Monster-san... smarter than he looks?

[Ashadar]

I just noticed in this comic (http://www.giantitp.com/comics/oots0550.html) , on the last panel of the first row, O-chul tells Mitd that he is learning the game of Go very quickly. As a player of Go myself I can tell you that Go is a great deal harder than chess, and being able to play against someone who's played the game for a long time is incredibly hard and often ends up in disaster.

And yet, in this comic (http://www.giantitp.com/comics/oots0651.html) we can see the monster clearly holding his own against O-chul, and even tho he is losing the game, black can still do some damage to white, and take the whole left side, as well as extend towards his bottom group. That is quite amazing for someone who's just started playing, considering he's playing against someone who must be quite high level, with a great deal more experience.

So I suspect that the monster isn't stupid at all, he may have a higher int score than many characters in the oots cast, he is just too lazy to use his intelligence to his advantage, and has low wisdom(like a child). Now I'm not sure if this has been discussed before, but it looks like good information for the monster guessing thread, no?

[lothos]

Hi,
The level of MitD's intellect (i.e. his Intelligence Score in D&D terms) has been discussed quite extensively in at least two of the MitD threads. You can find a summary of previous discussions about MitD's intelligence in the first post of the current MitD thread here (http://www.giantitp.com/forums/showthread.php?t=189676). Have a look in Section 2b - Abilities, then expand the spoiler for "Knowledge". Grey Wolf has spent a lot of time and effort summarising the consensus over the last 165 pages of discussion.

If you are interested in MitD, please come and join the discussion in the main thread. It's a really good idea to get "up to speed" by reading the first post in the thread (the one linked to above). It's a great way to see what's been previously proposed and what ideas have already been debated. Of course there's nothing to prevent you suggesting something that's previously been debated and it's particularly worth doing so if there's new evidence to consider.

Cheers.

[King of Nowhere]

As a player of Go myself I can tell you that Go is a great deal harder than chess,

As a player of chess, i must stand up for it, and say that you robably find it so because you never tried to play chess at some high level, or never found someone competent at it. I'm not speacking against go, but there'0s so much complexity iun chess that not even the world champions fully understand it.
Anyway, back on topic.

Mitd is childish. Children can be smart. Children can especially be good at mmind games (in chess, there was a guy who became great master at age 13. And there are only some 500 great masters in the whole world).
The mind is much more complex than just intelligence, wisdom and charisma. I think mitd has potential, but never knew to use it because he's been trained not to think for his whole life.

@ lothos: he wasn't speculating on what the mitd is, so this don't belong to that thread.

[Conuly]

Yes, but computers can be taught to play chess at a reasonably advanced level. This is not the case for Go.

[Ancalagon]

I have a question: How does Monster-san look? :smalltongue:

[Klear]

Yes, but computers can be taught to play chess at a reasonably advanced level. This is not the case for Go.

...which doesn't mean that Go is harder. Would you say that speech comprehension is harder than chess? Yet computers are much better at chess.

Besides, harder can mean different things - i.e. difficulty of understanding rules, being able to play without making glaring mistakes and the difficulty of mastering the game. I've never played much go, but I think it's safe to assume that it's quite incomparable to chess.

In any case, MITD being able to grasp the rules quickly and playing interesting games with O'Chul definitely means he's got some potential. In fact, it's probably noted in the comic to tell us that much. I don't see any problem with that.

[NYCharlie212]

Being a Go player myself, I got to say Go is much harder than chess. It is pretty much widely accepted among master chess players as well that Go is the harder of the two. http://en.wikipedia.org/wiki/Go_(game)#Computers_and_Go

Back on topic: MiTD has displayed many clues that it is actually perhaps the most intelligent character on Team Evil (maybe next to RC). Besides being able to learn Go in a relatively short period of time (even grasping the concept of the twin "eyes"), it derived Tsukinko's spell as "half a spell" with only a glance. With a higher wisdom score I can see MiTD becoming a much smarter character.

