Learn to Play Go
Posted3 months agoActive3 months ago
online-go.comOtherstoryHigh profile
supportivepositive
Debate
20/100
Go GameLearning ResourcesBoard Games
Key topics
Go Game
Learning Resources
Board Games
The online-go.com 'Learn to Play Go' resource is well-received by the HN community, with users praising its effectiveness and sharing their personal experiences with the game.
Snapshot generated from the HN discussion
Discussion Activity
Very active discussionFirst comment
3h
Peak period
73
12-24h
Avg / period
21.7
Comment distribution152 data points
Loading chart...
Based on 152 loaded comments
Key moments
- 01Story posted
Sep 27, 2025 at 7:50 PM EDT
3 months ago
Step 01 - 02First comment
Sep 27, 2025 at 10:33 PM EDT
3h after posting
Step 02 - 03Peak activity
73 comments in 12-24h
Hottest window of the conversation
Step 03 - 04Latest activity
Oct 3, 2025 at 3:48 PM EDT
3 months ago
Step 04
Generating AI Summary...
Analyzing up to 500 comments to identify key contributors and discussion patterns
ID: 45400376Type: storyLast synced: 11/20/2025, 8:00:11 PM
Want the full context?
Jump to the original sources
Read the primary article or dive into the live Hacker News thread when you're ready.
Personally, my favorite tutorial I went through was The Interactive Way to Go (https://way-to-go.gitlab.io)
Also notable is Sensei's Library (https://senseis.xmp.net) which is a very old and unbelievably thorough wiki on everything Go. It's a cool place to browse even if you don't play.
I learned chess in 7th or 8th grade and was easily able to get to about 700 Elo on chess.com after barely learning the rules, which is about the 60th percentile on the site. I only play a couple games a year now but can still hold my own against 1200 Elo opponents, which is in the 90th percentile.
I feel like I have put in just as much effort into learning Go. I bought a book and have been doing exercises. But I’m still in the 0.1 percentile on the site! (Yes, that’s not a typo.)
I’m sticking with it because it’s fun and that’s all that matters. But I definitely have a lot to learn.
According to the elo formula, a 700 elo player is expected to "hold their own"--draw or win--against a 1200 elo player 1 in 19 games (so they will lose 18 of 19 games), and against a 1550 elo player 1 in 134 games (so they will lose 133 out of 134 games). A 1200 elo player is expected to "hold their own" against a 1550 elo player about 1 in 18 games (they will lose 17 out of 18 games). However, the chess.com and FIDE elos are from very different pools--1200 elo at chess.com is probably equivalent to about 600-800 elo FIDE.
Why would that lower the average elo? Elo is a conserved quantity: for you to gain points, your opponent must lose that many points, and vice versa.
So there are two obvious facts about the average rating:
1. In a closed system, the average rating can never change, not up or down, not by any infinitesimal fraction of a point.
2. In an open system, the average rating can change, but for it to go down would require players with an above-average rating to leave the system.
In the much more common scenario where beginners come in, lose a bunch of games, and then leave, the average rating will go up over time.
Because the average chess.com player is considerably weaker than the average FIDE player.
> In the much more common scenario where beginners come in, lose a bunch of games, and then leave, the average rating will go up over time.
Weak players come into the chess.com pool faster than they leave.
By definition, someone entering the pool enters with the average elo (for a closed system). The rate of entry and exit doesn't affect the direction of average (open system) elo. (It does affect the rate of divergence; the faster entry is, the longer it will take for the open average to diverge from the closed average.) What matters to the direction is the elo players have when they leave, not how long they take to leave.
Consider what happens if people enter with 1000 elo at a rate of 10 per day, and leave with 800 elo at a rate of 3 per day:
> Because the average chess.com player is considerably weaker than the average FIDE player.This... is nonsense. That has nothing to do with the average elo.
The situation you describe is mathematically impossible. Please consider that before you repeat the claim.
700 elo on chess.com is completely normal for a new player who knows the rules, but not much more. I have a good sized sample to back this up based on coaching classes of complete beginners and getting them to setup a class specific account.
Chess.com seems to have 1.3 million players with 100 ELO which is their hard cap minimum, so that is literally the same, if not worse, as a random move generator. So there has to be some weird bias there.
[1] - https://lichess.org/stat/rating/distribution/rapid
It is essential to study these tactics in this website... if only because they are the only "ground truth" known about Go. But for rapid improvement, the only real way forward is to play lots and lots of games to learn how the early game flows. Direction of play, which side of the board is most important and other such details.
Seems like a reasonably good tutorial in terms of layout. But just pointing out: joseki and direction of play is "more important" in terms of winning. Its just damn near impossible to teach so maybe its best for beginners to ignore this incredibly important (and difficult) subject.
---------
To put it in perhaps more concrete terms: playing a "tactic" position may net you +10 points across a sequence of 5 moves or so. (IE: One well placed tactical move, and ~4 followup moves may capture 5 enemy stones + 5 territory simultaneously from your opponent). However, every single early-game move is worth nearly +20 points of territory if played correctly. I'm serious.
That's why when you watch top-level Go play, there's a lot of "teleporting" across the 19x19 board, searching for the most important positions. And there is also very, very loose play and possible sacrifices / aji. (Maybe its not a true sacrifice, but you'd be willing to sacrifice if the opponent over-extends).
A couple of things I love about go is that you don't need to memorize fuseki, and that applying joseki correctly is as much a matter of judgment as it is of memory.
