#1  [119164] |
| inehou-pxy06.compaq.com |
SWATCH BABY!!!
Fox |
| #2 Kili[119165] |
| ua180d22hel.dial.kolumbus.fi |
| 1st non-author! |
| #3 plus2longsword[119166] |
| asc8-grr-ip005.wmis.net |
| This was really cool. I tried Go, but the Chess player in me always takes over. I never win. About the only thing Chess and Go have in common is calculation. In Chess the most important part of the board is the center. Besides this, another big difference is how computers react to these games. Computers are just starting to be able to compete with the best humans in Chess. Rebel Century played a matchwith the Dutch number one, Loek Van Wely, and made a draw. I guess computers arn't to good at Go. This seems strange to me, because I've heard Go is mostly tactical, where Chess is a mix of strategy and tactics. And tactics is the reason comps play chess well! |
| #4 GoZuLu[119167] |
| 195.249.56.3 |
It sucked. Sorry for the non-constructive comment.
GoZuLu |
| #5 Casper...[119168] |
| spc-isp-tor-58-18-437.sprint.ca |
i tried go, couldn't get into it.
chess is just so much more rewarding imo.
not that i play much chess. |
| #6 Furious[119170] |
| bal-ts2-2500.tpgi.com.au |
| Single digets on a fox reports = me > u |
| #7 WaxAngel[119171] |
| 61.74.168.83 |
| very hard to follow |
| #8 Redemption6[119172] |
| cache-1.sbo.ma.webcache.rcn.net |
| errrr surprisignly..not so good. I found it hard to follow and boring as hell :) Anyways 5-6 but gold? |
| #9 titans-thesoby[119173] |
| station4.govwentworth.k12.nh.us |
i need more action...
6 |
| #10 apos'trophe[119175] |
| adsl-63-200-53-2.dsl.snfc21.pacbell.net |
| wtf r u ppl talking about?!? This isn't supposed to be an exciting StarCraft like br. Its a fuckin board game! |
| #11 plus2longsword[119176] |
| asc8-grr-ip052.wmis.net |
| yeah, you guys should stick to SC if this couldn't hold your attention. In fact, I found it easier to follow than most of the the SC battlereports here. |
| #12 'Final`Flare'[119177] |
| spider-wp072.proxy.aol.com |
| I personally liked this. A bit hard to follow at times, but I usually knew what was going on. Although I couldn't figure out how to navigate the IGS page :/ |
| #13 DarkRider[119182] |
| 198.134.2.2 |
| lol@ #11 |
| #14 Psychochild[119185] |
| 247.orlando-23-24rs.fl.dial-access.att.net |
This is obviously a report not aimed to those with tiny attentions spans, I.E. me. Iwould love a chess BR or a Mouse Trap BR (Mouse Trap kicks ass! J/K). Since I never enjoyed Go, I couldn't get into it. 4-6 for me. But thats why I'm not a rater.
Til next time,
Psychochild. |
| #15 Amoeba_Boy[119187] |
| spider-tl051.proxy.aol.com |
I dont think Id be real good at Go, I cant see things really far in advance. But a nice report
It was a little hard to follow a few times, like at the last part when you started talking about M16 and those other things but the pics helped out A LOT. Kinda lacked humor |
| #16 Kane[119188] |
| 12-233-180-134.client.attbi.com |
| Hard to follow. Maybe use animated GIFs next time? |
#17  [119189] |
| St-Catherines-ppp112530.sympatico.ca |
| kinky. I liked it. |
| #18 AlferdPacker[119190] |
| dhcp078246.res-hall.northwestern.edu |
The animated .gifs sound like a good idea.
My one complaint with the report is that with Go, you could just post the kifu to show what happened in the game. To write a BR about Go requires something ot make it worth while.
Your comments were great for someone who didn't know much about the game, but for someone who knows how to play even fairly well, its not terribly interesting except as something new on BR.com |
| #19 IcE-Inforcer[119191] |
| 209.105.210.34 |
| Woa... I clicked on the "explain how go works" link and I still don't get how it worx. Sorry Fox. |
| #20 ZerG~LinG[119193] |
| p434-089.ppp.get2net.dk |
Its nicely written and all, but GO cant really rock my boat :(
Chess is (imo) better, and makes for more interesting games (for the viewer atleast).
|
| #21 [119197] |
| 66.113.28.2 |
thanks for the comments. I wanted to try something new. Animated gifs sound good. If you ask me I find Go to be much more interesting then Chess hehe. I find chess to be so restricted. When I hear that at the top level chess players usually tie, then in Go at the top level players will resign when they are being crushed. But next game will go all the way to the end of the game and win by 1 point. IT's so different the way the boards look. Chess just seems like it's almost at the point where it's been figured out. I dunno when computer in chess can beat anyone but can't beat even the lowest ranked pro in Go. I think that's something special.
Thanks for the feedback!
Fox |
| #22 DarkRider[119202] |
| 198.134.2.2 |
lol@top chess players usually tie.
Statemates are not the norm.
