Rank and what you know
How much a player should know at a given rank?
exswoo: The most obvious thing to say on this topic is to just that a person A is X stones stronger than person B. However, what I'm wondering is if we, as Go players, have certain expectations for how much Go knowledge a person has when we look at their ranking?
This is a bit of a loaded question since there are a lot of ranking systems out there, but if we were to disregard that, what things would you expect a 20k to know and not know? 15k? 12k? 8k? 1d?
I think it would be interesting to see what our ideal vision for certain rankings are and make up some sort of a chart.
William Newman: In my experience, "technique" things, like ability to find tesuji, tend to increase more or less continuously, so it's hard to describe what different levels of players have. Concepts, at least the original insight, are more of a black and white thing. I'd reckon 30 kyus don't quite understand the concepts of life and death (like "two eyes live") and miai, 20 kyus typically have no more than a very foggy idea about shape and sente, and 10 kyus have no more than a foggy idea about kikashi, thickness, aji, efficiency, and sabaki. Of course, these concepts have a lot of technique attached to them, so a 1 kyu is usually much better at keeping sente than they were at 10 kyu, even if they grasped the basic concept by the time they were 20 kyu. (Strong players routinely beat me up on shape even though I've been thinking about it for twenty years or so). Also, I think some people manage to delay their acquisition of these concepts by as many as 10 ranks, so you might be able to find for example a 10 kyu who doesn't quite understand what's the big deal about sente; but it seems pretty difficult to delay longer than that.
Darak: A 10 kyu not understanding the advantages of the sente concept? This is hard to conceive. I think this would probably be one of the concept needed to go past the 15 kyu barrier. However I could perfectly conceive a 10 kyu or even stronger having no knowledge whatsoever about what is joseki and what isn't thus relying on his overall knowledge to define what the next best move is in the fuseki.
Alain Wettach: The importance of fighting for sente goes beyond understanding in abstract what sente means (this at least is not very difficult). Deciding to answer or not to answer locally in a given position should involve a positional assessment of the game and maybe also life and death considerations. (What will happen if I don't answer? If I lose something here, what can I get elsewhere?) If you watch a game between two players of different strengths, you will most probably notice that the strongest one will take sente more often. In my view, this makes much more difference than knowing if a sequence is joseki or not.
William Newman: OK, a 10 kyu player probably knows the definition of sente. But I have more than once had the experience of talking about different variations (like in joseki), and pointing out that in variation X, Black's position is stronger than in variation Y, but in variation Y, Black was able to take sente; and getting a response, or nonresponse, which I interpreted as "Bother me not with those dan-level details! The X position is stronger, I'm not going to worry about the other stuff." My impression is that this pattern of thought, where the player knows what sente is but never takes it seriously, is reasonably common at 15 kyu and not impossible to find at 10 kyu. It seems to me that if you're good at some important techniques like solving life and death and connection problems, you can get to 10 kyu with some pretty horrendous weaknesses in other areas. Someone in another discussion described the strategy of just making the game complicated and hoping to kill something.
Charles Matthews: Bottleneck theories would take a rather different and personalised line on this subject.
alter Pedro: 19k on the Dragon Go Server as of December 2002: I have been playing mostly turn-based Go[1] for a year.
keywords cut and pasted from the text above
- Life and Death
- I guess I'm too lazy to read the position out for all possible moves and reach a conclusion to whether it's alive or dead. I usually just play the most promising point (either to kill or make life) and hope for the best.
- Miai
- Never thought about that when considering my moves and their implications. I can see miai in opponents position (if I play A opp plays B, if I play B opp plays A) so I usually play C.
- Shape
- For now I go by the proverb "empty triangles are bad" and try to avoid them.
- Sente
- It's nice to feel in control ... until it all comes crumbling down.
- Efficiency
- Oh yes! I can very efficiently make a one-point jump towards the center or a two-point jump along the side.
- Aji
- When I don't know where to play and nothing seems urgent I may play somewhere where that stone might have some influence later (is this kikashi?).
- Kikashi, thickness, sabaki, joseki, fuseki
- ???
[1] Turn-based Go is just like e-mail Go: the players do not need to be online at the same time and a game can last for months.
Jasonred: Heh, simple. It's not what you know, it's how well you know it, and how well you put it into practice. Given enough time, and the fact that someone tells me that there's some nakade for a situation, I can eventually figure it out. But in a real game?
