O-Chul's Game

[Optimystik]

Sorry if this has already been discussed. In strip #550 (http://www.giantitp.com/comics/oots0550.html), O-Chul reveals that he is teaching the MitD a game involving placing black and white pebbles on a grid. This raised two questions:

1) Which game is it? At first I thought it was chess, but the various pieces in chess would be difficult to represent with formless pebbles. There are also other board games with black and white pieces, like Othello, or it may even just be checkers with white instead of red.

2) Why is he teaching the MitD anything, much less a game that will clearly increase its knowledge of strategy and tactics? Even without knowing what game it is, it's likely to be a strategic game since (a) a strategy game is likely the only one a paladin would learn or care to teach, and (b) the MitD comments on how "hard" it is. O-Chul's sotto voce comment indicates he is intentionally furthering the monster's education... but to what end?

Thoughts, playground?

[Kronski]

The game is called Go, I think.

Go is a game in which you place white and black pieces on a grid, and must surround enemy pieces and territory to win.

[Wolsnik_of_Filbawn]

Go is the game he is referring to, it is a game of strategy involving small, round, black and white stones. It is of Chinese origin, but was popularized by the Japanese:

http://en.wikipedia.org/wiki/Go_(game)

[TheSummoner]

Yes, the game is Go

I doubt O'Chul is teaching the monster it for any reason other than the monster is his friend and it gives them both a way to keep entertained without being thrown into a vat of acid.

[Jimorian]

1) The game he's teaching the MitD is Go (http://en.wikipedia.org/wiki/Go_(board_game)), and it is indeed a very deeply strategic game. Computers are no where near as good at Go as they are at Chess to give one indication.

2) I think O-Chul sees that the MitD has a lot of potential if he'd only learn to think beyond immediate wants and needs. A good debate has been had that the Monster is at worst Neutral alignment, but his naivete and eagerness to please Xykon and Redcloak are really all that are keeping him in the Evil camp. If he could see consequences beyond that, there's at least a possibility that he'd change sides or at least refuse to do their bidding so blindly. I think O-Chul understands exactly how much of a game changer the MitD will be when the ultimate confrontation comes.

EDIT: Wow, triple ninja'd! :smalltongue:

[Optimystik]

I didn't even think of Go! I do know what it is, but it didn't occur to me. *smacks forehead*

That still doesn't explain why. O-Chul is making a very dangerous gamble here - educating the monster may make it more sympathetic to the Azurites' cause, but it could just as easily use its newfound knowledge to make life much harder for the forces of goodness. Wouldn't it be a better use of his time to focus on the MitD's moral education rather than its tactical one?

[Eloel]

1) The game he's teaching the MitD is Go (http://en.wikipedia.org/wiki/Go_(board_game)), and it is indeed a very deeply strategic game. Computers are no where near as good at Go as they are at Chess to give one indication.

It gives NO indications. Computers don't "calculate", "strategize" or "think". They're given millions of games to "memorize".

Go is not as well played as Chess by computers, because no one cares about Go enough to program a good Go program. Also the gameplay is far more easier in Go than in Chess, which could indicate how deep the "strategy" is, going by your rules.

[Jimorian]

That still doesn't explain why. O-Chul is making a very dangerous gamble here - educating the monster may make it more sympathetic to the Azurites' cause, but it could just as easily use its newfound knowledge to make life much harder for the forces of goodness. Wouldn't it be a better use of his time to focus on the MitD's moral education rather than its tactical one?

In a lot of eastern philosophy, there's a linkage between general education and self-improvement and self-awareness, and moral development. Martial arts for one example. With the great power to be able to kick somebody's butt all over, comes great responsibility to see that it doesn't happen except under dire circumstances.

Simply by making MitD aware of consequences of an action, particularly if it's beyond just a one or two-move outlook, should in theory lead to an understanding of a moral code. Now you may be completely right and the wrong lessons may get learned, or the MitD's natural alignment very well may be evil, but O-Chul seems to think it's worth the risk.


It gives NO indications. Computers don't "calculate", "strategize" or "think". They're given millions of games to "memorize".

Go is not as well played as Chess by computers, because no one cares about Go enough to program a good Go program. Also the gameplay is far more easier in Go than in Chess, which could indicate how deep the "strategy" is, going by your rules.

Go doesn't work for brute force computing techniques because the decision tree gets insanely huge far faster than for chess -- that's just a function of the board and the number of potential moves. Tactics on the other hand are something that they excel at -- winning a corner battle is all about raw computing power. But extend beyond that battle, and why is placing a solo piece 7 spaces away a winning move, while 8 spaces away loses the game?