One player using wins
FYI |
| #23 silverleaftree[119203] |
| PA207.ias.edu |
i'm decent at GO actually if anybody wants to play over yahoo...IM me (silverdaemond)
on topic: you need better players- and finish the game! |
| #24 TIFKONIKA[119204] |
| CPE0080c6f147d4.cpe.net.cable.rogers.com |
| THIS GOT A 7?!?! HOW IS THAT POSSIBLE? |
| #25 ADRENALING[119210] |
| CPE014100005949.cpe.net.cable.rogers.com |
| FOXY IS POSTED AGAIN? YAY! =] |
| #26 Erad[119213] |
| pa-millscreek2a-368.buf.adelphia.net |
| i guess ill have to post again |
| #27 Doom-[119214] |
| mke-24-167-248-237.wi.rr.com |
| Going out is fun. |
| #28 Kane[119215] |
| 12-233-180-134.client.attbi.com |
Fox: As has been pointed out before, a computer's skill is no measure of a strategy game's complexity. I don't think Chess and Go can be compared. Of course, I've never played Go, but there are obvious differences.
Fox, tich me go? |
| #29 PsySal[119219] |
| h24-81-160-165.cg.shawcable.net |
| That seems WAY too early to have resigned. Why not play the whole board? If you can make a new fight somewhere else then you have a chance of connecting with your old pieces and saving them; hopefully you can tie your opponent up by threatening his pieces somewhere else on the board. Well, i suck pretty bad at Go anyways, but this is usually what i do when i start to get raped. Maybe Dref resigned coz he realized he was hopelessly outclassed? Or, maybe I suck... Hrmmmm |
| #30 Mistears[119222] |
| cpe-24-31-39-161.socal.rr.com |
| i thought this report was an extremely good shot at something new. the only problem is, Go is such a complicated game that it is very easy to lose the reader's interest when trying to explain the thought. a good shot at reporting one of the most complicated games in the world |
| #31 plus2longsword[119224] |
| asc3-ip51-grr.wmis.net |
| I think I can offer a little clearer explanation of how Chess and Go relate to computers. As far as complexity is concerned, it is up for debate, but Chess is far from being figured out. I doubt it could ever be. I dont know about Go computers, but Chess computers can only work in one way. Basicaly a series of algorithms, with different conditions, like never play a6, g5, is one bad one I saw. To actually figure a game out to the end is hard to fathom. I dont know the statistics exactly, but in the first seven moves of Chess there are like 500,000 possibilities. So were probably waiting for the quantam computer to figure that one out. Go has a bigger board, therefore even more possibilities, thus a bigger horizon, which makes it even more difficult for programmers. |
| #32 [119225] |
| inehou-pxy05.compaq.com |
yeah 19x19 for a go board compared to 8x8 with a chess board.
But that's not it. Everyone has openings. The thing is that Go's mid game is so hard, that a computer can't play it well enough. Early game it can do fine against a pro, but mid game comes and it gets raped.
Fox |
| #33 striz[119226] |
| CPE00a0cc2d2b10.cpe.net.cable.rogers.com |
nice to see you again - good stuff. a bit hard to follow, but anything decent has to be better than what we're seein lately.
gg, ty
striz |
| #34 Kane[119227] |
| 12-233-180-134.client.attbi.com |
| On top of that, a computer has a very easy time (well, relatively) assessing a chess position because things are basically definite. It's quite easy to program a computer to recognise open and half-open files, defective pawn structures, piece development, board control, etc. But a Go board consists of nothing but black stones, white stones and empty squares. I get the impression that assessing a Go position is very subjective. Course, I've only played Go twice and got my ass kicked both times =[ |
| #35 plus2longsword[119228] |
| asc3-ip51-grr.wmis.net |
| Also in regards to draws, by top level players. The end of a Chess game is determined by checkmate. Material plays a role obviously, and I think this is were the "tactics" vs. "strategy" issue comes up. Because Chess is not just who can win the other guys pieces, but who can checkmate King. When humans beat computers, its usualy because the human sacrificed a pawn for some positional compensation. I don't know if compensation exists in Go. If the computer can't "see" a reason not to take the pawn, then it will, only to find out several moves later that its King will subject to a powerful attack, or some other such device. But Kings can also be put in perpetual check, or a GM might realize the game result will probably be a draw, and agree to it early to save energy. |
| #36 Kane[119230] |
| 12-233-180-134.client.attbi.com |
A problem with actually battlereporting Go is that the board is huge and that games go on for hundreds of moves. Best solution I would come up would to make a flash movie of every little scuffle. Something simple, like "Click me to play movie", or maybe a little window with back and forward arrows. Probably would take up a lot of webspace though.
And I have no idea what a Go middle game looks like, since I've never watched a game and I personally never survive to the middlegame.
Fox, post some good sites with tips/puzzles. K? =] |
| #37 [119232] |
| inehou-pxy05.compaq.com |
Compensation does exist in go. There are many names of moves like a snapback, where placing a piece in a position to be killed, and then if it is taken, the color that took the piece has now placed himself in a position to lose many pieces. THere is a lot of things like that. Strategy is very huge in Go and the computer cannot replicate there. There is strategy in Chess, but it's more tactics then anything, which is why computers do much better in chess then in go. Since tactics can be replicated. Strategy cannot. But that comes down because of the amount of moves avialablet hat allows someone in Go to have more avialable to him. Chess board being smaller with pieces occupying half of it limits what you can do. But again I'm not saying that Go > Chess because of that.