[FujinAkari]

Yes, but computers can be taught to play chess at a reasonably advanced level. This is not the case for Go.

This seems incorrect (http://scienceblogs.com/evolutionblog/2009/03/computers_learn_to_play_go.php)

[Tilt]

Being a go player myself, I have to agree that go cannot reasonably be said to be "harder than chess". What does that even mean? Both are two player games -- they're as hard as your opponent is skilled. We're nowhere close to knowing perfect play or sure-win strategies for either.

Computers are much better at chess than at go, seen in relation to the best humans. This is partly because chess reduces better than go to things that computers are good at (tactical calculation), and partly because chess has been the preferred object of study for AI for as long as the field has existed. (There's also a widespread perception that computer go engines are weaker than they really are, as the field has been moving quickly recently. From what I'm aware the best go engines are now around amateur two dan, which I think many people would classify as "reasonably advanced".)

(Also, assuming that's the left edge of the board visible in the panel, I'm not sure we can reasonably assume that O-Chul is an experienced player, as long as we're over-analyzing this.)

[Nimrod's Son]

Being an Angry Birds player myself, I can say that it's probably easier than either Go or chess.

[Klear]

I guess the best way to settle this would be to cut a chessboard and a go borad in half, glue two halves together and let a chess grandmaster play with chess pieces and the go grandmaster place his go stones and see who wins.

Now that I think of it, it could be fun. The go player would have to surround the king piece, which is impossible (AFAIK) as long as the chess player doesn't move it from the edge, while the chess player would go on a rampage removing his stones while trying not to stay surrounded. The chess player would definitely have the advantage of easier captures, but on the other hand the go player's pieces are unlimited and would eventually probably swarm the chess player.

So - correct me if I'm wrong - chess player gets points for every go piece captured, but these points are only used when the whole board gets filled and the game thus ends. The go player can win by capturing the king, but he first has to force the king to move from the edge (by, for example, denying the chess player any other move).

Wow... I'm starting to thing that such a game may not be as stupid idea as I originally thought.

[Zejety]

I guess the best way to settle this would be to cut a chessboard and a go borad in half, glue two halves together and let a chess grandmaster play with chess pieces and the go grandmaster place his go stones and see who wins.

Now that I think of it, it could be fun. The go player would have to surround the king piece, which is impossible (AFAIK) as long as the chess player doesn't move it from the edge, while the chess player would go on a rampage removing his stones while trying not to stay surrounded. The chess player would definitely have the advantage of easier captures, but on the other hand the go player's pieces are unlimited and would eventually probably swarm the chess player.

So - correct me if I'm wrong - chess player gets points for every go piece captured, but these points are only used when the whole board gets filled and the game thus ends. The go player can win by capturing the king, but he first has to force the king to move from the edge (by, for example, denying the chess player any other move).

Wow... I'm starting to thing that such a game may not be as stupid idea as I originally thought.

http://fc01.deviantart.net/fs71/i/2010/001/8/e/Checkmate_by_zejety.jpg

[Ashadar]

Being a go player myself.. oh wait I said that already. Ok, the reason I said go is more complicated than chess is basically because with a 19x19 board it's very hard to have two identical games. For this reason it takes much, much, much longer to become a 9 pro dan player (the best kind) than it would for chess. I've seen a lot of chess players that just memorize like 30 opening moves, and know all the possible responses to them. The board is smaller, the pieces are limited, and so the game is easier to master. I am not speaking from personal experience with weak chess players or whatever, I was just comparing the time it takes to become a master of each of them. Plus, didn't Kasparov get beaten by a computer?

On topic: Boy... that Mitd is really smart am I right?

[Skavensrule]

I have played both Go and Chess but have not played them against an opponent that was any good for almost twenty years now. I would say that the rules for Go are easier to understand for a beginer but that the game itself is harder to master than Chess. In Chess you have fewer options because of the nature of the different pieces. Once you learn to think ahead you can quickly get to a decent skill level. What little I remember about Go is that trying to predict your opponents moves in advance was much more difficult.

