issue?
The first moves of the game, often in the corners in roughly a five-by-five area (Joseki) are only recently being evaluated for best move potential...
That can affect the outcome of professional matches played for big $$$$.
But more importantly for people like me, I can't play humans much... Kids, wife and home environment mean I can't spend 30 minutes undisturbed, so playing against human opponents is out for me.
Any technology that makes computer programs stronger, improves algorythms or makes me play harder will keep my morning bus trips interesting.
Because Go programs have got a long way to go if they are easily defeated against a human opponent with just 1 year experience.... Who would be easily classed as a novice let alone just a weak player.
GrpA
Enjoy science fiction? "Turing Evolved" - AI, Mecha, Androids and rail-gun battles. What more could you want?
The canonical game (7 columns, 6 rows) has been solved.
Well, on the one hand go is much harder, etc. etc., other people have explained this already. On the other hand, I don't think it surprised anyone seriously interested in computer go, that 5x5 can be done by brute force. Every serious go player can read out quickly that it is a full-board win for black. If Black's starting move is restricted, it takes a little more care to read it out, but I would be confident to read the out the correct play for both sides in a couple of minutes. Further, the essential key algorithm (position evaluation according to so-called "unconditional territory") used by Erik has long been known.
This is not to belittle Erik van der Werf's achievements. In fact to the contrary. His more interesting program is MAGOG, which plays 9x9 go. AFAIK, in the end of the game, it uses the same algorithm as MIGOS, and thus plays perfectly (given enough time, and not too complicated a position). Before that, it combines traditional goal-directed search (tactical search, "life-and-death-search") with a lot of brute force global search. Although his program is pretty young by computer go standards (ALL the top programs started to get developed in the 80's), it has shown to be a serious competitor in recent computer go tournaments.
And real Go can be played at KGS.
"Never attribute to malice that which can be adequately explained by stupidity." -- Hanlon's Razor
The other posters are correct. If you really RTFFA, you'll find that it was solved October 20th 2002, well over two years ago. Even the link you provide only mentions the corresponding doctoral thesis beind defended recently (January 27th 2005). Perhaps you should RYOFL (Read Your Own Frelling Links).
Honinbo Warrior was coded in UCSD Pascal and really did not play that well, but my boss and a few friends talked me into running ads in some Apple II magazines and marketing it. Working on that program was a fun obsession that lasted about 1 1/2 years.
Go is such a great game. In the 1970s, I got to play exhibition games with Miss Kobyoshi (women's world champion) and Mr. Lee (national champion of South Korea). The high level of their play really blew me away - getting slaughtered was a surprisingly great experience.
The Gnu Go program plays a good game, BTW. It is best to play against human opponents, but give Gnu Go a try also. Just like studying chess, if you get into playing Go, make sure you study complete master games: studying opening, middle game, and end games in isolation just does not cut it.
To put this into even more perspective
..., 9 dan.
In go, players can be given a rank on how strong they are compared to others. It's a fairly simple method.
Everyone starts out at about 30 kyu. As they get stronger, their kyu number decreases till it gets to 1 kyu. At which point starts a new number system that goes upward, starting at 1 dan and goes to 9 dan.
So..
30 Kyu, is weaker then a 29 kyu,... 2 kyu, 1 kyu, 1 dan, 2 dan,
Now that is for amateur rankings. There is a professional ranking system that starts at 1 dan pro and goes to 9 dan pro. I have heard that a 1 dan pro is roughly the same strength as a 7 dan amateur.
There is a handicap system where if you take the rankings of two players and subtract them, it determines the number of handicap stones given to the weaker player. Thus a 10 kyu playing against an 8 kyu, the 10 kyu player gets to play first by placing 2 stones on the board (one set of rules allows black to place the stones anywhere on the board, another set of rules, the stones must be played at specific spots). The rule of thumb is that each handicap stone is worth about 10 points. Another rule of thumb is that each handicap stone "erases" one mistake by the weaker player.
Normally one doesn't play with more then a 9 stone handicap. Mainly because beyond 9 stones, black really isn't "learning" much
To prevent ties, a half point is awarded to white in handicap games, in an even game (where both players are of equal strength), white is given 6.5 points (this has been changing around some -- depending on the rules you are playing with).
Usually after the 1st game or so a 30 kyu player learns enough to drop to around 28 kyu or there abouts.
I have heard that the amount of time and study to go from a 10 kyu to a 1 kyu rank is about the same as going from a 1 dan to a 2 dan.
A game between two weaker players can result in scores of anywhere from just a few points to 100's of points going to the winner. As one gets stronger, the wins are usually only a few points, or someone resigns.
I have seen strong dan and pro players when playing weaker players their goal is to try to get the score within a half point (always in their favor).
In Go, the game really doesn't start to get interesting till about 30 to 50 moves into the game (in chess, the game is usually over at that point).
Currently on one of the online go playing servers, GNU Go (among the top go playing programs -- though not the strongest) is roughly around 11 kyu in strength, A weak dan player can give gnugo a 9 stone handicap and the dan player will still win.
Several years ago, Janice Kim gave the top go playing program a 28 stone handicap and she still won the game (I believe it was a 28 stone game).
To get to a professional level player, it is best to start playing when you are very young. Expect to dedicate your life to the game. To get to a strong amateur dan level, also expect to dedicate a good chunk of your life to the game.
His name is Eric van der Werf.
a professor at the University of Maastricht
He is not a professor. He was a Ph.D. student. He received his Ph.D. title January 27 of this year.
in Holland.
That should be "The Netherlands". Holland is part of The Netherlands, but Maastricht is not located in Holland.
At about a quarter of the full-board version, 5x5 go
That's about 1/14th of a full board (25 points as opposed to 361 points).
is miniscule, similar in scale to "solving" 2X2 chess.
It is similar to solving 5x5 or 6x6 chess.
The fact that a programmer
Calling Van der Werf a "professor" is a bit too much, but calling him a "programmer" is not enough.
would even consider this a noteworthy challenge is itself a remarkable testament to the game's complexity.
Basically, it was not done before, and could be done with a couple of weeks computation time. That's not to belittle Eric's work; it is only a small part of his work. Read his thesis to see what he has done for the field of Go research.
Van Der Werk's
Again, it is "Van der Werf".
approach is described in detail in an article at the Netherlands Organization for Scientific Research (NOSR).
That should be NWO, not NOSR, and the approach is not described in detail in the article. For details, visit Eric's website.