http://deadspin.com/federal-judge-rejec ... 1789364005

Hard to believe this was even in dispute. Can you imagine if Go matches like the Meijin were somehow not allowed to be played live on various online platforms?

Hi Drew, I feel the opposite: whenever anything is on the line --
$, power, other resources, etc. --
there's bound to be some people, somewhere, fighting.
No dispute is very strange to me ( just look at various random threads here! )
Peace & harmony are very rare and precious indeed.No need to imagine; been the norm for years:
KGS cannot condone unauthorized broadcast of any of the usual suspects: Meijin, Kisei, Judan, Oza, etc.

Ed: Karjakin ought to have followed your advice, taking a another meaning of 和 - draw!



I think it is important to point up an important difference with chess. In Japan, at least, the go world for the most part acts as a community and acts responsibly and irrespective of whether it's allowed by law or not (still an unknown in Japan), they do not publish games until the sponsor has had first bite at the cherry (which is really all the chess sponsors were asking for in the present world championship dispute - copyright was a side issue). They often go even further. Even after the Nihon Ki-in has published a game, someone else writing about that game later will often partly respect copyright by publishing only some of the moves. There is a common but thoughtful further practice: the published game will typically have the event with the name of the sponsor included. This is all done voluntarily and in a responsible manner.

I suspect that this correlates with the fact that Japan alone can provide between 2 and 4 times the number of players who can make a decent living from tournaments alone compared with chess worldwide (100, according to Stephen Moss), and is also weathering the recent lack of international success.

The apparent counter-argument is that Korea and China also each host more well-off pro tournament players than there are chess, yet community behaviour is more like the chess world than in Japan. There are probably deeper cultural differences at play, anyway, but I always have a sneaking doubt about the pro scene in Korea and China. The turnover of sponsors is much, much higher than in Japan. Their fickleness is one reason for Mickey Mouse time limits in Korea. In China, I have a feeling that sponsorship is sustained to a large degree by nationalism and the long battle to get to the top of the go world. Now that they are there (and/or now that AlphaGo is on the scene) will that continue?

Obviously that description of the go scene can be challenged or nuanced, and I don't know how the future will pan out. But I do feel convinced that the lack of a strongly cohesive community in the Japanese go sense has harmed chess at pro level, and things will get worse in chess. That seems to me to be a good reason to think hard about fostering community spirit in western go. Join your go association, join a real club, support go publishers - that would be a good start. I've just bought another two Nihon Ki-in books, so I've done my bit

Hi John & others more knowledgeable about Japanese culture:
Curious about the accuracy of this text:One can also find certain parallels of the very good collective human behaviors in times of crisis:

• The drop in crime rates in NY immediately after 9/11.
• The calm, orderly behaviors of society immediately after the 2011 Tōhoku tsunami.

Curious how much of this aspect of their culture seeped into Japanese go.

The Carlsen v. Karjakin time settings seem to be similar to the one-day title matches in Japan ? ( About 8 hours total, finished in one day ? )

The pro chess world is rife with "mickey mouse" time limits which are perhaps imposed by the tournament sponsors. As an example it seems that often pro tournaments include a blitz phase in addition to longer time limit sections and winning the tournament is partially determined by winning the blitz section. I could be wrong about this because I am not an active chess player. The blitz section sometimes use unusual starting arrangements of the pieces. In my opinion these blitz sections are for the entertainment of the fans, thus increasing the exposure of the sponsor, i.e. advertising. I wouldn't begrudge the sponsor their desire for advertising to fans but I don't like the inclusion of the blitz section in determining the overall tournament winner.

The blitz phase is used to determine the seeds for the "classical" phase, so the extent to which it "partially determines the winner" is limited to sometimes awarding someone an extra game with the white, rather than the black, pieces. The alternative would be for the seeds to be random.

Ties after the tournament is over are also sometimes broken by rapid games. There are some tiebreak formulas that could be used instead based on the tournament results themselves, but they are all insufficient in some way. I think everyone involved prefers the rapid tiebreak games.

There are very occasional tournaments with random arrangements of the pieces (called "Chess960" or "Fischer Random") but they are not rated and are "for entertainment only".

The Carlsen-Karjakin time limits are about as long as they get these days. In the past, serious games would sometimes stretch into a second day if enough moves were played, but now that superhuman computer analysis is available, adjournments are considered a bad idea.

I've been reading bits of Reddit Chess for the last couple of days. it is interesting to see how some things are viewed by chess players.

About time limits there seem to quite diverse opinions, with people favoring even faster games, others blaming the shortening on sponsors, tv, etc. Nothing really different than what we can read here about go

What is more interesting to me is the debate about the number of games per match. I don't remember seeing it mentioned here but thinking about it, I do find it is something quite important too. I like best-of-seven matches because they provide more opportunities for the match to come to a life of its own. There is more room for comebacks, reversals, questions about how much psychological factors will weigh in, and the likes. A best-of-three generally cannot, I feel, provide the same depth*.

Another interesting aspect is the use of computers by the chess community during title matches. Some people seem use it to solve their self-esteem issues (ah, how satisfying for the ego to call grandmasters on their "blunders" while safely ensconced in a couch with a chess engine running on your laptop). But on the whole, I find that there isn't that much discussion about moves. I don't know if this is due to people having engines at their disposal, to Reddit not being very user friendly as far as technical discussions are concerned, or to some other reason.