Kane a good puzzle site is www.goproblems.com and you can watch professional games at www.gobase.org.
Fox |
| #38 ...[119233] |
| dsl092-169-188.wdc1.dsl.speakeasy.net |
| I think this game is either the game "Wei Chi" in case you ppl here don't know what that is it is a chinese game played just like this board here, or a replica. Unless WeiChi is a replica of Go then .... ^^ because I played this game A LOT OF TIMES in China. |
| #39 [119235] |
| inehou-pxy05.compaq.com |
mr ...
Go has been around for 3k years. Longer then any board game. China is huge into Go I use there opening often called Chinese High Low. Wei Chi or whatever it was not the basis for Go I can assure you.
Fox |
| #40 Kane[119237] |
| 12-233-180-134.client.attbi.com |
Fox: Thanks for the sites.
But I disagree with your claim that "chess is largely tactics". Chess is mostly strategy. A large portion of the game is spent forcing tiny strategical advantages— exchanging pawns to open a file, or trading pieces to inflict weakness in a position, or roundabout manouvers (sp?) taking many moves to get a single piece into the proper position. Only after many moves of strategising does the oppurtunity for tactics present itself. The most famous chess games are decided by cheese tactics, but they are the exception. The chess games you never hear of (i. e. almost all of them) are decided by good old strategery. Tactics are invariably the result of a good complex strategy, and tactics only pop up if you properly execute a strategy or if your opponent makes a huge blunder. Just because there are more tactical oppurtunites in chess doesn't mean the game is "mostly tactics." |
| #41 Kane[119239] |
| 12-233-180-134.client.attbi.com |
(still continued)
I will make the admission that many beginners get fustrated about chess because they see it as "too tactical". In chess, there are many tactical possibilites and it is hard to see all of them. But once one get a basic mastery of tactics, it's almost all strategy and "strategical combonations." Trust me.
I have too much spare time. |
| #42 [119240] |
| Ottawa-ppp3517228.sympatico.ca |
Chess is without a doubt by far the most complicated and creative game I've ever played. Go looks nice but it just doesn't seem to have the potential for a shocking move or artistic play like Chess does.
Kane, you put more work into this comment section than your last br :O |
| #43 Kane[119241] |
| 12-233-180-134.client.attbi.com |
Dagomar: Please don't get Fox started about the richness and depth of Go. He's almost as bad as me talking about chess. >:)
I had something to add: In the aforementioned game Stienitz did use tactics a lot, but it is important to note that EVERY SINGLE ONE OF HIS FIRST 21 MOVES WERE MADE FOR A STRATEGIC REASON. I cannot overemphasise how important this is (which is why I used caps). He used tactics to gain a strategic advantage, but using tactics is only one way to gain a strategic advantage. |
| #44 Kane[119242] |
| 12-233-180-134.client.attbi.com |
And one can argue that it was a strategic assesment of the position that led stienitz to make move 22. I mean, any good chess player that looks at that position will immediately think "Wow! Look at the power and poise of White's queen and knight, and look at the horrible position of Black's king! With such a strategic advantage, there must be some way to invade Black's position."
Okay I'm done for now. |
| #45 [119243] |
| inehou-pxy05.compaq.com |
The game is mostly tactics. Which is why a computer can play it very very well. Go is more strategy which is why a computer has trouble playing it. I mean that's how it is.. Sure there is strategy in chess, but if there was to much then it would not be playable by a computer. A computer cannot create strategy, but they can execute tactics.
BTW Dag please read up on Go. Chess master after Chess master praise the game so much and play it a lot. But some say they stick with chess because they can compete there but not with Go. It's true. Also artistic play? There are less avialable moves in a chess board then a go board. So there cannot be less. Also shocking move? Please just read about go. THere are history books of the 1700s in japan of games that are recorded that have moves that are the most godly. An example is the "ear reddening move"
Fox |
| #46 [119244] |
| inehou-pxy05.compaq.com |
-continued from the previous comment-
there is a game between 2 of the greatest players ever. One was shusaku, who is considered to be the greatest EVER. All the top players and pros agree that he is. Anyway he was facing an opponent Genen. During the game Genen seemed to have the upper hand, and there are many go pros watching this and see that he does have the upper hand. Shusaku makes a move and no one really notices anything. But one of the observers is a doctor and says, " I think shusaku is going to win" So the other observers ask why. The doctor says, "I don't know much about go, but when shusaku played that last move, Genen's ears flushed red. This tells me the move Shusaku made caught him by suprise and made him uneasy" So the observers look at the move and then realize how good it was. Today pro after pro goes off the hook about how godly a move it was. Etc etc.
Fox |
| #47 [119245] |
| inehou-pxy05.compaq.com |
-continued from previous comment-
So what I'm saying is that you are judging a game you know nothing about. I mean if I came in and said, "omfg the best rts ever is black and white, nothing comes close. This starcraft doesn't touch it." but i had not played a lick of starcraft. Chess btw is not black and white. But you see what I mean. Don't judge it until you have the info.
Fox |
| #48 Kane[119246] |
| 12-233-180-134.client.attbi.com |
Fox did you even read MY comments? =[
There is plenty of strategy involved in Chess. The only reason that computers can beat most players is because computers can see one bajillion moves ahead. Nothing to do with the game's actual strategical content.
Look at Starcraft. Starcraft AI is total ass. Does that mean Starcraft is superior to chess strategically? I think not. |
| #49 Kane[119247] |
| 12-233-180-134.client.attbi.com |
| I'm talking about computer AI, not unit AI, though dragoon AI is pretty crappy |
| #50 [119248] |
| Ottawa-ppp3517228.sympatico.ca |
Probably the greatest catalyst for creativity and artistic play in Chess is the variety of pieces, which allows for the astonishing amount of different moves and strategies available. It's humourous, back when I played Chess I'd usually be too lazy to memorize openings or bother thinking more than 6 or so moves ahead. However, many times everything seemed to come together in one spot of the board, like the clouds parting after a storm ( to be ocvrly dramatic ), and suddenly with a single move I'd blow open the entire game. Chess is amazing that way.
|
| #51 PsySal[119251] |
| h24-81-160-165.cg.shawcable.net |
Chess is frustrating for newbies (me) because it's just too high-stakes; it only takes a few wrong moves to completely screw yourself over. Go has a much different feel, as a new player, at least because the point system lets you gauge quite accurately how you did (in Go you have to tally up territory gained at the end of the match, and the person with the higher score, minus a handicap on the person who got to go first wins).
Go is a very, very human game. I seem to read everywhere that advice to new Go players is to lose your first 50 matches as quickly as possible, so you will develop good instinct, rather than try to think out your moves in detail (as seems to be popular in newbie Chess players that I play against).
YAAAAA FOR GO! =) |
| #52 [119254] |
| inehou-pxy04.compaq.com |
Kane 1st off I've spoken with a programmer or 2 that is into starcraft. They said an ai could easily be made that would make the comp probably beat top players about 50% of the time.
Also I read your post, but you're wrong. Every chess person that I've talked to says that chess is mostly tactics and some strategy and one them is a former chess prodigy. So I respect there opinion more then yours. Sorry =[ Maybe if you could show me some proof that it's more strategy. Um looking ahead a billion moves doesn't have much to do. Since that's just information. A computer just has to be made faster/stronger for that to work. I mean so are you saying based on the billion moves ahead. That a computer 5 years from now with the same type of go program will be good enough to beat pros?
nope I don't believe that.
Fox |
| #53 plus2longsword[119256] |
| asc3-ip83-grr.wmis.net |
| Actually I dont play Go, as I have said. But I've read that go is primarily tactical. (I'm not saying I believe that.) But I think there is confusion in what Strategy and Tactics is. Even amongst Chess players I know there has been a lot of confusion on this. And between to opposing games there certainly would be a lot. Alex Yermolinsky give an excellent definition of Strategy, "as opposed to tactics is a longterm thing. Being a strategic player means being able to create and follow plans. A strategic plan can be conducted by tactical means, independant of the positional principles it was bad on, and tactics, "A tactic relates to the position we see on the board at the moment and continues along a calculated line as far as the moves are forced, and thus calculable. |
| #54 plus2longsword[119257] |
| asc3-ip83-grr.wmis.net |
| continued...Also computers are getting there, but arn't there quite yet when it comes to Chess. The top players generaly don't have a problem in long games, Ilya Smirin, ranked in the top 15 I think, played one of the best computers recently and made it look pretty silly. Loek should have won that match with Rebel (he tried to press a win in the first game, which he had an easy draw in, but stepped into a forced mate instead) Vladimir Kramnik, who is World Champion, and ranked 2nd in rating will play an enhanced version of Fritz soon I think..nobody in the Chess world has any doubts about what the result of that will be. And Deep Blue's defeat of Garry Kasparov was largely psychological manuevering by IBM. I dont think anybody will disagree on their ability to be underhanded. Yasser Sierawan, a top US player, and computer playing expert, immediately challenged the IBM team, after the match finished, but they declined. |
| #55 plus2longsword[119258] |
| asc3-ip83-grr.wmis.net |
| One thing I am pretty sure of, how computers react to these two games, is largely unimportant in terms of..anything. They are both great games that have stood the test of time, and will continue to do so. |
| #56 plus2longsword[119259] |
| asc3-ip83-grr.wmis.net |
Also something to think about..There was this Russian guy, named Rashid Nhezmetdinov, probably the strongest tactician that Chess has ever seen, he even had a good score against Mikhail Tal, a former World Champion, who liked to wax tactical with him whenever the met. Anyway, this other guy Yury Averbakh, who is a Grandmaster, known for his keen positional (not to be confused with strategy-see def above)
skill, explained that whenever he faced Nhezmetdinov he would opt for positions that were deviod of tactics, and it was here that he could be easily beaten. Nhezmetdinov for all his tactical worth never obtaind the Grandmaster title. |
| #57 plus2longsword[119260] |
| asc3-ip83-grr.wmis.net |
^ COMMENT WHORE ^
okay..now I'm done |
| #58 GoZuLu[119261] |
| 195.249.56.3 |
| You guys shure have a lot of spare time ... |
| #59 Amoeba_Boy[119264] |
| spider-mtc-tj064.proxy.aol.com |
| You should use some of your spare time to learn how to spell sure. |
| #60 ThePalisade[119267] |
| 170.158.213.70 |
| lol, Fox you should look up to the comment made by "silverleaftree" who claimed that "you need better players". lol |
| #61 IcE-Inforcer[119269] |
| 209.105.210.34 |
Woa... go comment whores :P
Just kidding. |
| #62 Kane[119270] |
| 12-233-180-134.client.attbi.com |
Fox: How good are these chess players? (And I said one bajillion, not one billion.)