Go was never programmed with the vigor of chess because computer scientists always knew that they didn't have the power to accomplish what they could with chess. Now that chess has been "conquered" (computers can beat the human world champion (as opposed to being "solved" like checkers, where the best computer will never lose to a human)), Go is a likely next target -- both for the raw challenge, and because it's different enough where they'll learn new lessons from the attempt.

I've played tournament chess for years, and have followed the development of computer chess since around 1982. I dabble in Go, but there's still strategic conceptual leaps that I can't get my brain around, even though I can "calculate" tactics fairly well. The fact that Go is more strategic than chess isn't a slight, it's the way it is, just like chess tactics are more involved than Go's because there are a greater variety of pieces to coordinate.

[Rotipher]

O-Chul is making a very dangerous gamble here - educating the monster may make it more sympathetic to the Azurites' cause, but it could just as easily use its newfound knowledge to make life much harder for the forces of goodness. Wouldn't it be a better use of his time to focus on the MitD's moral education rather than its tactical one?

That would only pose a risk if the MitD were in a position of command. Even cursory observation of how Xykon and Redcloak treat the monster would've taught O-Chul that they'd never consider placing it in charge of troops. (For good reason, too, after the "Taco Night" incident! :smallwink:) The tactical insights it might learn from Go aren't relevant to the circumstances under which Team Evil might deploy it, as Monster-san is apparantly a solo shock-trooper, hence can't execute the maneuvers (e.g. surrounding the enemy) that Go pieces perform cooperatively.

[Kaytara]

Besides, what does O-Chul have to lose? They already have a strategist on the team, i.e. Redcloak. Even if the MitD remains on the evil side, him having a better understanding of strategy won't necessarily tilt the balance.

And on the other hand, there's the chance that with learning to think for himself he might learn to make his own judgements and decisions, as O-Chul has already been prompting him to. It's not really a gamble, more of a shaky investment.

[Studoku]

I know it's a little off topic, but this thread made me lose the game.

[kerberos]

Besides, what does O-Chul have to lose? They already have a strategist on the team, i.e. Redcloak. Even if the MitD remains on the evil side, him having a better understanding of strategy won't necessarily tilt the balance.

And on the other hand, there's the chance that with learning to think for himself he might learn to make his own judgements and decisions, as O-Chul has already been prompting him to. It's not really a gamble, more of a shaky investment.

Also let's be honest, the MitD is catastrophically stupid. Making it slightly less catastrophically stupid is not going to make a significant difference, even if there is a link between Go-strategy and military strategy.

[hamishspence]

Stupid, or just very childish?

[David Argall]

Now while the game is likely Go, we have to keep in mind the "What is this 'Japan' you speak of?" factor. There is no necessity the game is more than vaguely similar.

As to why O-Chul is playing, what else is there to do? Being a prisoner is so boring that O-Chul might almost welcome the torture as simply something different. Playing a game is really popular as a timepasser. [I have DMed postal D&D games for a number of prisoners, some of which were quite passionate about the game. But those that got out stopped playing.]

There is no good reason to think O-Chul has any deeper plan than to chase away the boredom. We have no evidence that he is of any deep intellect and he could be somewhat deficient. And if he is not, he should be aware that some of the people he plays with are real jerks, and be skeptical of any idea he can reform the MitD this way.

[Optimystik]

I know it's a little off topic, but this thread made me lose the game.

Install the anti-virus. (http://xkcd.com/391/) :smallwink:


As to why O-Chul is playing, what else is there to do? Being a prisoner is so boring that O-Chul might almost welcome the torture as simply something different. Playing a game is really popular as a timepasser. [I have DMed postal D&D games for a number of prisoners, some of which were quite passionate about the game. But those that got out stopped playing.]

While boredom is a possibility, I don't think it's likely. Truly pious people are exceedingly patient - if they are in a situation beyond their control, either their deity wants them there or will soon provide an opportunity for them to extricate themselves. Observe Miko, who was content to remain in prison until she saw a sign; observe also Durkon, who was separated from his team for hours in a dungeon without getting flustered AND who stood sentry for them in a forest despite his overpowering dendrophobia.

O-Chul is no exception. His piety is such that he interprets the rain (much as Durkon did) as a sign that he is keeping his faith well. He does not chafe or act bitter at his situation, and takes every available opportunity to strike at Xykon despite being unarmed. I doubt he would do anything merely out of boredom; his every action is directed at avenging his beloved city.