[weeping eagle]

Would you say that speech comprehension is harder than chess?Yes, yes, a thousand times, yes. Chess is definable by a tiny set of rules, whereas speech comprehension requires a good understanding of vocabulary, syntax, context, non-verbal communication, shared culture and experience, etc.


Yet computers are much better at chess.Exactly. Chess is easier.


I've never played much go, but I think it's safe to assume that it's quite incomparable to chess.No, it isn't safe to assume that, at least for you. (To be honest, I *might* agree that chess and go don't compare, but probably not the way you mean.)

For the record, I play both chess and go, and I think that go's vastly broader combinatorial explosion, coupled with the difficulty of rigorously defining even whether a group of stones is alive or dead, give it a much longer learning curve than chess - blows it out of the water, in fact.

I'll go ahead and claim that activities with longer learning curves are "harder". So there.

[gooddragon1]

Being an Angry Birds player myself, I can say that it's probably easier than either Go or chess.

I disagree, I played angry birds and chess (not go) and angry birds seems so complicated to me. Jk.

[Whiffet]

... yes, I'd say the MitD more intelligent than he seems. Kinda like the absentminded professor in stories, except he's not a professor. So he's not that much like the absentminded professor... well, you get what I'm trying to say, right?

Huh, this made a lot more sense in my head.

[Klear]

Yes, yes, a thousand times, yes. Chess is definable by a tiny set of rules, whereas speech comprehension requires a good understanding of vocabulary, syntax, context, non-verbal communication, shared culture and experience, etc.

Yet anybody can recognise speech without thinking and you'd have a hard time finding somebody who could claim higher speech recognition ability than you.

The word you need is complex, not hard. It's possible to have something very complex, but once you learn it, it becomes extremely easy.

Mind you that I'm not using this to prove anything about chess/go, since like I said, I never plaed go much. I just want to show that the arguments of computer programs and board size are somewhat flawed.

[Incom]

SSB is harder than Go, SO THERE ;)

Srsly, as I understand it Go is harder, but I've never played it and I love Chess to death sooooooooo... *goes to find some online Go and learn it*

I wouldn't be surprised if MITD gets a throwaway Chess gag after this thread.

[Knaight]

Yet anybody can recognise speech without thinking and you'd have a hard time finding somebody who could claim higher speech recognition ability than you.

Find one person who has any where near as many hours of practice in chess as in speech recognition. Consider the whole 10,000 hours to become a master adage. At age 10, mastery of speech comprehension would require only 1,000 hours a year. Which is less than 20 hours a week. Which is less than 3 hours a day. This is 2 hours talking to someone, hearing live speech, hearing recorded speech, listening to music, so on and so forth. Good luck not getting that much.

[Klear]

Find one person who has any where near as many hours of practice in chess as in speech recognition. Consider the whole 10,000 hours to become a master adage. At age 10, mastery of speech comprehension would require only 1,000 hours a year. Which is less than 20 hours a week. Which is less than 3 hours a day. This is 2 hours talking to someone, hearing live speech, hearing recorded speech, listening to music, so on and so forth. Good luck not getting that much.

Yeah. It's complex and hard to learn. But once you learn it, it's not hard at all. Chess will always present interesting challenges, no matter how long you train.

[Toper]

This seems incorrect (http://scienceblogs.com/evolutionblog/2009/03/computers_learn_to_play_go.php)
Science journalism is almost always sensationalist and unreliable, sadly. The computers were given several handicap stones. The Wired article just barely alludes to this when it mentions a "head start". Go programs are far better than in the past, but still much worse relative to human players than chess programs.

"Difficulty", obviously, is in the eye of the beholder.

[Jay R]

It is generally accepted by mathematicians who study game theory that Go is more complex than chess. The complexity of six different types of pieces in chess is real, but not as deep as the extra complexities of Go. There are 361 positions rather than 64, there are over a hundred pieces on each side instead of 16, a game takes hundreds of moves rather than around 50-80, and the play of a single piece can affect hundreds of pieces at once.