(I am a 1 dan go player but haven't played much in the last 15 years.)
Yes, at some point when people are somewhat able to take a decent lead home the fuseki becomes important. Before that, beginners really need to understand how to „move“ their stones, how to defend and connect their groups and how to cut and capture.
If you see a strong player win against a weak player with a large handicap it always goes down the same way: the strong player places stones all over the board such that eventually many many skirmishes appear all over the board and then she is patient to take small advantage after small advantage, manifesting groups and territories out of what looks like thin air to the other player.
At a somewhat higher amateur level and above the fuseki again loses importance and the distinguishing factor is fighting skills and judgement, fuseki and prep just becomes table stakes.
Agreed. Now check the tutorial series.
There's no movement. There's no connecting of groups. It is ENTIRELY cutting and capture, and life-and-death puzzles.
-----------
https://online-go.com/learn-to-play-go/bl1-stretch/2
Look at this puzzle. WTF is this?
The "stretch" ("correct move" according to this tutorial) seems wholly inappropriate compared to 1-point jump, horse-move, diagonal move .... or hell just running in the other direction (to the right, escaping towards the center).
To know the appropriate move requires knowing what is going on around the whole board (and not just what's going on locally).
----------
https://online-go.com/learn-to-play-go/bl1-stretch/6
Or this one is perhaps more egregious. This is simply the wrong direction of play entirely. 20+ point mistake to play the "correct" move here. The correct move is a 2-point jump (or greater) along the 3rd or 4th row northbound.
> The "stretch" ("correct move" according to this tutorial) seems wholly inappropriate compared to 1-point jump, horse-move, diagonal move .... or hell just running in the other direction (to the right, escaping towards the center).
The stretch is good shape here. I can't even tell which keima[0] you have in mind but that makes no sense here. A diagonal also just looks strange, unless you mean hane, which lets opponent cross-cut and will end with losing that stone. Similarly, a 1-point jump (I assume you mean upward) lets opponent wedge in and will end in gote, and running to the right (I assume you mean another 1-point jump) allows opponent to push you around.
The stretch is important here because it prevents opponent's hane, which is very severe. It also threatens to turn and block opponent's expansion decisively, so it may be sente depending on the rest of the board. In general it will make sense to push twice here, then make a one-point jump.
> Or this one is perhaps more egregious. This is simply the wrong direction of play entirely. 20+ point mistake to play the "correct" move here. The correct move is a 2-point jump (or greater) along the 3rd or 4th row northbound.
Per AI, on an empty board, the stretch is about 1 point worse than those jumps (establishing a base); a one-point jump in the same direction of the stretch is about as good as making a base; and best is tenuki. There is simply no urgency to settle those stones without anything else in the area. But stretching or jumping damages Black's shape and limits further expansion of that group. The point is simply to prevent Black's hane there.
Regardless, when they say "correct move", this is not supposed to be about what would be best overall to play in the position. It is supposed to be about a) recognizing the shape and its purpose and b) doing it in a sensible direction. And you really have to learn that sort of thing by example and by physically laying it out, because trying to give the rules for choosing a direction to stretch is harder.
[0] By the way, I'll spot you that Western teachers usually do use English for ikken tobi and kosumi; but "horse-move" is cringe. Most Western players who actually talk about the game will probably know "keima", but you can at least say "knight move".
But this kind of discussion is missing from the tutorial in any regards. This is where I believe I'm stuck in the ~9kyu level and feel like its been mostly about this middle-game movement / direction of play stuff since 15kyu or so.
Sure. But you put in the effort to get that far. You can't just put everything in the beginner tutorial. They'll get confused. And there's only so far you can go with one-size-fits-all lessons, without feedback from the student.
After having checked the positions you mentioned, I had put you somewhere in the 4-10 kyu range. For me this was the time when I found out how important it is to play efficient and fast and play big spots. In order to improve further, I had to learn to appreciate thickness more because of the downstream benefits. As the fighting of both players becomes stronger, thicker shape makes your side of the fight much easier and prevents everything from crumbling.
Take your first example where you suggested 1-point jump or keima instead of nobi. The main point here is about 1) liberties and 2) having your group‘s „head“ run ahead. There is also this concept of a hard head or a soft head and the nobi creates a wonderful hard head ahead of your opponents stones that can not be bullied. If you play any other move, a stronger opponent will (locally) immediately and gladly play that hane without thinking, which takes a liberty and forces you to play again to defend your stones. Depending on your moves, the opponent would turn your group into a big dumpling and collect very nice stones on the outside in the process.
Perhaps you would find the games of Kitani Minoru or Lee Chang’ho inspiring.
This is an unintuitive aspect of go because it is different from virtually every other strategy game. In most strategy games, "macro" (large scale logistics) is what determines winners at all levels of play, and then at higher levels where logistics skills are similar, small-scale tactics start to discriminate winner from loser. In low-level go, you'll find "micro" (small-scale battle tactics) determine the outcome of most games.
This is because of the "teleporting" you mention. When the opponent can materialise units and start a battle anywhere they want -- including inside your base -- small-scale tactics becomes important. (I once read the analogy that "if you were able to drop a siege tank into the opponent's main base at the start of the game, micro would end up determining low level StarCraft games too" -- only players that excel at local tactics would survive to see the end game with any base worth mentioning.)
For each hour of training, exercises in reading and local tactics is what will improve your rating the most. At least for the 20 or so first grades. Someone who is good at reading will obliterate all positions of someone who only knows the more subtle aspects.