There is no good reason to think O-Chul has any deeper plan than to chase away the boredom. We have no evidence that he is of any deep intellect and he could be somewhat deficient. And if he is not, he should be aware that some of the people he plays with are real jerks, and be skeptical of any idea he can reform the MitD this way.

You're underestimating O-Chul. (No, I have no intention of turning this into another Chuck Norris-esque adulation thread.) He may not be a scholar, but he is a tactical genius, as shown by his actions in the shark tank. He also outsmarted Redcloak on the wall. What basis do you have for thinking he might be deficient, his failure against impossible odds?

More to the point, he is CLEARLY up to something! Why else would he comment on how quickly the MitD is learning his game? If it was merely an idle diversion, the speed at which Monster-san picks up on it would be irrelevant to him.

[chibibar]

Out of comic context.

Paladins requires

Strength
Constitution
Wisdom
Charisma

So... while Int is not on the list, it may be a "dump" stat, but Wisdom can make up the difference in terms of knowledge. I am pretty sure O'Chul is wise being one of the toughest Paladin out there. He does have ability to lead (Hinjo did assign him there) so O'chul is no dummy. Tactically he should be able to handle many situation (even impossible odds and survive) which has been shown. I there isn't a time where he has shown incompetence in his action (yet)

[ericgrau]

It gives NO indications. Computers don't "calculate", "strategize" or "think". They're given millions of games to "memorize".

Go is not as well played as Chess by computers, because no one cares about Go enough to program a good Go program. Also the gameplay is far more easier in Go than in Chess, which could indicate how deep the "strategy" is, going by your rules.

People try hard to make a good Go program; it's a major geek goal in fact. But it is way too hard for a computer to handle. Computers still can't strategize the slightest bit, they can only try every combination (or sometimes only some/most of the combinations with some good pruning). Go has way too many combinations thus computers are lousy at it. They don't memorize much; even one game done this way would take way more memory than the whole world has. They do have saved opening move books for chess (the same ones humans learn), but never entire games.

I've mused about adding an invisible "spy" piece to chess. It wouldn't even need to be that strong. This simple change would add an insane number of possible combinations since after only a few moves it could be in a large number of places. Thus it would require a million times more computation to solve. This simple change would make computers lousy at chess. Likewise playing on a 16x16 board or any similar thing which, while it may not have much impact on a human, would really throw a wrench in the works for a computer. The only reason a computer can handle the game at all is b/c of the fairly limited options in chess.

[admerwill]

Fun fact Microsoft in the HPC lab has started building a concerted effort to topple the Go problem. I work as one of the facilities people in that area of campus, and they have dedicated substantial (more then 20) offices, and a full lab of computers, including several crays in order to investigate paths for this great problem.

[Jimorian]

That's very cool. It sounds like the kind of problem that a working quantum computer might be able to tackle someday. For existing technology, I'd think something like having a processor for each space, calculate a weighted value for each space to scream "Me me, pick me!" to a central decision processor that chooses the best one. As spaces get filled, those units can then go help the others or shift to "strategic" thinking.

Ob-OOTS: Maybe V's 4 words are "I understand Go completely." :smallwink:

[Dalenthas]

O'Chul is playing Go with the MitD because he's BORED. Plus, it's probably his favorite boardgame (and/or the only one he can make a board of with the tools at hand), given the vaguely eastern setting of Azure City and the eastern origins of Go. And he probably got fed up with playing Monopoly with the Monster because he kept eating the money.:smallbiggrin:

Go is fantastically complex. Saying that computers aren't as good at as chess because fewer people worked on the problem than with chess... is just flat out ignorant. Wheras a single person can write a program that can reliably NEVER LOSE Tic-Tac-Toe in about an hour, Chess is orders of magnatudes more complex and Go is even more complex than that. Getting a computer to play a game like, say, Warhammer or D&D and you'll cry at how bad it is. (Well, D&D 3.5 at least. 4th edition could probably do OK (yes, that is a poke at 4E)).

To give you an idea of how complex Go is, there are more possible ways for a game of Go to be played than there are atoms in the solar system. And for a pure strategy game with no luck factors whatsoever and finite boardspace, that's pretty damned complex.

[admerwill]

I think that's more or less what there going for at HPC... mass scale parrel computing . the real problem comes in when predicting what your opponent is going to do next... its not just about what helps you the most, its also about what hurts the other guy the most... and predicting what the plan is .

[Optimystik]

O'Chul is playing Go with the MitD because he's BORED.

Again, I see no evidence of this. If anything, O-Chul has too much excitement being a prisoner of the very inventive Xykon.