Jay R., Ph.D. (Operations Research, including game theory)

[Dr.Epic]

I just noticed in this comic (http://www.giantitp.com/comics/oots0550.html) , on the last panel of the first row, O-chul tells Mitd that he is learning the game of Go very quickly. As a player of Go myself I can tell you that Go is a great deal harder than chess, and being able to play against someone who's played the game for a long time is incredibly hard and often ends up in disaster.

The game they're playing is called Go? Then I guess that makes him...

Monster-in-the-Darkness A-Go-Go:smallbiggrin:

[Conuly]

At age 10, mastery of speech comprehension would require only 1,000 hours a year.

Except that at age 10 you're starting to bump into the end of the very real critical period for language learning. You know how people criticize American schools for starting second languages so late? At least you can LEARN a second language if you start it in your double digits. At that point it is incredibly difficult (and getting to be impossible) to learn a *first* language. If you, for some reason, have had little to no exposure to language already, it's going to take a lot of effort on everybody's parts - and fast, because in a few years it'll be too late and you'll be well and truly screwed.

What's really sad is that children will try to communicate even if they have no language (that's how signed languages and creoles develop, groups of children with only a few words (home sign and pidgin, respectively) start multiplying their vocabulary and spontaneously creating grammar if you put them in school or town together), and you have to work HARD to isolate them that much. But it has happened.

[Matamane]

I'm a chess player, and have dabbled in go, so correct me if I'm wrong, but it looks like MiTD is winning.

[Cloaked Bloke]

I'm a chess player, and have dabbled in go, so correct me if I'm wrong, but it looks like MiTD is winning.

Not so sure about that. (http://www.giantitp.com/comics/oots0651.html)

Going off of O-Chul's dialogue, it seems that MitD is losing because he's allowed himself to be flanked like that. He only has enough room to not immediately lose, I suppose.

[Stille_Nacht]

Yes, but computers can be taught to play chess at a reasonably advanced level. This is not the case for Go.

there are many go computers that can play at an advanced level. the play is more conducive to programming actually.

i would say that go is more difficult to become competent at, but chess is a bit more complex at an extremely high level of play.


back to this, yeah i am reasonably sure that he is very intelligent

[rewinn]

Except that at age 10 you're starting to bump into the end of the very real critical period for language learning. You know how people criticize American schools for starting second languages so late? At least you can LEARN a second language if you start it in your double digits. At that point it is incredibly difficult (and getting to be impossible) to learn a *first* language. If you, for some reason, have had little to no exposure to language already, it's going to take a lot of effort on everybody's parts - and fast, because in a few years it'll be too late and you'll be well and truly screwed....
This is a key point in comparing chess to speech comprehension. Human beings evolved to be speech-comprehension machines. We have a whole lot of specialized circuitry (...speaking loosely...) for just that purpose.
OTOH the circuitry specialized for chess-playing isn't there. So which "is" harder? It depends what you mean by "is". In some cosmic, absolute sense, speech recognition is vastly more complicated: chess involves the manipulation of less than 100 symbols according to very simple rules; it's difficulty only because it's such an artiface.

Now go, in my slight experience, takes better advantage of humanity's innate esthetic sense; we can to some extent evaluate positions according to their beauty in a way that doesn't work as well for chess. But that is probably a very small advantage compared to its greater number of potential moves.

Finally, as to MitD, while I'm sure it's a lot smarter than its lack of experience/education would otherwise indicate, still some of its success in the go game shown may have been with the help of handicap stones. One of the many delightful features of the game is how well it handicaps by the simple preplacement of an appropriate number of black stones; the result is a very playable game. In contrast, if you add or take away chess pieces to handicap a game, it changes the game more radically.

BTW this is a delightful thread. I wonder if we could get MitD to play the OOTS game?