No, it isn't. As a decent chess and go player I can tell you that they're both just tactics until you approach the master/dan level. And what is strategy if not just a longer form of tactical play?
At the end of the day, strategic play is just play that sets up tactics later on.
Or, to quote Fischer: "Tactics flow from a superior position"
You cannot tenuki in Chess.
In Go, especially at the 15kyu to 10kyu double-digit dan level, the opponents are full of opening and middle-game mistakes. The best response is often to ignore your opponent and play the most powerful move elsewhere on the board.
Knowing when to tenuki (ie: ignore the last move, play elsewhere) is a HUGE point in Go strategy. Its exceptionally difficult to play sente / forcing moves. Playing a sente vs gote sequence is what separates the 1-dan (experts maybe 1800+ Elo equivalent players) from the rest of us mere mortals. But recgonizing that the last move was gote (non-forcing) is maybe a 10kyu / 1200-Elo kind of thing.
-------
Chess is almost all tactics. Go on the other hand, is Strategic, as the concept of sente/gote/tenuki allows you to validly ignore the opponent's plan and work out your own plan.
You still need a solid tactical basis in Go. You cannot just run away from the opponent forever. But you might be surprised at how "valid" tenuki moves are.
------
For me, the growth from 15kyu (maybe 1200 elo in Chess) to 9kyu was a strong focus on sente, gote, tenuki, strategy, joseki, direction of play, strong vs weak. All "strategic" play that often sacrifices local tactics for greater point gains elsewhere.
Indeed, "weak" play in Go (ex: a 2-point jump) is WEAKER in terms of tactics. You are explicitly making an area weaker and easier to kill in exchange for moving faster on the board. A 2-point jump will ALWAYS be the worse tactical choice than a 1-point jump or solid connection.
This isn't like in Chess where a sacrifice immediately becomes apparent either. It can take 50+ moves before a position is played out and the difference between strong-connected play vs a 2-point jump shows up.
In any case, even 20kyu beginners can improve their games if they play 2-point jumps (or other weak / loose patterns) appropriately. Yeah you need the basics of tactics there otherwise the 20kyu player just loses all their stones at all. But protecting your stones / strong play is actually very very weak and will trap you as a beginner. You MUST play faster (but weaker) connections if you want to break through double-digit-kyu. Players just get too strong by 9kyu or 8kyu to rely on tactics alone.
You absolutely can! If it's during a tactical sequence we might call it an intermezzo. If it's not during a tactical sequence we don't usually have a name for it in chess but it happens all the time. The mainlines of the KID are famous for having theory where white goes for a queenside attack and largely ignores the kingside and vice-versa for black, just as one example.
> Chess is almost all tactics. Go on the other hand, is Strategic I used to think so too but I think this is a meaningless and superficial comparison now. Each go move is simpler by itself but to counter balance that you get a much bigger board and generally much longer games with much longer tactical sequences (I'm sure you've had games where you spend over 50 moves in a long tactical battle over a mojo, for example).
In Go, I think everyone feels like they're a "beginner" for years.
In my experience, absolute beginners (30kyu or weaker) should study tactics. You have nothing else to study after all and need a baseline. But even by 20kyu or 15kyu, you _WILL_ stall out and be unable to continue if you're unable to recognize when a 2-point jump, horse move, running on the 4th line vs 3rd line is appropriate.
And you probably should be studying joseki theory in any case, because you need to start the game with _SOME_ move. And then you need to connect your opening theory with different, strategic level moves somehow.
----------
Is it not strange to you that there's a 50+ set tutorials with not a single one discussing the 1-point jump, horse move, or 2-point jump, or diagonal move?
This set of tutorials gets to Tiger Mouth (they call it a "hanging connection, section 3.11) before it gets to diagonal moves (ie: never).
There's no discussion on the appropriateness to sacrifice stones to gain momentum or territory. Etc. etc. This is perhaps more of a 10kyu level concept. But seriously, some of this stuff (ex: 2-point jump vs horse move) is simple enough for a 25kyu beginner.
Well yeah. But look at these tutorials. They're all local tactics. Nothing on early game or middle game strategy.
Despite dozens of tactical positions, I don't think a single one of these holds a one-point jump, two-point jump, horse move, diagonal move, 4-row move, 3-row move or similar pattern.
I'd say this stuff gets extremely important around 15kyu, where your tactical knowledge is passable (maybe not great, but passable). It becomes more important to move around the board with 2-point jumps, recognize sente vs gote and tenuki vs weak gote moves from your opponent.
Tactical knowledge will only get you maybe to 20kyu or 15kyu at the best IMO. Then you're forced to learn "squishy" and "opinionated" discussions that no one really knows how to teach. There's patterns (ie: 2-point jump or horse-move), but its not like there's any ground truth to knowing when these moves are appropriate.
----------
I guess people's point here is that if you choose a 2-point jump as a strategic connection when a horse-move was more appropriate, you might lose 1 or 2 points. But both of these moves are likely worth 5, 10, 15, or more points.
The more important bit is knowing when to play strong and connected (often because each move is strongly sente / has momentum and forces an opponent's response), vs when a sequence is gote and the opponent (or yourself) should consider a move elsewhere (ex: possibly start an invasion).
Fuseki (opening) doesn't matter much for most players. AI confirmed that a wide variety of openings, even weird ones far removed from the usual credo, work very well. At worst you may lose a couple points in doing so, but unless you play at a high level that's negligible and will never be the reason you lost.
Joseki (corner sequences) is also not that important, and certainly not something any beginner should spend time on. In fact, a common Go proverb is to "learn Joseki and lose two stones" (get weaker). We often see beginners learning Joseki, getting confused when their opponent doesn't follow the sequence they have in mind and ultimately blundering their corner. Or they ask "how to punish that?", without realizing that many moves are good even if they are not Joseki, and there's nothing to punish.
With so much emphasis on cut and strong play, anyone completing this set of tutorials is going to be an absurdly strong tactician and then lose 20 points as the opponent horse-moves around the board.
Surely you've played the beginner who favors tactics and capture at the expense of easily captured territory? Given this set of tutorials, do you think any beginner will understand sente, gote and tenuki? And even if a beginner somehow understood it, what basis of play will they have? There's literally no tutorial or discussion on walls, influence, 2-point jump, 1-point jump, strong vs fast play (etc. etc.)
I'm also not sure how that relates to your earlier point about "Joseki", which I was disputing (like many others).
Look at the tutorials. Do you not see the lack of strategic discussion? Its evident from the outline.
There's no strategy here. There's no Joseki theory. There's no movement tutorial. There's no middle game, direction of play or other such tutorial.
---------------
Like read the tutorials I'm trying to describe. They're full of sound tactical advice but any beginner who only uses this set of tutorials will have difficulty on the strategic aspects of their game. I think Joseki theory would help them very much (as well as many other kinds of tutorials or discussion).
There's nothing wrong with one online tutorial to focus on tactics. But its also important that any reader (especially the beginners who come into this discussion) realize what they're missing. Joseki is perhaps the most obvious missing element from this set of tutorials.
IE: Every beginner coming in here won't know what to do for their first move of a real game even if they complete this set of tactical tutorials.
------
What's your idea of the readers coming into here? My mental model is that this is Hacker News, and that very, very casual beginners on the order of ~30kyu are going to start reading these tutorials and maybe start playing Go.
I think my advice from the first post is helpful to them. If you complete this set of tutorials, definitely study something like Joseki (or other ignored topics).
--------
Go isn't a game like chess. In Go, 80% of your pieces will often survive into the endgame. You don't need "strong" tactics because in the vast majority of cases, you can survive. The difference between "strong" tactics vs "weak" tactics is only a few stones or points of territory.
In contrast, the difference between playing in the correct direction of play and not is easily worth 20+ point moves, as you capture large swaths of territory. Maybe you as a ~single-digit-kyu or low-dan player have forgotten these kinds of mistakes. But I assure you that double-digit kyu / beginners will make these mistakes all the time.
There's no discussion of 2-point jump, 1-point jump, horse-move, diagonal move, loose play or connected/strong play in this set of tutorials. Or the value of 4th row (center-oriented influence play) vs 3rd row (edge-oriented territorial play). I'm not seeing any discussion on invasions or defense of invasions.
-------
I'm like 9kyu. I'm no expert. But I'm certainly at the level where tactical geniuses beat me in many positions, but I just wreck a lot of n00bs in the opening, and hold out until the ending.
The good news about opening theory or middle-game theory is that if the opponent is playing tenuki (ie: they ignore your most recent move and play somewhere else on the board), you're probably focused on the incorrect area.
On the other hand, if you're up vs a weaker player, YOU need to be the one playing tenuki. Its surprising how awful players are at double-digit kyu is at this. You will only see tenuki opportunities if you have superior opening/middle game skills than your opponent.
Nah, they aren't the most important (you can do without), not are the usual ones particularly difficult to learn.
Source: I'm a European 4 Dan.
(To expand on this: if you're a beginner, joseki don't matter. When you become a strong player (several of my friends are professional players), joseki is something you can usually come up with, or you come up with a similarly good non-joseki move, which is also ok. Practically, the game is usually decided in middle game fighting.)
https://online-go.com/learn-to-play-go/bl1-stretch/6
https://online-go.com/learn-to-play-go/bl1-stretch/9
I'm saying that this "correct move" is a middle-game error. Maybe its my 9kyu brain being bad here, but there's nothing about this tutorial (and many others like it in this tutorial series) that strikes me as a strategic error.
And I certainly think that nobi is a bad concept to teach if the tutorial hasn't covered diagonal moves, horse moves, 1-point jump and the like. Its not that "nobi" or "stretch" is a bad thing to teach, but its just one option in a sea of valid middle-game options.
------------
If the tutorial is going into things like this, it really should be going into at least the basic 3-4 Joseki and why various moves are considered good.
The first one is indeed probably just a little bit worse than a jump.
The second one, I'd extend too: it strengthens white and black has to come back to live. You could think about the hane instead, but at worst black can play the cuts and come back to live at 2-2.
I'm baffled by this comment. Both professional commentary and AI evaluation confirm that most joseki mistakes are very small, often on the order of 1-2 points, because the temperature is low. There are specific joseki that turn into fights (e.g. Taisha), and it is possible to lose 10-20 points in those, but first of all, it's typically possible to play conservatively and avoid those joseki, second, most errors in them are smaller, and third, mid-game fights end up being even bigger (an error in a capturing race can be an almost unlimited number of points, having 40 points at stake is common).
I'm curious what level you are? As a 4kyu (European), I can confidently say joseki is less important than reading up to my level. I believe stronger players say the same well into the amateur dan level.
I'd say that at 9 Kyu level, my main gains were playing and abusing Tenuki. Refusing to respond to the opponent's (weak) early game moves and instead playing significantly stronger elsewhere on the board.
If the opponent were stronger than me, I'd pay attention when they ignored my moves. Its actually very difficult IMO to play sente every time as a double-digit kyu (or even high-single digit kyu). Recognizing that the last move wasn't forcing and that playing elsewhere is a surprising way to get ahead.
Vs stronger opponents who can tell when the board is sente, gote, appropriate to Tenuki, and is able to count up Ko threats... well yeah. I lose. But there's significant skill in this part of the game and NOTHING in this set of tutorials that teaches it.
> To put it in perhaps more concrete terms: playing a "tactic" position may net you +10 points across a sequence of 5 moves or so. (IE: One well placed tactical move, and ~4 followup moves may capture 5 enemy stones + 5 territory simultaneously from your opponent). However, every single early-game move is worth nearly +20 points of territory if played correctly. I'm serious.
Sorry, but this is complete nonsense. "Play lots and lots of games" is the only part I can get behind. I have seen tons of people get well past the beginner stages entirely self-taught and with a focus on fighting because that's the easiest thing to self-teach. A few basic joseki sequences go quite far (especially now that AI analysis has done such promotional work for early 3-3 invasions) and people who try to figure out the rest by intuition generally make smaller mistakes than those who can't read out squeeze sequences or don't understand semeai theory. Of course joseki are important to know, but emphasis on them often leads to a false confidence in having mastered the opening for students who aren't thinking about direction of play or other fuseki issues.
If a tactical sequence appears to net you 10 points against an opponent who is playing correctly, you basically already had those points. (Well, half of the points for most capturing sequences, since they'll usually end in gote; but then you have to consider the opportunity cost of tenuki.) This is as true for capturing sequences as it is for invasions and reductions. Which is why students are counseled to be very conservative in counting frameworks; they are not territory.
If a tactical sequence actually does net 10 points because of your opponent's misplay, that doesn't mean that your moves were only worth 10 points, or even that each move was only worth 10 points. It means that your moves were worth 10 points more than the opponent's were. Similarly, correct early-game moves may be worth a lot of points (compared to passing, the reference value), but most incorrect moves are worth almost as much. Even things that are marked as clear mistakes in a joseki textbook, with clear refutations, might only cost one or two points (although, yes, they can be catastrophic; and of course pros do have to worry about every point). So this is very much an apples and oranges comparison to the "value of moves in the opening".
But also, that value is about 13, definitely not "nearly 20". We know this because even now that we have ferociously strong AI players (who inherently make moves with a higher average value), they still accept a komi of 6.5. And if you ask them to evaluate the first few moves and let some of them be passes and see the change in the score, you'll rarely see 20-point swings in any opening, but you can trivially create such positions (and much larger swings) in middle-game fighting.
> That's why when you watch top-level Go play, there's a lot of "teleporting" across the 19x19 board, searching for the most important positions. And there is also very, very loose play and possible sacrifices / aji. (Maybe its not a true sacrifice, but you'd be willing to sacrifice if the opponent over-extends).
Sure, but 10 kyus can do a reasonable facsimile of this as well. They just have bad timing, or make bad choices about sacrifices, or have wrong ideas about how to use the aji.
[1] https://www.youtube.com/watch?v=WXuK6gekU1Y
When 9D-master Sudol attributes human qualities to the beauty of his AI opponent's creativity upon formulating certain moves... is definitely eye-opening. Hubris replaced.
I think with the rise of KataGo, its becoming clear that AlphaGo's "dumbfounded" strategies were instead incredibly strong tactical play with hilarious levels of blindness with regards to ladders.
It feels like modern AIs (like KataGo, which are hundreds of times better than AlphaGo) are getting closer to what humans consider appropriate strategic play.
Go players must be humble because if the opponent is stronger then the opponent wins. But Go AI Programmers don't necessarily have to be. Go AI programmers look for the weaknesses, lean upon them and yes, prove that AlphaGo/AlphaZero never learned ladders. Ever. Or other such concepts of strongly forcing moves (ex: loose ladders).
That's one of KataGo's biggest innovations. Explicitly programming a ladder solver so that simple ladders can indeed be factored in by the neural net.
----------
I do wonder how Lee Sodul will react if we told him that the superhuman AI he played could not see ladders... and had other such key weaknesses.
I wonder what Kasparov thinks of the fact that he was beaten by such a primitive chess program as Deep Blue's!
Its very strange to me that AlphaGo / AlphaZero was unable to ever learn ladders. It shows that the way humans learn and AIs learn is quite different (or at least, MCTS + Neural Net machines learn differently).
Yes, Go is a strategic game of patterns and perhaps we humans overemphasized the ladder. Nonetheless, its a concept that humans can see and calculate with reasonable speeds that the (earlier) AI was unable to do (and now we've built stronger AIs that can prove this weakness).
---------
This 100% makes the 2016-era discussions about the "strategic brilliance" of AlphaGo come up into question. Now that KataGo is superior (ie: MCTS+NeuralNets + dedicated Ladder Solver), we need to double-check all those "strategically brilliant" moves with the newer AI and see if a ladder messed with them.
Etc. etc.
The things AlphaGo sees aren't necessarily useful to us humans, nor are they useful to modern AI-levels of Go. They're just... that. Trapped in 2016. There should not be any great mystery assigned to the 2016 era game, aside from it being a pivotal moment for AIs.
The game itself is now suspect, now that we know all of AlphaGo's flaws. As Go players, we have better things to study.
Thanks for this detailed response/thought. As a layperson, cannot wait for the documentary on KataGo (AlphaGo II).
DeepBlue was the big chess moment. No one writes or cares about Stockfish vs DeepBlue.
Well, chess players performing opening analysis for their games care. Go players care about KataGo being more readily available and stronger than AlphaGo or AlphaZero. But it's back into the weeds and grind, it's not too interesting a story outside the Chess or Go world.
Fox, without a doubt. I recommend Weiqihubb for a quick-access app to asian servers on all platforms, with puzzles (https://walruswq.com/WeiqiHub).
If you dislike chess because you don't like abstract total information strategy board games you will not like go. If you dislike chess because
- it has too many rules,
- the board is too smalll,
- the pieces move around too much, or
- it doesn't involve adversarial, collaborative construction,
or any of the other things that make go different from chess, you have a chance of liking go.
I think that's why I don't like chess. It seems to me that a winning strategy would be to think as far ahead as possible by enumerating all the permutations. A few heuristics exist however.
So to do well you have to learn how to support your reading ahead with heuristics and a feel for the game.
A famous amateur player and advocate for the game once went through all the game records of Go Seigen in order to digitize them. This means having to pore over hand-written diagrams looking for the next number in the sequence of moves. Obviously this is easier if you can guess where to look. But, if you guess them all correctly, then you are playing just as well as the old master! After spending a good few months on the task, he was a significantly better player!
I'd never heard of this!! Who're you talking about?
If you think this is equivalent to the description in GP then you are quite simply incorrect.
If you think it's just the actual reason you personally don't like chess, then I'm not sure why you asked in the first place; of course go is an "abstract total information strategy board game".
Yes, in principle such a game has such an algorithm for perfect play. In practice, computers cannot and do not do that for chess (although they make a reasonable approximation) and approach go quite differently (in much the same way that earlier attempts at computer chess did, before Deep Blue's much more brute-force approach proved effective given enough computing power). Getting AI for go to its current superhuman level involved multiple complete revolutions in how the systems were programmed.
Even Connect Four wasn't strongly solved until 1995, and checkers is only weakly solved and that not until 2007 (https://en.wikipedia.org/wiki/Solved_game).
Progress in chess (and I assume Go) is about training your subconscious so that your mind naturally pushes you in the right direction with minimal effort. Think about something like writing. When you're writing something you aren't really thinking through each word in your vocabulary, comparing them, and picking one - it all just kind of flows without you even trying. The same thing happens with chess mastery.
This is why some people say you're not "really" playing chess before a rather high rating. Less experienced players will struggle to simply not leave material hanging. Then as they improve that will no longer be an issue but then they'll still struggle to avoid simple tactical ideas. But once you move comfortably beyond that phase, the game becomes much more about the things people want it to be about - strategy, plans, big picture stuff that's lots of fun. It's one of the way the game draws you in - it gets more and more addictive, and rewarding, the better you become at it!
In go there are some sequences of "standard" moves (joseki) but it's highly controversial whether memorizing them even helps at all, see another thread in this same comments section.
The reason less experienced players get obsessed with openings is because they make regular tactical mistakes in the opening, and then blame lack of opening knowledge, as opposed to their lack of tactical ability. In other words they try to memorize their way out of tactical mistakes, which is impossible. At that point, after it inevitably fails to work, they claim they've plateaued, lack the IQ, maybe the memory, maybe are too old, or whatever other excuse.
This is made even worse by the fact that opening knowledge is ostensibly easy to improve whereas tactical vision seems very difficult to improve. In reality it's the exact opposite. Exact opening lines will fade from your mind rapidly (though general ideas will stay with you forever - but that's another topic), whereas grinding tactics might not 'feel' like you've learned anything, but overtime will permanently train your intuition to where it needs to be to start seeing major gains.
On top of all of this - one could simply play 'freestyle' chess (the starting pieces locations are randomized) and suddenly there is 0 opening theory. But you'll find that your freestyle rating is going to be strongly proportional to your 'normal' chess rating!
This is different from opening understanding. Understanding the importance of tempo, development, controlling the centre, the different pawn structures, middle games and endgames that result from different openings, the plans and motifs typical in various opening complexes. Any late beginner to intermediate player needs to pick and study an opening. The problem is that instead of studying the opening, players try to memorise lines without improving their understanding of the resulting middlegames, and the plans they should be playing for. Then, when their opponent diverges from the main lines(which in my experience happens in 99% of games between players below 2000, because it's very rare that both players have memorised the same long line), they don't know what to do.
I'm a 1900 FIDE player, I have an opening repertoire of sorts. For instance I play the modern benoni with black. An extremely theoretical opening, and yet I have only a small handful of longer lines actually memorised, because they're simply too complex for me to figure out over the board(e.g the b5 lines against Bd3 h3 Nf3 setups). But what I have studied extensively is the strategic landscape of the benoni, games by strong players in the opening, etc. And I have years of experience playing the opening. I know what kind of exchanges typically favour me, or my opponent, what pawn breaks each player should be trying for. And all of that knowledge is crucial for me to get anything out of the opening. I have beaten players tactically much stronger than me in this opening simply because my understanding of this specific opening was better than theirs.