And I'll ask you the same question I asked David: if he is playing Go merely to pass the time, why is he concerned with the MitD's progress?

[TheSummoner]

Chess computers work by examining every possible outcome of every possible move (usually they look to about 20 moves into the future) and then deciding which course of action would be best. There is no memorization of every game ever, just a great deal of foresight. The reason it works for chess and not for Go is because theres is a great limit to what chess pieces can do. Even the Queen, the most versitile piece in chess can move to at most 27 different places in any given turn. In Go however, a piece can be placed in almost any unoccupied place on the board on any turn. Sure, there are more possibilities for the next 20 moves in chess than I'm willing to calculate, but the possibilities for Go blow that out of the water...


[...] a program that can reliably NEVER LOSE Tic-Tac-Toe
*raises hand high* I never lose Tic-Tac-Toe! Ever!!! People can force a draw, but its impossible for anyone to beat me! *childish glee*

[Optimystik]

*raises hand high* I never lose Tic-Tac-Toe! Ever!!! People can force a draw, but its impossible for anyone to beat me! *childish glee*

Consistently forcing a draw does count as never losing. :smallwink:

[David Argall]

While boredom is a possibility, I don't think it's likely. Truly pious people are exceedingly patient - if they are in a situation beyond their control, either their deity wants them there or will soon provide an opportunity for them to extricate themselves. Observe Miko, who was content to remain in prison until she saw a sign; observe also Durkon, who was separated from his team for hours in a dungeon without getting flustered AND who stood sentry for them in a forest despite his overpowering dendrophobia.
I see two people who within a couple of hours of standing around went completely off the rails. That's not very encouraging for O-Chul having to suffer over 100 days.


You're underestimating O-Chul. He may not be a scholar, but he is a tactical genius, as shown by his actions in the shark tank.
On the available evidence this seems to be a fairly typical feat for a 10th level character.


He also outsmarted Redcloak on the wall.
What makes you think this? He tried to lie and failed.


What basis do you have for thinking he might be deficient?
The simple fact of being a paladin will do for starters. And being a fighter-paladin likely makes it worse. Of course his inability to catch Shojo doesn't speak well for him.


More to the point, he is CLEARLY up to something! Why else would he comment on how quickly the MitD is learning his game? If it was merely an idle diversion, the speed at which Monster-san picks up on it would be irrelevant to him.
And it is. But telling the beginner he is improving is a good way to keep him playing. So he tells him that to have somebody to play and thus avoid boredom.

[Optimystik]

I see two people who within a couple of hours of standing around went completely off the rails. That's not very encouraging for O-Chul having to suffer over 100 days.

Durkon's breakdown was for the purposes of humor, and Miko was perfectly herself (meaning, she wasn't very sane to begin with - her solitude had no effect on that, supporting OR deleterious.) Neither speaks badly for O-Chul.


On the available evidence this seems to be a fairly typical feat for a 10th level character.

Perhaps, but all evidence also suggests 10th level characters are themselves atypical in this world. Kasumi and Daigo make good benchmarks and are around 6th - 8th level, as well as the majority of the Thieves' Guild.


What makes you think this? He tried to lie and failed.

Yet he called Redcloak on his faulty logic and succeeded.


The simple fact of being a paladin will do for starters. And being a fighter-paladin likely makes it worse. Of course his inability to catch Shojo doesn't speak well for him.

You're forgetting that Shojo easily fooled the comic's brainiest character: Roy. Outwitting O-Chul is a cakewalk by comparison.


And it is. But telling the beginner he is improving is a good way to keep him playing. So he tells him that to have somebody to play and thus avoid boredom.

His sotto voce comment ("very quickly") was not directed to the MitD, it was an aside to himself. I don't need to quote Shakespeare for you to realize how important asides are to a storyline - they betray a character's secretive line of thought.

I'm convinced he's planning something, but what?

[TheSummoner]

I still think he was simply happy to have a way to keep himself entertained. Just because paladins tend to be patient and stoic doesn't mean he wouldn't welcome a way to pass the time (he isn't Miko afterall...). One of the best ways to get through a difficult time is to focus on something else.

Atleast how I interpreted it, his comment about how quickly the monster was learning was directed towards the monster, and even if it wasn't, it could still be a simple expression of happiness that the monster is learning to play a game that can keep him entertained.

And whether or not my opponent is able to force a draw depends entirely on how well they understand Tic-Tac-Toe... its an incredibly simple game and the pattern for how to never lose is remarkably easy, but unless you can see it you don't stand a chance.