Let's face it, I could get a newbie to hang around Sensei's Library and just read up every single Go concept here, memorize the terms, etc, but not attempt any problems or actually play Go. I don't think he's going to break 20 kyu. Though I've heard of prodigies, not in age, but in experience, despite only playing 3 games in their life, they've "suddenly" mastered tsumego, miai, sente, shape, and their playing jumps three stones a game.
kokiri: I think there's definite mileage in the idea of improving by adding new concepts to one's game (e.g. bottleneck theories) As a 3 kyu, recently I've 'discovered' the ability to play much more lightly in my opponent's areas, whereas six months ago I'd not have thought of, say, attaching one of their stones. Whilst I knew, and could recognise in pro games, the idea of sabaki, actually using it in practice was a different matter. The yose, too, is something that I am starting appreciate for the first time. I wouldn't have thought that everybody picks up the same ideas at the same strength though. I also think that it's one of the benefits of studying pro games, or those of players stronger than yourself. I see something I'd not have thought of playing myself and then when a similar occasion comes up in my own games, I try it out (and get cut to shreds the first few times, no doubt).
amc: This discussion leads me to a question: Do you think it is *bad* to read too much at the start? I mean, I have played two games of Go in my entire life, but I've been reading this wiki for a couple weeks now. I *think* I understand the basic concepts of miai, life and death, sente, joseki, tenuki, etc. just from reading about them here and thinking about (non-Go) situations where they might apply. Still, I have no prayer of looking at a board and actually recognizing any of the things I know about.
This leads to a problem I seem to have when actually playing, where I try to think way too much on each move, and end up making poor moves that I thought were implementing those ideas instead of trying to read the board at my level and try to play according to what I know about the game.
This also, I think, puts my opponent in the position of considering me a bad player instead of an inexperienced one.
What do you guys think? Should I get off the advanced parts of the wiki and stick to beginner books for now?
Charles Matthews: I've just checked, and there are 470 pages on SL actually marked as Beginner - which seems quite a large number. Well, Go is above all a leisure activity, and people should approach it in some way they find enjoyable. There is plenty here that I'd say is counter-indicated for beginners, but then the site actually aims to be comprehensive. I would recommend to anyone taking up the game two things: play reasonably quickly, since you have to be nearly dan level really to bring deep background to bear; and get an introductory text and simple problem book.
amc: Hmmm, this leads me to an idea. Should pages in this wiki be rated for intended rank or something? because due to the way wiki works, a beginner like me can enter discussion about basic concepts and in two or three clicks be reading cross-eyed some mumbo-jumbo about the taisha joseki :)
Confused: Speaking as a relative newcomer too, I don't think that all this reading will do any damage. When I was reading about sente, tenuki and miai the first time, I thought I understood the concepts too. Now, at 15 kyu, I'm a step further, and I'm pretty certain that I have no real clue about them, but I slowly start to encounter them in my games; most of the time by destroying my cunning plans.
About the advanced topics, why not just read them casually, without trying too hard to understand them. Once you can relate to them, they'll start to make sense, and if you can't, ignore them. You never know, what piece of information will pave the way for the next enlightenment.
I found it very helpful, to go quickly over my games afterwards and look for what I think was my biggest flaw. This may be a good topic to read up on next, even if it won't help immediately. It's more like a constant pecking on the rock.
Charles: It's like using an encyclopaedia. I suppose a few people need to be told that starting reading at A isn't the best method, but not so many. At the moment the keywords only recognise four levels (and three in practice). But surely material at the wrong level is just uninteresting.
Baz: The concern about reading something beyond your skill being detrimental reminds me of a discussion in a Philosophy class about the hypothetical dangers and ethics of introducing highly advanced technology to people in the past (during a lesson on the logical consequences of time travel). Reading about techniques and concepts beyond your skill is like giving cavemen a TV and a machine gun... it won't make a difference until you have an electrical outlet and bullets :-)
Kungfu: What I remember: When I was 25 kyu I didn't understand the basics. When I was 20 kyu I smelled sente, and knew enough to make eyes.
At 15 kyu I could not progress until I understood how josekis flowed into each other on the sides, sente, and studied life and death problems.
At 10 kyu or abouts I could not progress until I discovered concepts such as furikawari and korigatachi. After a study of these concepts in the josekis I knew, and some life and death study, I became 5 kyu.
Now I am 1 dan in the Canadian Go Association. I became this rank by studying some professional games and articles, watching Hikaru no Go, buying an expensive Go board, and calming myself down. Playing a lot of games also helped. However 1 dan CGA means about 3k or 4k on IGS / KGS, to me.
How I plan to improve in the future:
+1 stones - Study of life and death problems graded for dan strength players.
+1 stones - Study of joseki - goal is add 50 josekis.
