American Go E-Journal » Your Move: Readers Write

“It’s interesting to read about the work of University of Edinburgh to use machine learning to improve the level of playing in computers, (Scottish Neural Network Takes Computer Go to Next Level 12/16/2014 EJ)” writes Nin Lei, Distinguished Engineer and CTO, Analytics and Big Data, STG IBM Systems and Technology Group. “However, the title in their article creates an impression that their research is creating a program that can beat the best human players. If their probability of guessing their next move is only 44%, then their chance of guessing it wrong is 56%.  In a sequence of 10 moves, the chance of getting the complete sequence correctly is 0.44 ** 10, which is a very small number.” Noting that checkers “has been solved via machine learning,” Lei says that “it appears it is promising for go as well.” But because machine learning predicates that there is a pattern in the underlying data set, Lei warns that “it could be so complex that machine learning can only attain a certain level of accuracy.  It seems to me a program needs to have very high level of accuracy before it can play a good game at strong human level.” Lei also says that “Since machine learning is based on pattern recognition, I wonder if a professional can trick the program by using moves that may not be optimally locally but will create patterns that the program has not seen before. I applaud the work they are doing,” Lei concludes. “It is innovative by using a different approach than the existing strong computer programs.  It will be interesting to find out if someday they can come up with an algorithm that can improve the accuracy significantly.”
12/22: the chance of getting the complete sequence correctly has been corrected to 0.44 ** 10 (from 0.56).

“In your recent article (Your Move/Readers Write: Where to Play Go in Japan 9/13 EJ), Devin Flake states that the Diamond Go Salon is ‘mainly for women,'” writes” Adam Harding. “I am a long term member of that salon and I would say that DIS (Diamond Igo Salon) is not as much ‘mainly for women,’ but more for young and middle-aged players. The salon owners do run a monthly ‘Igo for women’ session which is for women only.” Harding says that Diamond’s other strong points include “a strong connection to the professional world; the owner runs her program on the Igo/Shogi channel; the atmosphere is that of a high-class wine bar instead of smoky back-room, with drinks and food available and the age range of players is about 20-50 on Wednesdays and Fridays instead of 40-60 as seems to be at most other places.” While Harding says DIS “is most slightly more expensive,” he notes that membership brings the entrance price down to that of other salons.” Click here for DIS lesson and Go Circle information and the club’s instructor listing (all in Japanese).

“You mention that you’re looking for a January 2002 article about go by Katy Kramer (Go Spotting: Northeastern University Magazine 6/7 EJ),” writes Harald Zellerer. “I really liked that article also and republished it on the website of the Amsterdam Go Club.” Click here to read “Go: With the Flow.”

Bob Joyce also sent us a copy of the article, noting that “featured is Sangit Chatterjee, who authored Cosmic Go, Galactic Go, and provided game commentaries for the book Go! More Than a Game by Peter Shotwell.  He describes the game’s complexity as ‘Go is like six chessboards joined together, with all six games happening at the same time.’” Joyce extended special thanks to Joan Lynch, Managing Editor, Marketing and Communications, Northeastern University,who provided a copy of the article.
Editor’s Note: This terrific article would make an excellent handout for local clubs to beginners or at public events.

History is Not Offensive: “Regarding the ‘offensive’ qualities of the ‘Highbrow’ item (Your Move/Readers Write: High/Low Brow Matrix Offensive 9/14 EJ), I’d like to rebut on behalf of the long-gone authors,” writes Peter St. John. “The thesis, which I believe was new in that era, is that appreciation of, and interest in, higher levels of abstraction is ‘high-brow’.  Go is more abstract than chess in the sense of being a level further removed from physical combat, the way Eisenhower immersed in logistics was a level removed from George Patton deploying tanks, who was several levels removed from the gunner pulling a trigger.” St. John also notes that “At the time of the article [1949] the only places to find go in America would be in the math and physics departments of universities. My dad learned, around that time, in a science laboratory from a mimeograph of a German article, because German scholars collaborated with Japanese scholars after the Russo-Japanese war. I urge people not to be offended by history. We can learn from it, not in the sense of learning from an Authority but in the sense of learning from an Experience. The grid is a bit of history.”