Chess is not "mostly tactics." Chess does involve a large amount of calculation, but i'm willing to bet Go does, too.
Plus2longsword is 100% correct in saying that you can win a strategic advantage by tactical means.
On that chess prodigy who said it was mostly tactics, are you sure he was a prodigy? If he was, I'm not suprised. Most of the great chess prodigies looked at chess from a tactical viewpoint. But that's only one way of looking at chess. |
| #63 Kane[119271] |
| 12-233-180-134.client.attbi.com |
(continued)
To say that chess is "mostly tactics" indicates that you have absolutely no understanding of chess's inner game. When I watch high-ranked people play, EVEN IN BLITZ GAMES, almost every move they make is for a strategic reason, whereas only about half or so have tacical reasons (depending on the player).
If you still believe that chess is mostly tactics, try reading the chess book "The Amatuer's Mind" by Jeremy Silman. No other book I know of shows how important strategy is. |
| #64 [119274] |
| inehou-pxy02.compaq.com |
plus2longsword makes a point in post 56#
I agree with the tactic part. That a tactic is something that relates to position on the board and continue along a calculated line. I see that more so then I see created strategy. If you don't believe this. If you don't believe that a game of chess and more then 50% of the moves are moves that are done on things that relates to the position we see on the board at the moment and continues along a calculated line as far as the moves are forced, and thus calculable. If you don't think chess's moves are more then 50% that, then there is no reason to continue this conversation.
|
| #65 Fox[119275] |
| inehou-pxy02.compaq.com |
I play chess and I play Go. I stopped playing chess and I started to play Go. I'm surrounded by chess and go players. I don't think that tactics is less or better then strategy. I'm just making a point from the start that a computer cannot create strategy very well at all. But it can use tactics since tactics are calculated and math based. Now I understand that computers in chess dont' dominate top players. But they are good enough to be around the pro level. A go computer pogram can't even beat the lowest level ametuer player. Much less any pro.
That's what my original point is.
Fox |
| #66 [119276] |
| inehou-pxy02.compaq.com |
that above post #68 is me.
Fox |
| #67 Kane[119277] |
| 12-233-180-134.client.attbi.com |
| I don't see what computers have to do with anything. Computers don't think like humans. A computer looking at a Go board would be overwhelmed by the possibilities of moves, whereas any human would zero in on a few critical positions. A human has a much easier time finding good moves. In Go, there is a huge ratio of stupid moves to good moves, which is why the computer fails. |
| #68 [119278] |
| inehou-pxy02.compaq.com |
It would just take a computer longer to search for each move. That's all. If you are trying to say that there are to many moves so that's why a comp fails. That's dumb, since all the means is the computer can just take longer to process each move. The number possibilities has no bearing on why it fails. Unless you are some programmer and know things I don't and can show me how it is the reason they fail. I mean if you think that the computer fails because there are to many moves. Then are you saying that every year when computers are faster they get better at go? I find that hard to believe.
Fox |
| #69 [119279] |
| inehou-pxy02.compaq.com |
BTW you don't see what computers have to do with anything? I was just commenting on how I think it's neat that they can't make a program that is good at go, but they can at chess. That's neat to me. Is that good enough to answer your question?
Fox |
| #70 Kane[119280] |
| 12-233-180-134.client.attbi.com |
Look at ANY chess game. Except for a select few rare exceptions, tactics never make an apperance. And a forced series of moves is NOT neccesairily a tactic. It's just a forced series of moves. Tactics always threaten to win material or force checkmate. That's (more or less) the definition most chess authorites use, and that's the definiton I use.
The famous phrase "chess is 90% tactics" is a mantra for fustrated newbies and hopeless patzers. The book "The Inner Game of Chess" makes a correction. Chess is 90% calcualtion. Most people get calculation confused with tactics because tactical play involves a heavy amount of calculation. Calculation is simply finding, looking at, and examining moves, and looking ahead. Tactics are threats and carrying-outs of a win of material, a descicive play, or checkmate.
Fox, it looks like you've got tactics confused with calculation. I know that Go also involves a lot of calculation, so you should know what I'm talking about. |
| #71 Kane[119281] |
| 12-233-180-134.client.attbi.com |
to number 71
wtf? If a computer had enough processing power to calculate everything, OF COURSE IT WOULD WIN. I was saying that a computer fails because it does not have the processing power to calculate everything and has no way to tell dumb moves from good ones. |
| #72 [119282] |
| inehou-pxy02.compaq.com |
hahahahahah ok if you think that then please no reason to even discuss the computer thing anymore ok?
Fox |
| #73 Kane[119283] |
| 12-233-180-134.client.attbi.com |
| Chess computers just calculate every move on the board. Even my iMac can examine 50,000 positions in a second and calculate 8 moves deep in a complicated opening position in a matter of seconds. This is in reality the only feasible way to do a strategy board game, because a computer sucks at strategy no matter what the game. |
| #74 [119284] |
| inehou-pxy02.compaq.com |
to Comment 73
I know Go involves calculation. Ok I guess the defenition out of websters dictionary that I'm reading is confusing me and I don't know how to interpret the defenition of tactics. Because from what I see that's what happends in Chess.
But I guess it doesn't. So is strategy calculation? Since you say chess is 90% calculation. Also that post by longsword quoting Alex Yermolinsky "A tactic relates to the position we see on the board at the moment and continues along a calculated line as far as the moves are forced, and thus calculable." So I mean, I dunno. Am I not understanding things correctly?
Please help me understand.
Fox |
| #75 [119285] |
| inehou-pxy02.compaq.com |
to comment 76. So computers suck at strategy... but they play well in chess? But if chess is barely any tactics, how can they do well? Is the calculation? Because chess is 90% calculation is that why they do well?
thanks for the feedback.
Fox |
#76  [119286] |
| 12-222-61-61.client.insightBB.com |
| Go is fun. The more I learn about it, the better it gets. Fox, I'm pissed at you for reporting this game because I honestly gave you a much better game in game #2. Much more interesting and worthy of reporting... ;) And reading this report makes me want to go play a game of Go which is what any good battlereport should do for its game. So 7, even though I'm biased ;) |
| #77 Kane[119287] |
| 12-233-180-134.client.attbi.com |
A tactic is hard to define, and I came up with the best defenition I could come up with.
Generally, if you play a forced series of moves that leads to winning a pawn or busting open the king's position, people will say "That's a tactic."
But if you play a forced series of moves resulting in a subtle advantage in queenside pawn structure (which occurs fairly often) people will say "Oh, that's strategy." |
| #78 [119288] |
| inehou-pxy02.compaq.com |
I did this game Dref because it was short. I didn't want to try game 2 which was like 230 moves. But I'm working on it now. But holy shit there is a lot of stuff. Also I wanted to do this game to test the waters and get some feedback about it.
Fox |
| #79 [119289] |
| inehou-pxy02.compaq.com |
btw Dref you going to be on IGS later?
Fox |
| #80 Kane[119290] |
| 12-233-180-134.client.attbi.com |
| Computers do well because chess is easier to calculate. A chess position might have at most 20 possible moves (i'm being generous), where as a Go position might have around 100. If a computer were to calculate 5 moves deep, it would have to see "only" 3,200,000 positions (which my iMac could calculate in about a minute) whereas a Go computer would have to see 10,000,000,000 to have the same kind of depth. |
| #81 Who'sGotLag.IDO[119292] |
| usrns160.dialup.hawaii.edu |
For clarification...
Kane, game AI's for chess and go DO NOT calculate ALL the moves possible for the game. If the fastest computer in existance right now attempted to accomplish this right now, the calculations before the first move would not be completed within any human player's lifetime (theoretically the program would never finish calculating, the program would get stuck in infinite loops). Your calculations do not take into account the fact that there is an opposing player who's moves must also be accounted for and factored in. |
| #82 Who'sGotLag.IDO[119293] |
| usrns160.dialup.hawaii.edu |
Part 2
If a computer calculated every option: For an opening move there are 20 options (10 moveable pieces with 2 choices apiece) and the opposing player also has 20 moves possible. From there the computer has to calculate 20*20=400 possible 1st round opening moves by both players AND THEN start on the players' second moves. The second move by the first player (assuming that both players moved the farthest left pawn [relatively to them] would have 21 possible options (22 if a knight was moved toward the center instead and more possible with more mobile pieces, a king in the open has 8 possible moves just for itself). All derivative moves for each possible option would then need to be calculated until the end of the game before the first move according to your statement. |
| #83 Who'sGotLag.IDO[119294] |
| usrns82.dialup.hawaii.edu |
Part 3
My point (which you stated) is that brute force computation is not possible. Shortcuts have to be taken resulting in that a program is only as good as it's programmer(s). A program never understands what it is doing, it just does what it's told to do in the ways that it is allowed to do.
The game program either compares the pieces on the board with certain strategies programmed into it and responds accordingly and/or randomly does moves and calculates optimum outcomes in a limited derivative scope. |
| #84 Ling-erie[119295] |
| host62-7-82-149.in-addr.btopenworld.com |
| omfg u guys are all geeks! |
| #85 Who'sGotLag.IDO[119296] |
| usrns82.dialup.hawaii.edu |
Part 4
The majority of the 10 billion moves you mentioned for Go would largely be pieces randomly thrown on the board, which hardly denotes strategy unless you maneuver the playing on the board to take advantage of it (which in itself is using a preprogrammed strategy on a clear spot of board). When the board becomes too cluttered for preprogrammed strategies, the program probably begins to rely on random moves with limited derivative scope calculation which is hardly an effective strategy. The program simply picks the best choice in it's allotted time period/random move calculations limit (which may have 0 overall strategy involved, according to scope). |
| #86 Who'sGotLag.IDO[119297] |
| usrns82.dialup.hawaii.edu |
=P
We are creative thinkers! |
| #87 testosterone[119298] |
| mkc-65-26-50-59.kc.rr.com |
| 1 word, tic-tac-toe |
| #88 testosterone[119299] |
| mkc-65-26-50-59.kc.rr.com |
;], "yah this is my 3rd or so project" 1 min later...=]
ill read after i finish reading up on go |
| #89 plus2longsword[119300] |
| asc8-grr-ip048.wmis.net |
| Giving it some thought I've come to the conclusion that the best Chess players use strategy constantly, but always at the whim of the tactics present on the board. They then use these tactics to justify their strategic aims. I'm going to do a Chess report I think, and that might shed some light on what I'm babbling about. I don't really believe you can simplify it down to...Chess is this percent of tactics and this percent of strategy. In Chess, Strategy and Tactics are like two different things that intertwine with one another, such that you can't really seperate them. Probably the same is true with Go. |
| #90 plus2longsword[119301] |
| asc8-grr-ip048.wmis.net |
| Oh Oh!, I just remembered something, that I had found in Yermo's book when I first got it, that relates directly to this(I can't find it again though..doh!) It was a game between Anatoly Karpov, 12th WCC, and some strong American Gm I think. Anyway, Yermo was commenting on a really deep strategic principle that Karpov was utilizing in the game. It was marvelous because the move didn't really make a lot of sense from a traditional point of view, which incidently most people from intermediate on down accept as gospel..things like castle early, never trade bishops for knights, and sorts of other rubbish, anyway the move would have been good except it lost a piece..this is something Karpov completely overlooked..interesting eh? |
| #91 goldrush945[119302] |
| f-219-171-res.mts.net |
| How can you really do a chess report? That's got to be hard... |
| #92 Kane[119303] |
| 12-233-180-134.client.attbi.com |
#84-88
You do have a point. However, because Go is so abstract, the computer has an extremely hard time deciding which moves are good and which ones aren't.
I'm not sure how discriminate chess engines are, but mine usually looks at around 500,000 positions to get to a 7 or 8 move deep main line. And it makes strategic blunders all the time. And this is on the confined, clear-cut Chess board. |
| #93 [119323] |
| 12-222-61-61.client.insightBB.com |
| What does it say when I can't beat the Go AI? |
| #94 [119326] |
| inehou-pxy05.compaq.com |
Dref the Go Ai is probably around the 10k level. So for us it's hard, but as you know there is 9k-1k 1d-9d and finally 1p-9p.
I'm like a 25k and I can't beat it.
Fox |
| #95 vincible[119349] |
| 1Cust96.tnt17.sfo8.da.uu.net |
| Let me back up everyone else saying that chess is very intensely strategical. I used to play a fair amount--I beat the Boston youth champion in three straight games, though that was years and years ago, and he was out of practice--and I can say that for most of the game you use tactics to advance your strategic aims. To gain control over critical central squares or of open files, or to gain a slight edge in the speed of your development, for instance. These don't necessarily help you gain material or checkmate the king until much much later in the game. Any book on chess will tell you the same thing: tactics are chosen so as to advance the player's overall strategic goals. |
| #96 vincible[119350] |
| 1Cust96.tnt17.sfo8.da.uu.net |
Anyway, as for go computer programs being worse than chess programs: I agree with most of what's been said above about how go is harder computationally because the board is much bigger.
However, you've neglected one other aspect, which is that much more effort has gone into creating chess programs than go programs. This will naturally lead to chess programs that are more likely to be able to compete at high levels.
All the best chess computers, the ones that compete at master level, were designed to play chess and do nothing else. The hardware was specially designed to create a chess-playing machine. These were very expensive projects. I don't think anyone has ever done this for go, as it is a less popular game. |
| #97 vincible[119351] |
| 1Cust96.tnt17.sfo8.da.uu.net |
| And I'll grab comment #100 while I'm still typing. |
| #98 vincible[119352] |
| 1Cust96.tnt17.sfo8.da.uu.net |
In chess the computer doesn't search the entire decision tree. As said previously, evaluating each of the leaves for any significant search depth would be prohibitively expensive. There are lots of ways to "prune" the tree for any search, though.
There are lots of ways to make the computer take account of strategic concerns. Briefly, you need an "evaluation function" that tells the computer how desirable a given leaf (outcome) of the search tree would be. The most important term in the function is material but you can add terms for the things that usually matter in strategy, like pawn structure, control of the center, ability to castle, etc. You find the minimax given these positional "strategic" factors as well as material "tactical" considerations. So, even though the computer can't do long-term strategic planning it will take into account the positional factors necessary for a good strategic position, thus seeming to strategize. |
| #99 vincible[119353] |
| 1Cust96.tnt17.sfo8.da.uu.net |
| Whew, okay I'm done. I hope someone actually reads all that. |
| #100 GrayDoom[119356] |
| 35-140.port3.gettysburg.supernet.com |
Nice report, even though I don't understand Go all that well. I know to get a general idea of what's going on, but not enough to see deeply into things.