As a side note, there was a move played by Carlsen in game 10 (move 62 I think) which wasn't considered the best move by engines, some people started calling it a blunder, or at least saying that Carlsen lost a bit of his advantage, but it seems that after it was played engines suddenly went crazy and Carlsen's winning rates skyrocketed.
Not sure if this is exactly what happened or if this is just how some redditors have experienced it, but if true, it is nice to see that engines can be surprised.


* I vaguely remember reading something about a player (Sakata?) commenting on how Go Seigen was at his best in the second half of ten-game matches and that only there could his full strength be experienced.

In the chess world this is a dispute that has been going on for a very long time already. Emanuel Lasker had already tried to establish that the moves of a game of chess are the shared intelectual property of the players. He was seeking royalties for his games that were published. He never succeeded, and neither did anybody else.

In most countries the ruling is that broadcasting moves is sort of equivalent of reporting on a sporting event, be it of the result or of the ongoing events.

The only option left was to claim rights to the footage. This is what the organizers did here as well. The additional attempt to also claim rights to the actual moves failed as it always has done sinces Lasker's times.

Brace for what is going to happen in the go world as soon as everybody has an AlphaGo on their smart phone. The move mentioned by you is prime but rare example for a move where human judgement is still superior to that of the engines. It simplyfies into a trivially won endgame after White's trade of knights at move 67, but the silicon friends usually don't recognize quickly that it is so easily won. Especially, if they hadn't calculated for a sufficient amount of time.

Nevertheless people are prepared to berate top grandmasters as idiots when they play moves that are not the engines' first choice, even though there are lots of reasons human competitors might have for that. For example, the principal variation that is backing up the evaluation of an engine's first choice might simply rely on a tactic deep down in the tree, and a human grandmaster might decide that his margin for error going down a complicated line is too big.

Yes, I'm not really looking forward to that day if only for all the cheating problems it will create.

I did some more reading and saw it was mentioned a few times that there are certain types of positions (don't remember which ones) the computers are not confortable with, i. e. that their judgement could not really be trusted when such positions arise. Maybe I misunderstood, but if not, I find it quite surprising.

The most prominent positions that engines cannot evaluate are fortresses. Another type are some theoretical endgames where a material advantage is not sufficient for a win.

Below is a simple fortress where the advantage of queen versus rook is not enough to win. As you can see, the endgame is guaranteed drawn according to the endgame table base entry.

Nevertheless the engine asserts a winning advantage for white. Of course, no progress can be made, and the evaluation of exactly the material value of a queen versus a rook reflects that. But to an engine this is still a winning advantage, even though it doesn't improve towards checkmate. So, if this position occurred along a critical variation, the engine might trick itself into drawing while believing it is winning.

Thanks for the example!

Interesting how such simple things can end up being loopholes for engines, it reminds me of the ladder problems many go playing bots have.

It will be a few years before even a derated version of AlphaGo is available on smartphones. Even for a version that can play at pro shodan level, less than 10 years is a good estimate.

It’s getting off-topic, but …


Remember CrazyStone with it’s server connection subscription.

We’ll very soon live in a world where we’ll have constant ’net access, many of us already have it. And the ’net is getting faster also (though of course with bandwidth and speed the noise and clogging increases at the same rate as the signal).

I can imagine a few ways Go programmers could offer the “services” of their programs online … and this is actually what I expect DeepMind/AlphaGo to do in a year or two, maybe as a spin-off company (Google is NOT Xerox!).

I mean, with all the recent and current hullaballoo about Zen19[insert character] playing on KGS (yes, I’m also excited and sometimes I also watch her playing!) I can well imagine that we could see soon could see some “pay-for-play” with pro-strength bots, I can imagine a LOT of strong players willing to spend $1 or a € or perhaps quite some more for “a game against the bot that beat [insert name]”.

Am I totally off here?

_________________
“The only difference between me and a madman is that I’m not mad.” — Salvador Dali ★ Play a slooooow correspondence game with me on OGS?

My initial reaction was "huh"?

But then, I didn't have much connection with small children and schools while I was in Japan.

But then.... I've been told of some very stiff entrance exams to influential kindergartens and primary schools ("influential" as in: their graduates are fast-tracked to the next stage of influential schooling, all the way up to university and the "best" jobs). And I've seen a lot of rather small children going to "juku" (after-school cram schools), but then, maybe they are just going there to add some polish to all that socializing they get at school all day.

But then.... maybe once they are in those good schools, there are no internal exams until 4th grade?

Dunno, basically.

Also dunno how to persuade this software to show the image in the quote....

And here I am, calling GnuGo(6k) an adequate and satisfactory partner.


There is something quite pleasant about giving handicap to a program that used to wipe the floor with you.

I wonder if they will ever host chess in the US again since the organizers got shot down by a US federal judge for not getting their way.

Maybe the user has to first save the image locally to their computer (or any device), then [include] it again inside the quote...

There is no link to the image in the quote, and the page source doesn't have anything useful either.

If the user can view the image from the original post, and if they can then save it on their computer or device, then they can [Add the file] and [Place inline] it on any future posts. Example: