Kasparov vs. The World: It's all different

Faber@FICS writes "I just checked how the match Kasparov vs. The World has been covered here. Today, more than fifty moves into the game, it is interesting that nearly everything that was said about this has been shown to be wrong. (1) This was no easy win for GK at all -- quality chess at its best, with very good drawing chances for the world after fifty moves. (2) Computers, although heavily used all over the net, did not play a significant role in this game so far. (3) This is no longer Windows-only -- Microsoft removed that requirement rather early in the game without comment. Surprise, surprise... Read a very insightful interview with Irina Krush, one of the official expert analysts, and check out the World Team Strategy BBS, where the world is at work. "

Comments

Democracy

[Kinthelt]

It was still kind of a one-sided match. I mean, the way the World chose moves might have been the best way to get alot of people involved, but it did not produce the best moves.

I would warrant a guess that the majority of people who play chess are average players. And almost everybody is a worse player than Kasparov (there might be some undiscovered talent out there). So, the grand majority get to choose moves that will ultimately lose, while the few genius moves will be out-voted.

This didn't happen as much as it would have since there were advisors on the World's team. And people never voted for a move that was against the advisors. So this wasn't really a match against the world, but against a few selected chess players. The voting mechanism was just to formalize everything and make people think their ideas counted.

Re:Democracy

[Possum Man]

It is true that advisors were essential in this game. In fact, the world basically followed Irena Krush's game (I'm not sure whether there's a single move against Krush's advice). However, it is important to note that Krush paid a lot of attention to the players' suggestions. She posted lines of play on her web site developed by a variety of different players on the internet. She constantly upgraded her opinion based on what different players suggested. She herself freely admits that she could never have played a game of this strength without her huge support base.

I would argue this method reflects the development of the Linux kernel (and other free software). The Linux kernel has an obvious leader. This leader would not have gotten very far if not for the respect and support of various people putting in almost as much as him. These giants in turn would not have gotten very far if they hadn't had the support of a much larger group of somewhat influential designers. These people in turn were supported by a much larger group of slightly influential people.

Originally I thought that the democratic element of this chess game was unnecessary. Why not just put one person in charge, and let the world advise them. But, as in the development of the Linux kernel, it is necessary that the people get to choose their leader. The other three advisors did not pay a lot of attention to the world-team players, and therefore were not listened to nearly as carefully as Krush was. The democratic element allowed the world to choose a leader who would represent them.

Re:Democracy

[MindStalker]

So, what your saying is that if we could ever get our representatives to listen to us, and vote on thier ultimate suggestion, we could have a kick ass government. I like it. As it seems we could never get our (my) government to do something like this, (they only pretend to listen). Who out there is all about creating an island of oil tankers and see how far the slashdot effect will take us :)

Distributed Computing vs the humans

[Possum Man]

The next interesting experiment to try is distributed computing vs humanity. Clearly computing power can defeat any single player, as was evidenced in the Kasperov vs Deep Blue game. And this new Kasperov vs the world game has shown the power of collective human chess playing.

A distributed set of PC's throughout the world should be able to create a player easily capable of defeating Deep Blue. Wouldn't it be a great match to witness (and play): the human world vs. the PC world?

The One Bonus Humanity Has

[Skyshadow]

I see a lot of posts which point out that the world can't beat Kasparov 'cause they're, taken together, the most average chess player possible.

While this may be true, they have one very major advantage: They must be very difficult to predict, which is what these chess types count on to decide their next moves.

That plus they have a couple of grand masters helping them out.

----

Re:The One Bonus Humanity Has

[XDG]

While this may be true, they have one very major advantage: They must be very difficult to predict, which is what these chess types count on to decide their next moves.

This is, I believe, a misconception about the game of chess, that shows why human/computer matches are often misunderstood.

For humans, chess at the highest levels is not played by considering each move, predicting a response, and continuing through some search depth. Instead, chess masters always assume that the opponent will find the best response. To the extent that the response to a move must be "predicted" -- the master must find the one, single best move in response. This is why the "predictiveness" of the world is irrelevant -- the combination of the grandmaster coaches and collective voting push the game towards correct responses.

The real question is how to develop a strategy in the game that leads to small advantages. Each move, particularly early on, has effects on various factors: space, material, development, etc. This requires the ability to understand the balances and imbalances inherent in the board at any particular snapshot.

That is the difficult part of programming the computer. Searching out combinations is just a matter of power. The trick is programming the computer to analyze a snapshot of the board in order to determine the relative balance. Then the computer can use that information to examine in further depth the stronger lines.

If the programmers do a poor job of programming the analysis of the snapshots, a human player can exploit that. In several of the Kasparov vs computer matches, Kasparov's first few games were all about him examining how the computer was programmed to think, then exploiting that in the later games. For example, if the computer over-valued material advantage, Kasparov could use that knowledge to "convince" the computer to take a sacrifice that on the surface looked to offer a material advantage but that would lead to a positional imbalance that would favor Kasparov later.

(This is one reason that Kasparov is annoyed that the programmers never released information on how they programmed the computers.)

Further, the earlier comment about the effect of grandmaster coaches was dead on. One of the things that made the game so interesting to chess afficionados is that Krush pulled out a move she had been preparing for some time and "discovered" a new line in the old opening. (Look at Kasparov's comments on this.) If you look at the history of the analysts commentary, the voting became a personality contest between the analysts taking quieter approaches to those taking active approaches.

As a final note, another comment in the history is that Kasparov mentioned at one point that he was doing something less that ideal because didn't want to give away certain preparation he's making for his world championship match against Anand.

My overall take:
Is GK vs The World interesting? Yes.
Does it "mean something"? Not particularly.

XDG

What does this undertaking show?

[jabber]

Well, besides the exercise of getting so many voices contributing to a strategic decision, a'la online voting?

This is a way for Kasparov to get his 'face' back, by defeating the collective chess expertise of the world single-handedly.

What I would like to see next is the world vs. Deep Thought. If the greatest chess mind in the world, capable of defeating the whole world, was himself defeated by a computer, does transitivity apply? Can the world be beaten by the computer? What would be the result on the human psyche, to be defeated by a machine? Would governments halt AI research funding out of fear as thoughts of the W.O.P.R. and SkyNet dance in their heads?

Yes, we all know that DT was coached, and in fact designed, specifically to defeat Kasparov. It was programmed with Kasparov's strategies and game history... But still, it makes one wonder how the collective ego of humanity would respond to having it's collective hinny wipped by it's own invention.

Altered Games, or Thinking vs. Memorizing

[wass]

I don't really like these chess matches because it judges who has mastered this particular game more, and who has a better feel for how the pieces move and different strategies. It does not judge who is better at abstract thought and general strategi thinking. IMHO, of course.

I have a different idea of a wits game, and I've talked with some serious chess players who say this would be a very interesting test. What you do is to have a chess match like this, but change the rules, maybe even only slightly. This would more adequately match wits against wits. For example, let's say you change the moves, such that knights cannot jump over other pieces, or that the queen can only extend moves up to a range of 4 squares, or that you can teleport from the left side of the board to the right, a la pacman. You get the idea. It would drastically change the strategies that many chess players have already memorized.

Even more interesting would be a human vs. computer match like this. Say you give each party maybe a few days or 1 week to prepare for the game. The human tries to get a feel for the game, the computer guys re-program the computer's code during this time. I bet you'd find in this case that the computer demolish the humans. Kasparov has had 30 someodd years to memorize all these strategies and moves, but given only 1 week I think he'd be just as much a newbie in a changed game as I currently am at normal chess.

Has anybody ever heard of competitions like this, or do you think people would be willing to try them? I'd be very curious to know how chess masters compare to newbies at the altered games.

Re:Altered Games, or Thinking vs. Memorizing

[Shanoyu]

"Who has mastered this game more", Ugh, this is like holding up a big sign that says "bring weapons, i'm a troll" to chess players. =P

First of all, it's been proven thru tests of all sorts that high level chess players are intellectually very strong. For existance the Grandmaster and World Champion Emanuel Lasker was a good friend of Einstein and talked with him on par about Einsteins theories and all kinds of subjects.

On the flip side of the coin, Chess players can be extremely autistic and/or not so socially responsible, For instance Bobby Fisher today is insane and also an Anti-Semetic who claims that the Jews are out to get him, (You can hear his radio interviews at chess-space.com , I think.)

You don't Memorise games, you memorise lines, admittedly for the opening lines this is a pre-requisite to do well in chess, but this hasn't stopped people from developing new openings like the Pirc Modern and the Dragon.

However memorising these opening lines is pretty easy, and if you are going to be playing on a tournament level, studying the major lines, (Indians, Kings pawn, etc.) isn't really that much of a pain. More important in chess are concepts and general strategies.

Chess is a studious game, and people who study chess play better, and on the higher levels of chess a large vocabulary of opening lines is required, but believe me, games would most likely have the same result given these 'altered rules', I could mention bughouse and myriad other variants but I wont, needless to say, it has little effect as long as the general strategic concepts are the same, and in chess they are sweeping.

"Good god man, you're addicted to chess and now you are going to take up bridge? Once you get the hang of bridge you'll be useless for any worthwhile occupation!" - Emanuel Lasker


-[ World domination - rains.net ]-

Re:Democracy / optimization / search

[Oestergaard]

Deep though / deep blue consisten of a very large number of very bad chess players (they only knew the rules and how good/bad one move would be). But the were organized so that a good number of moves could be considered (including possible moves from Kasparov - by guessing), and the best move picked.

This is not democracy. The move with the most backing (the move to be chosen) was not the one which most of the ``chess-players'' wanted, but the one that looked best considering the whole set of moves and possible moves.

You can't find an optimal move by distributing the search for single moves to dumb players, and then voting in a democratic way. Maybe this could have worked if the human voters' moves had been considered by a computer (or a human with too much time on his hands). But then again, studpid computer chess-players can produce a much larger number of moves than smarter human ones.

Stupidity parallelizes well, and that can be exploited if it is just reckoned that it _is_ stupidity. Parallelizing intelligence is different. A large number of intelligent thoughts cannot be parallelized only by considering their outcome. You must consider the reasoning behind them too. I'll leave that as an exercise to the reader :)

Re:Proverb

[SteveX]

How can you say this?

The world may score a draw with Kasparov. This is no easy achievement - that's better than almost every computer (every computer?) has been able to do, and it's completely against what was predicted of this game (by the people here on Slashdot at least).

"A consensus of chess intermediates will suck compared to a single chess expert", a quote from the earlier Slashdot thread on this.

I've been hoping from the start that this would work out - and maybe someone should look into whether the combined efforts of so many people contributed to a larger strategy that none of us could have created alone. Isn't that how the mind works? Isn't this what so many Sci-Fi books have pontificated on?

- Steve

Is it possible to Slashdot effect the moves ... :)

[dustpuppy]

Is the voting for the moves checked (and vetoed) by humans?

Cos if it isn't, I could see an amusing result if the /. effect was rolled out over the voting form and we had a solitary pawn on the far left of the chess board advancing one square at a time completely ignoring any move that Kasporov made ... ;)

Why bother with Chess?

[Anonymous Coward]

There still isn't a Go program written that can defeat a human master. It's a shame it isn't more popular in this country, I think it would be more interesting if IBM and such put effort into it.

Change the game...

[jedrek]

I think that you're overstating the 'power' of DT. Humanity won't be 'obliterated' by loosing to a computer at chess. Chess is a very finite game and the number of positions can in fact be calculated, stored and referenced. In fact, I think it's time to take chess down as the 'game of games' - especially when competing against a computer. As computing power grows, chess goes the way of tic-tac-toe.