Celebrating Progress: Noah Doss agrees, saying that the matrix “simply records historically the type of people who, in the time period observed in the matrix, were most fascinated with go.” He goes on to say that “Nowadays, go is not, in America, a game of the elite, but just because modern man has made some progress in quashing these societal imbalances in some respects doesn’t mean we need be offended by the fact that they once existed. I truly believe go is for everyone and, to be honest, if it was the way of polo or lacrosse in that everyone I met playing go had a trust fund, I would probably not love it so much, coming from a dirty farm town and an undersized stucco house. At a point in time, go was ‘highbrow’ but I think we should honor the fact that it used to be ‘highbrow’ and now we, as a society, have fixed that problem.”

“I am probably in a very small minority on this topic, but I found the ‘High Brow – Low Brow’ matrix (Go Spotting: A “Highbrow” Game in 1949  9/6/2014 EJ) extremely offensive,” writes Gordon Castanza. “As with almost all pseudo sociology that passes as junk science, this crap is another example. Go, if anything is classless, colorblind, and non-age-specific. Besides, the entire matrix applies to any number of people at any given time. Glenmorangie Taghta can be enjoyed by anyone (of legal drinking age); it also knows no class, income level, working condition, color of one’s skin, or country of origin. I know you’re trying to show go in the most positive light possible, and I’ve enjoyed the ‘Go Spotting’ feature in the past. However, to have some C-grade sociologist put go, or any other human endeavor, into a hierarchical range is just plain stupid and intellectually dishonest.”

Correcting the Games Database: “I checked out the AGA game database from a recent E-Journal (AGA Game Database Test Version Online 8/12 EJ) and really liked it!” wrote Shawn Ligocki. “But I noticed that a tournament I participated in seems to be double counted. I went 4-0, so maybe I shouldn’t complain.” We got a number of emails like this, pointing out various errors in the database. Thanks for flagging these; the programmers are working to update and correct the American Go Association Game Database (AGAGD). Comments and corrections should be sent to journal@usgo.org

Looking for Japan Go Tips: “I will be traveling to Japan next spring,” writes Ben Bernstein. “Do you have any advice, or can you point me to a source of information (about where to play go)?” He’s specifically interested in Tokyo, Nagoya and Kyoto; email your tips to journal@usgo.org

Where’s the 2015 Go Congress? “I’ve heard that the next Congress will be in Seattle or in Minneapolis,” writes Wayne Nelson. “Which is it?”
St Paul/Minneapolis, AKA the Twin Cities: see 2015 Congress Website Launches 8/14 EJ.

Congress Player Profiles a Big Help: “I have really been enjoying reading the go player profiles (US Go Congress Player Profiles: Chen, Liang, Lee & Chiu 8/4 EJ, & US Go Congress Player Profiles: Sun, Ko, Koh, Lin, Teng & Ye 8/5 EJ) for the upcoming Go Congress,” writes Dennis Wheeler. “It’s really going to help me get a better idea of who these top level US players are as I watch their games.”
It’s going to help him as a Congress game recorder for the EJ, too; watch for our live broadcasts starting Saturday afternoon with the Pandanet-AGA City League Finals and then the US Open starting Sunday morning. 

More Clossius! “The Clossius commentary (Clossius in the Land of Baduk: At Home Abroad 8/5 EJ) was great,” writes Chris Uzal. “I hope that is not the last one.”
We’re pleased to welcome Shawn Ray as a regular EJ contributor; look for his next column soon!

“I said for decades that I did not think I would ever be beaten by a computer playing go,” writes Phil Straus 2D in response to Go Spotting: IEEE Spectrum 7/16 EJ. “I was wrong. I was first beaten by a computer program on KGS sometime in the last two years. Laurence Sigmond and I watched Deep Blue beat Garry Kasparov at the Philadelphia Convention Center in 1997. To pass the time between moves we, of course, brought a go board. I showed the go board to Hans Berliner, one of the iconic chess programmers. He looked at it, and just shook his head. He said ‘maybe in 20 years.’ I was even more pessimistic. Go looked impossible in 1997. We were both wrong.”
Straus is a former president of the American Go Association. photo: Rémi Coulom and Crazy Stone. Photo: Takashi Osato/WIRED