And on computers -- though they aren't anywhere near it yet, and probably won't be for a good long time, eventually computers will be able to "solve" both Chess and Go, using brute force techniques. However well the computers may be doing right now, in the end both games are solvable. It may take longer or use more processing power to solve Go, but it can be solved, just as Chess can. |
| #101 Illgetdroped[119371] |
| nycmny1-ar1-4-62-120-104.elnk.dsl.gtei.net |
| Go... love that game, too bad i cant really win, most of the time anywa. that was a gg and great thinking fox on making a br about a new game. |
| #102 Slings[119376] |
| bgm-24-169-35-188.stny.rr.com |
| interesting, u know that Alexander Yoshi, from old shox, is pretty into Go also I think, u three should get together |
| #103 Napoleic[119377] |
| 1Cust73.tnt9.dial.tor2.uunet.ca |
| GrayDoom is right, all turn based strategy can be solved, unless there are random factors, in which the game can be solved to the greatest percentage of chances to win, though not purely solved. I don't think that a real time game like SC can be solved though, because of the simultaneous nature of play. However you could certainly devise an ai to kick the crap out of any human player. |
| #104 Kane[119400] |
| 12-233-180-134.client.attbi.com |
| How exactly would you coax an AI to micromanage? |
| #105 NotDefault[119424] |
| h24-76-248-3.vs.shawcable.net |
| I really liked that report, but it was rather confusing, unfortunately. :( |
| #106 vincible[119444] |
| 1Cust140.tnt17.sfo8.da.uu.net |
| Of course you could make an AI beat a human in SC. Imagine a tank/goliath/vulture push accompanied by SCVs, where all the SCVs had medic AI. Unstoppable. |
| #107 i wanna know how to PLAY go[119506] |
| evrtwa1-ar3-4-65-149-020.evrtwa1.dsl-verizon.net |
| i mean i don't have to go to a store an buy it do i? can't i like goto a gaming zone of some sort and be able to play it for free? |
| #108 Kane[119507] |
| 12-233-180-134.client.attbi.com |
| 109, imagine even trying to program a computer to push tanks and vultures. |
| #109 vincible[119510] |
| 1Cust182.tnt17.sfo8.da.uu.net |
Kane (#111) I don't see why you say it would be prohibitively difficult to program a computer to execute a competent tank push. Any reason in particular you think so?
I've thought about writing a story around the idea of a Starcraft AI. Hopefully I'll get to it over the summer. |
| #110 Kane[119525] |
| 12-233-180-134.client.attbi.com |
I think that a computer would have trouble with proper tank and vulture placement for the same reason that a computer has trouble placing stones on a board of Go. And keep in mind that it would be virtually impossible for a Starcraft AI to calculate, so it would have to rely on pure strategy.
Pure strategy hasn't worked to well for computers. |
| #111 Kane[119526] |
| 12-233-180-134.client.attbi.com |
| *too well |
| #112 vincible[119529] |
| 1Cust51.tnt17.sfo8.da.uu.net |
I agree that it's virtually impossible to calculate.
However, I don't think tank pushing is all that complicated a strategy, certainly nothing compared to chess strategy. I do think it could probably be boiled down to a few simple rules about relative placement of tanks and mines, how many are needed for a given enemy troop count, etc. I think the main limiting factor in executing it for a good human player is handspeed. |
| #113 Napoleic[119548] |
| 1Cust11.tnt9.dial.tor2.uunet.ca |
This has been gone over in the other forum at least once. I believe it was generally agreed there that if the amount of effort that went into programming a chess ai went into programming starcraft ai, we'd have an ai at least as unbeatable.
For the purely technical aspects, I'm not an expert, but in theory you could program the computer to micro every battle perfectly while always proceeding along a build order that perfectly counters your own, and macroing perfectly the whole while. I don't see why not. |
| #114 Kane[119578] |
| 12-233-180-134.client.attbi.com |
Yeah, so if they make an AI that plays perfectly, it will never lose.
SOUNDS EASY! |
| #115 Napoleic[119632] |
| 1Cust64.tnt9.dial.tor2.uunet.ca |
| probably easier then landing a guy on the moon |
| #116 vincible[119634] |
| 1Cust16.tnt17.sfo8.da.uu.net |
| Certainly not easy. But I bet if a really good programmer spent as much time making a Starcraft AI as a pro gamer like Boxer has spent playing the game, he'd come up with a program that could beat the gamer (assuming of course he could see all of Blizzard's code). The advantages in micromanagement that the AI would have would be overwhelming and I think not all that hard to program. Perfect focus-firing would be trivial. Ghosts that would beat twice their number of goons with lockdown would also be trivial. SCVs with medic AI for vults/tanks--that would almost be a "copy and paste" affair. Hell, it could even micro its mining SCVs. As for macro--the comp would never have a surplus of minerals because it was busy somewhere else on the map. The main "strategic" problem is the question of tech choices and where and when to scout/expand. I don't think either would be impossible. |
| #117 [119669] |
| inehou-pxy04.compaq.com |
comment 110
you can find all sorts of go sites that assist with learning the game. Particularly the american go association.
Fox |
| #118 testosterone[119979] |
| mkc-65-26-50-59.kc.rr.com |
| this sure has a lot of comments |
| #119 [125382] |
| regncache4.sasknet.sk.ca |
| babo 18 |
#120  [158653] |
| 68.79.174.39 |
I always thought GO was the same as Reversi, guess not!
I like Reversi, heh. |
| #121 FARAH[160652] |
| 203.135.46.14 |
| FARAH I LOVE YOU SANA SHEERAZ GOOD :@ AO AO AO AO |
Home
Forum
Starcraft Tournaments