Re:Proverb

[mischief]

Too many cooks spoil the broth is a proverb because they all fight over what's going in the actual "broth"; with an organised system lots of cooks would theoretically make the perfect soup. Because there's an voting system for choosing the next move, the world should hopefully be able to filter out all the bad moves, and be able to at least draw with Kasparov (which it seems might be a possibility now).

--

Go, not Chess

[Anonymous Coward]

Go is, in computational terms, much, much more complex than chess. This is why it is an interesting challenge; conventional minimaxing is so computationally intensive as to just not work well.

There are many strategies used by go programs including analysis of local situations, conventional positions and heuristics based on conventional wisdom, all of which have shown some promise.

I think the chief interest of Go in this context is that it is sufficiently complex in all its manifestations that it has reached the point where alternative approaches (including pattern matching, fuzzy logic and other so-called AI techniques) might bear more fruit. I would really, really, really like to see where this goes. I have a gutfeeling that if we can write a champion Go program, we will learn an immense amount from it.

Democracy, with all the pitfalls

[Enoch Root]

If you look closely at how this works, this isn't truly Kasparov vs. the world. It's Kasparov vs. four chess masters, with a randomising factor thrown in to determine whose move is selected.

If you truly did a Kasparov vs. the world, you'd just invite everyone to vote for the move they want, without any advice. Then, indeed, you would have Kasparov vs. the world, and the world would be a very, very average player indeed.

Rather, this is more like democracy. Four masters "suggest" moves, which means the clearer (or cryptic and brilliant-sounding) their analysis is, the more likely they are to sway the majority in their favour. Oh, sure, you can vote for another move that isn't suggested, but when the average player has a chance of going with one master whose opinion he believes he shares, or thinking up his own move that none of the four masters thought up, what do you think he'll do? And even if he does, will there be enough votes from the other players? No.

So the little genius sitting at home and ready to beat Kasparov is not going to weight much in the balance. He doesn't have access to a visible, publicised advice posting like the other chess masters have. He could post on a BBoard, but what if he really sucks at English and cannot communicate his analysis properly?

So, what is left, is the ability to sway the crowd. Chess ability matters, but when the average player can't figure the game out one move ahead, he's gonna go for the more convincing, not necessarely the better one. And so, we're back to good old democracy: don't elect the one you know to be the best qualified, because you yourself are not qualified to make that call. Just elect the candidate that sounds the most convincing and seems to know where he's going. And he'll take care of the rest.

Interestingly enough, this is the only way the world can win (or draw, by looking at the game) against Kasparov. If there was no expert vote-swaying, it would be anarchy. Now it's democracy.

This game is a statement on human politics!

"There is no surer way to ruin a good discussion than to contaminate it with the facts."

Re:Democracy, with all the pitfalls

[0xdeadbeef]

So how do you know it isn't the "statement" itself and not the presentation that sways that catageory of voters?

It's the same old excuse. When the vote goes my way, it's due to enlightened leadership and an educated citizenry. When goes the other way, it's because politicians are manipulative and the masses are dolts.

In reality, both situations are somewhat true. No one is an expert on everything and no one is the definative expert on anything. But everyone has the capacity to think critically, and learn, and to judge the credibility of those who claim to be experts.

Chance at a draw == none.

[foodini]

ERGH! Damn browser.

Anyway, The world blew it when they advanced the
pawn to force the white rook to take. Had the
other pawn been pushed, the world would now have
GK in check and have tempo. As it is, GK will
soon have another queen and we're dead. Game
over.

I am farily convinced that GK just wasn't taking
this game seriously early on. Everyone just has
their hearts so committed to patting themselves
on the back that they fail to see this.

Don't overlook, either that this game was 5 advisors against GK, everyone just voted for the advisor that they believed the most.

The worth of Human Communication.

[Thomas Charron]

If anything, this contest was a good experiment in what the internet can provide. It brought EVERYONE who submitted moves together to determine what would be a good move to make. This would simply not have been possible without the internet.

It's not giant Beawulf clusters that did it, it was a giant parallel processing human network that did it..