Tactical ability is obviously important, but it's definitely not everything.
And it has nothing to do with memorization. I mean you mentioned the b5 stuff against Bd3/h3/Nf3 setups. You might not be able to calculate the depth of what happens if white manages to hold onto his extra pawn, but you can certainly calculate to at least the point of 'okay, I'm getting my pawn back in most lines, disrupting his center, and getting my play going. if the one line where he holds onto it (Bxb5 stuff) then he's going to have a bit of difficulty castling, his pieces look disorganized, his extra pawn and b2 both look weak.' That's more than enough on general principles to go for the sac I think.
Just my engineering brain picking up on it.
> Chess also has this "problem" thanks to rules like castling and en passant capture.
Chess is intended to be stateful. If you forget whether a castle has occurred (and then the king walked back to its starting position, and a rook repositioned into the corner) or not, chess players will note that you've messed up the game. The castling rule is there to stop you from castling more than once.
Go is not intended to be stateful, and if you forget that a particular board layout may have come up in the past, go players will not note that you've messed up the game. It doesn't matter. The ko rule isn't there to stop you from repeating a board twice. It's there to stop you from having to repeat a board layout an infinite number of times, because things like the need for food and sleep would interfere with the game.
Wrt the implementation concern, this distinction means, for example, that you must always track the castling state regardless of whether a player asks for it, whereas you're fine not bothering to track the history of a go board unless a player asks for it. You can just say "if you want to invoke ko, press this button, and we'll remember that board layout, and if it's already been flagged, the game will draw". That isn't done, but it could be done.
You say:
> if you forget that a particular board layout may have come up in the past, go players will not note that you've messed up the game. It doesn't matter. The ko rule isn't there to stop you from repeating a board twice.
Wikipedia says:
> Rule 8. A play is illegal if it would have the effect (after all steps of the play have been completed) of creating a position that has occurred previously in the game.
> Consequence (ko rule). One may not play in such a way as to recreate the board position following one's previous move.
> While its purpose is similar to that of the threefold repetition rule of Western chess, it differs from it significantly in nature; the superko rule bans moves that would cause repetition, whereas Western chess allows such moves as one method of forcing a draw.
To me that sounds like you do need to track this, in both chess and in Go, though for different reasons (to force a draw vs to prevent an illegal move). Is this not enforced in practice?
Since then, we’ve been going to the local Go club in Warsaw, and it’s become our main hobby. We play each other almost daily, travel to tournaments (sometimes abroad), and even spend our vacations at Go summer camp.
The camp is actually a magical event. It takes place at a campsite in the middle of the Kaszuby Lake District. The conditions are spartan – you either live in a tent or a five-person cabin, and hot water is scarce. But the crowd that gathers there is incredible. Over breakfast you might get an impromptu intro to lambda calculus, in the evening you might end up in a deep philosophical conversation, or hear travel stories from far-off places, or suddenly learn way more about knitting than you thought possible. When we first went, it felt like discovering our long-lost family.
The Go community is much smaller than chess, but also far more tight-knit and welcoming. I’ve heard chess can be more cutthroat, while in Go there’s this unspoken understanding that if you drive people away, you’ll have no one left to play with.
When I travel, I like to drop in on local Go clubs. It’s always been a great experience – I especially enjoyed visiting the San Francisco Go Club in Japantown.
I play almost exclusively over the board. I prefer long, thoughtful games, and I can’t really focus the same way on a screen.
Oh, and the anime about Go, Hikaru no Go, is really good (you should watch it even if you don’t care about the game).
> in Go there’s this unspoken understanding that if you drive people away, you’ll have no one left to play with.
Definitively not in online Go. I ran into some people who clearly thought racist trash-talk was a way to reduce the competition.
Yes. Part of the reason Yumi Hotta's manga (which the anime was based on) was written was to get younger people into the game, and it is credited in part for reviving the popularity of the game in Japan. Traditional board games like Go and Shogi have faced a lot of competition from video games over the past few decades.
It kind of follows a general trope in manga/anime where characters' eyes are smaller or thinner to indicate age/seriousness/maturity/intellectualism. The Korean kids tended to have the same kinds of eyes as the Japanese kids, like Hikaru's main Korean rival (forget his name) having almost exactly the same trapezoidal design as Akira. Hikaru's eyes also change into this later in the series, changing from his original carefree wide-eyed design.
I look forward to it the whole year. I've been going there for the past 20 years and been the main organizer the last 10 years. The magic happens by itself though.
Thank you, Sir! I have learned so much from these pdfs.
Some other interesting aspects of the camp:
The event’s currency are Łosie, which you get by taking part in classes and winning tournament games. By the end of each week there’s an auction where you bid for prizes. You can use your Łosie from previous years, but Tasuki implements an inflationary monetary policy to keep old-timers from becoming too rich (every year Łosie rewards get doubled).
Some people have been coming from abroad for many years, and at some point just figured out it makes sense to learn Polish (not the easiest of languages).
I really enjoyed the Chinese drama adaptation of this - more so than the original anime somehow.
https://mydramalist.com/45437-qi-hun
What does "lid" mean here?
"Drive by" game, to me, sounds like they've got the board set up somewhere in the house, and make moves over the course of a day (days?), rather than sitting across from each other and having dedicated time for the game.
It took me a bit to decode also.
I even lost the game I played with my son, teaching him the moves.
It's a great game.
truly, the thing I love about Go is two things. Firstly, the handicap system means you can play a good game between neophytes and high-dan players, it's mutual. Secondly, its built on such simple foundations. No horsey-horsey-moves.
Thirdly, even innumerate people who can't count to 3 can play it. You're playing a game substantively unchanged from before chess
We rarely play anymore, I should invite him over sometime :-)
Sometimes I play a move with a huge, but hidden threat behind it. If he plays elsewhere instead of answering locally, I get to play a clever sequence and capture some stones. I could just wait for the blunder and win. Instead I give a quick lesson in tactics: here is my plan, if you want to play elsewhere, your move needs to have an even bigger threat behind it.
He is learning, and now I face my clever moves being player against me. This makes it harder for me to win (it is about 50:50 with the handicap), but also more fun for me to play.
You could ask your spouses brother for a "teaching game" or a larger handicap, or a bit of both.
Our games feel relatively close, it's just evident that he is always ahead.
Handicap stones give a greater advantage on smaller boards, but 19 x 19 is the standard size. I've not seen any specific guidance for smaller boards.
Source: I am playing both Go on OGS and chess on Lichess.
It's extremely instructive for players at all levels.
[1] - https://www.youtube.com/@DanielNaroditskyGM/playlists
There was an older guy running the local club at the university, a strong player who had a passion for teaching the game. This is in a small town in quite a remote area. I later heard he had learnt Go in the 1970s, first by reading books, then by meeting Japanese fishing crews who had come to the harbour to play. We're still friends today.
Go has many proverbs which act as rules of thumb for playing well (e.g. 'hane at the head of two stones'). IMO an underrated one is 'make friends playing Go'. As great as resources like the link are, I still prefer to play in person, over a board, for this reason.
Maybe you'll find them useful or fun :)
I randomly tried a few exercises on the linked web site - nicely done!
Off topic, but I wrote a commercial Go playing program in the late 1970s. This was a great hobby.
I'm a chess guy, and I like to play blitz with 5+0 and bullet (format equivalent to tik tok) so games tend to be frenetic, but it's quite rare to find a Go game on those formats, they tend to have 40+ min in time. And honestly, this is a big W for Go.
The site is doing this exactly because those volunteers haven't shown up. Be the change you want to see in the world.
But, damn, this is a beautiful game. It's fun and there's no upper limit to the challenge. The way the game ends (by agreement that there are no further moves of benefit to either player) is amazing, as are the handicap rules (which make it so players of differing (to a point) strengths can still play all-out against one another).
The boards and pieces are works of art in themselves, and the game is steeped in tradition.
That said, the boards and stones can be expensive, so lots of people DIY which is also fun. There's a web page out there that talks about making boards with MDF and Sharpies. And you can buy melamine stones relatively cheaply, though if you can find glass or ceramic stones for a little more, they're much nicer.
When we were starving students in junior college and first learned about the game, we went out on a country road and gathered all the relatively small, round stones we could from the shoulder. Then we went to my parents' garage and found cans of black and white spray paint to paint the stones, and some leftover wood paneling that we cut in a square and lined with a magic marker.
The set was, arguably, offensively ugly. But I still have it in my closet.
I just tried to play 2.16 is the group alive. It told me I was wrong for saying black was alive when it clearly had two eyes.
Then I tried 2.17 and it showed me a screen of a black formation that has two eyes, but then says "White to play. Make two eyes." This is clearly the wrong prompt for this board. Not only is the color wrong, but the formation already has two eyes so I don't know what you want me to do here.
3 says I'm wrong for saying its alive. It's clearly alive. It's a big black formation with one eye in the corner and one eye at the bottom
4 Says I'm wrong for saying its alive. It's clearly alive. It's a big black formation with two eyes diagonal from each other, one on the edge.
In 2.17:
2 shows me a black formation with two eyes already made and says "white to play. Make two eyes."
3 is doable but it also says "white to play" while showing a black formation.
4 shows me a black formation with two eyes already made and says "white to play. Make two eyes."
5 shows me a black formation with two eyes already made and says "white to play. Make two eyes."
6 shows me a black formation with two eyes already made and says "white to play. Make two eyes."
FYI I'm using firefox
"Alive" pieces remain permanent throughout the game. In both 2.16 examples, white can capture black by filling the gaps.
"Eyes" only have one space each and are fully surrounded by a single color. In both 2.16 examples. there are no eyes. Look at 2.17 with this new definition to see where the gaps are not yet eyes.
2.17 (3) is asking you to place a white piece within the black formation. This renders correctly for me in firefox.
It's interesting that the shading for the white pieces is preserved. I wonder if a dark mode extension is affecting the colors.
For example:
> In 2.16:
> 3 says I'm wrong for saying its alive. It's clearly alive. It's a big black formation with one eye in the corner and one eye at the bottom
Should look like:
(I'm also using Firefox and it's rendering correctly for me)|
| B B B B B
| B B B B B
| _ B _ B B
^^^^^^^
Link to screenshot: https://imgur.com/a/odDYg9K
When I use Librewolf I think I get the intended render, but using Firefox this is what I get
and a great anime in its own right
1. https://gomagic.org/ , it has free and paid content, and I learned a lot here.
2. The European go journal. A nice print publication, I’ve not lived in or near the Asian countries with a stronger go history so I’ve enjoyed getting printed problems and go news.
3. The “so you want to play go” book series by Jonathon Hop. I like his writing style.
https://lichess.org/learn
1 more comments available on Hacker News