[Tass]

To give you an idea of how complex Go is, there are more possible ways for a game of Go to be played than there are atoms in the solar system. And for a pure strategy game with no luck factors whatsoever and finite boardspace, that's pretty damned complex.

Puh-lease, so there are for chess as well. And many more for go. :)

Visualizations of this kind is not very helpfull with numbers this big. Make it the number of elemental particles, in the entire vissible universe - thats just a few tens of orders of magnitude more(!) Go still has way more combinations.

Even elemental particles in the universe is only estimated at roughly 10^80. With an average game lasting 150 moves, and more than 100 possible moves each time you get 100^150 = 10^300, and that is not counting insanely long games with repeated big captures and play inside.

"Well 300 is less than 4 times 80, its not so much more :smallyuk:" Well remember that 10^81 would be the number of particles in ten universes, 10^82 would be the number of particles in a hundred universes. 10^160 is the number of particles if every elemental particle in our universe contained a universe including trillions of stars. And we are only halfway there, sorry I'll stop now :)

[Iranon]

There is no real need to bring up the relative depth and complexity of chess and go.

A good chess program will outdo any human player when it comes to tactical accuracy while remaining average at best when it comes to strategy.

There are several lines that are theoretically playable but will be avoided by humans because of the tactical complications - at least in competitive over-the-board play ('one weak move and I lose. If my opponent tries something I'm not familiar with, I will be in time trouble'). Computers will happily play those because they don't care how terrifying a position is, they only care if it's favourable or not.

At the same time, a few players have made it their speciality to beat chess computers. This usually involves very quiet play, building up long-term strategic advantages that are beyond even the best programs.

Since humans and computers excel at different aspects of the game, a human player assisted by a chess AI will beat the program he's using as well as human players considerably stronger than himself.

***

Since Go is a more pure strategy game, one would expect the computer to be worse at it. It's rare that tactics take over and the AI can force a sequence of moves where it's playing to its strengths.

[Nerdanel]

I think O-Chul is (in addition to just passing time) teaching :mitd: about the meaning of friendship so that it could recognize who is a true friend and who is not and thus someday potentially defect from the side of Evil.

[David Argall]

Durkon's breakdown was for the purposes of humor, and Miko was perfectly herself (meaning, she wasn't very sane to begin with - her solitude had no effect on that, supporting OR deleterious.) Neither speaks badly for O-Chul.
They are being called as character witnesses for O-Chul. So excuses for their behavior merely makes them worthless, and leaves us assuming O-Chul is really really bored, except when he is being tortured.


You're forgetting that Shojo easily fooled the comic's brainiest character: Roy. Outwitting O-Chul is a cakewalk by comparison.
Roy is quite brainy by fighter standards, but still is the mental inferior of V, and Haley does penetrate the trick. We can assume she has a good sense motive of course, but O-Chul is constantly around Shojo for years. That's a whole lot of chance for a roll of 1 and 20, not to mention circumstance bonuses.
We can't conclude from this that O-Chul is stupid, but we do get reason to doubt he can formulate some deep plan.



His sotto voce comment ("very quickly") was not directed to the MitD, it was an aside to himself.
Which here merely means he is telling the truth.

[Xerrik]

I wouldn't necessarily say V is smarter than Roy. Remember the incident with the Mind Flayer? Instead of going for V's "Delectable 18 intelligence right before you", it went after Roy's fighter brain. It saw him as a turkey dinner, while it only saw V's as a sandwich. That's saying something.

[TheSummoner]

The giant explained that as V is has higher intellegence, but Roy has a higher combined total of intellegence, wisdom, and charisma... a more balanced meal

[Poit-Narf]

I wouldn't necessarily say V is smarter than Roy. Remember the incident with the Mind Flayer? Instead of going for V's "Delectable 18 intelligence right before you", it went after Roy's fighter brain. It saw him as a turkey dinner, while it only saw V's as a sandwich. That's saying something.

Right. The sum of Roy's mental stats is greater than the sum of V's mental stats.

[Xerrik]

The giant explained that as V is has higher intellegence, but Roy has a higher combined total of intellegence, wisdom, and charisma... a more balanced meal

Sorry... I never saw that.

[Elfey]

Go has a spiritual connotation to many in Japan, but has had a history of philosophy tied to it that's very introspective. One person I read, forget where, said it's an enlightening game that teaches one to live in harmony, as better players don't need to kill to win.

Actually killing in Go becomes wasteful and stupid over time. There's more than a few Japanese that use Go with different Zen philosophies.

I think O'Chul is trying to teach :mitd: philosophy and wisdom that the MitD doesn't have to do Evil if he doesn't want to.