[Klear]

This is a key point in comparing chess to speech comprehension. Human beings evolved to be speech-comprehension machines. We have a whole lot of specialized circuitry (...speaking loosely...) for just that purpose.
OTOH the circuitry specialized for chess-playing isn't there. So which "is" harder? It depends what you mean by "is". In some cosmic, absolute sense, speech recognition is vastly more complicated: chess involves the manipulation of less than 100 symbols according to very simple rules; it's difficulty only because it's such an artiface.

Now go, in my slight experience, takes better advantage of humanity's innate esthetic sense; we can to some extent evaluate positions according to their beauty in a way that doesn't work as well for chess. But that is probably a very small advantage compared to its greater number of potential moves.

Finally, as to MitD, while I'm sure it's a lot smarter than its lack of experience/education would otherwise indicate, still some of its success in the go game shown may have been with the help of handicap stones. One of the many delightful features of the game is how well it handicaps by the simple preplacement of an appropriate number of black stones; the result is a very playable game. In contrast, if you add or take away chess pieces to handicap a game, it changes the game more radically.

BTW this is a delightful thread. I wonder if we could get MitD to play the OOTS game?

Thanks for summing up my arguments.

[Ashadar]

I'm a chess player, and have dabbled in go, so correct me if I'm wrong, but it looks like MiTD is winning.

Mitd can get a little more territory by playing a piece at the 6th row and 2nd column. I can't really see some of the right side of the board but I think it's safe to assume that besides the left side and that huge group with 2 eyes, O chul pretty much has the rest of the board all to himself. So no, O-chul seems to be winning this one. Also, if this were a handicap game, it can't be more than 3 handicap stones for monster-san because there's no black stone in the bottom left corner, and if it were more than 3 handicap, the corner stones would have to be put first.

Now on off topic: Chess is much more limited than go, I don't know why people don't see this. In chess you will have the exact same openings over and over again because you are forced to it due to the initial placement of the pieces. You will have to move the pawns out of the way first, in order to move any other piece than the horses. also, any move you make only affects a board of 64 squares. In go each and every move you make has incredible ramifications over 360 other spots. Sure, the number of spots you affect decreases as you play the game, but the same happens in chess, because by the end of the game there will be fewer and fewer pieces left on the board.

I'm not saying chess is easy to play, but I think there's a reason why people have to train all day, every day to become professionals, and only 3 of them make it each year.

[Whiffet]

People don't "see it" the way you do because whether something is easy or hard is relative. Some people will see Go as harder, and others will see chess as harder. Neither group is wrong.

As for the MitD, it would be fun to watch him play some more Go and win against a goblin.

[Alias]

I always looked at the game in that panel and say a picture of CiTD. One doesn't normally fill up their territory like that. Kudos to Rich for going to the trouble of making the board position look reasonable, but I don't think the game in progress was the point.

As to chess vs. go, I'm reasonably competent in both. Though hinted at in posts before the essential character of the two games hasn't been stated out in the thread so I'll state it. Go is cumulative. Chess is reductive. That is, during the progress of a chess game the number of possible moves is ultimately reduced to 0 - indeed that's the object of the game - to force the opponent into a position from which the king cannot escape capture. In go there are a huge number of possible moves from the start and the position does not reduce during play - instead it becomes more complex.

I am a computer programmer by trade so I can give some insight into the problems programmer face with both games. With chess you can have a reasonably powerful program just by programming the accepted book lines in the openings and studies of the most common endgame positions.

Indeed, when I teach chess to children I start with endgame positions and work my way backward, only introduction openings and the starting positions after many months of practice with endgames. My students often best stronger opponents as a result provided they can get to an endgame, since that's where the majority of my teaching has been.

While go has opening formula, they aren't forcing. Further more than one of them can be used in concert. To further cause problems for a programmer, there are few moves in go that are truly *forced* whereas in chess forced combinations are quite common. The lack of ability to truly steer the opponent a given direction makes the game difficult to program.

Chess is different than go. I don't think either game is necessarily easier. I will say that chess is demonstrably easier to program for because computers will always fair better against a problem set that reduces in complexity over time.