+1 stone - Replay commented professional games, memorizing own games, self-analysis and general Go world immersion.
+1 stone - Concepts such as direction of play, increasing reading strength.
+2 stones - Taking lessons from a pro dan strength player.
Tamsin: I suspect that the relationship between knowledge and strength is different for every individual player. What is certain, though, is that you can improve if you fulfil two goals:
1) Increase your knowledge about different areas of the game
2) Improve your understanding of that knowledge (i.e., recognising when to apply various ideas, sequences and techniques).
Also, as you get stronger, you tend to take old knowledge and understanding for granted. This is the "not seeing the wood for the trees" effect: you so much want to try out the sharp variation of the large avalanche joseki that you have just learned that you forget to ask yourself how it fits with the rest of the board. To correct this, I believe the use of checklists can be helpful: for example, make an effort to remind yourself throughout the game to "look a the whole board", to "seal the opponent in while not getting sealed in", to "look for the biggest move" and so on. Striving hard to base your thoughts on the most basic concepts can be really beneficial, because this usually causes the more advanced ideas to fall into their natural places.
Tropsy: I don't think there are "milestones" between ranks. I mean, I'm a KGS 20k, and just being a bit careful I beat an overconfident 10k yesterday in an even game. Just looking at the whole board more often, reading before playing moves that may be problematic, and trying to let your opponent play as few big moves as you can while keeping sente can make you a few stones stronger. Being careless and thinking you will have no problem because you're playing a weaker player can make you drop two or three stones easily.
OneWeirdDude: That could actually add to the point. Perhaps one milestone is to learn to think more often before acting, as it were - and voila! you get five or ten ranks on the spot.
tartuffe: While these concepts can be grasped on some level before the given level, I think this is about right for "application of the concepts". Feel free to correct me!
25k-21k: The net, basic dead shapes, eyes, ladders, ko, corner - side - center, how to make shimari
20k-16k: empty triangles are bad, sente is good, don't let the snake into your moyo, 3-3 invasions, the most common joseki, create eyeshape, bamboo joints, miai, throw-ins, big points in fuseki
15k-11k: attention to direction of play, a few common endgame tesuji, the table shape, attention to fast play, the difference between the third line and fourth line, avoid unnecessary early play on the second line, invade before the opponent completely solidifies territory and passes, joseki selection based on developments of other corners, ladder breakers
10k-6k: probes, two stone edge squeeze, timing of invasions, proper use of thickness, balance of territory, invade vs. reduce, better understanding of direction of play, size of endgame points
5k-1k: light play, reasonable expectations for invasions and reduction
In my experience from 30k-4k strength at life & death and tesuji account for most of one's rank. These can be improved rapidly by reading problem books and playing many games.
P7A77: I've recently shot up many levels to 7k AGA, so I've been pondering this a bit lately. I hit 15k or so almost instantly when I started playing two years ago, so I don't remember much before that. For me, getting to 12k was a result of learning more about L&D. 10k was after I had started reading more on fuseki and playing the whole board (tho' my understanding was minimal). My recent ascention into single-digit kyu range is because all of the concepts I've been picking at are starting to work together, bolstered by my study of direction of play and influence, and the fact that I finally understand at least a little bit about true sente and the value of a single stone. From what I've seen, that's a big difference between DDK and SDK. DDKs are still hesitant about their games and have gaps in their knowledge, whereas SDKs are thinking more completely, have concepts of influence and sente, and can see the game as a whole. I don't feel there are any major concepts I've not yet studied (excepting joseki), so now it's simply a matter of refining my tactics and deepening my strategies and mastering the elusive sabaki.
erislover: P7A77 said: "I don't feel there are any major concepts I've not yet studied (excepting joseki), so now it's simply a matter of refining my tactics and deepening my strategies and mastering the elusive sabaki." (laughs) When you've done this (I have not yet myself, so please don't take this comment as being critical), you'll be playing people who have also done this. --What "simple" matter will there then be to improve? At the time of this writing I'm AGA 6K. I can say for certain that there are holes in my picture of the game that I cannot even tell are there. Sometimes even "simple" life and death problems have given me trouble because I completely fail to even consider certain moves. If this happens in life and death, what sort of gaping holes exist in my strategic picture of the game?
Patrick Taylor: I have some thoughts that might help lead to a WikiMasterEdit.
- There is a difference between knowledge and application. Just because someone reads a lot doesn't mean it has to affect his rank.
- Around 1d, one I imagine there is a body of shared knowledge. Otherwise, you'd get eaten alive.
- Below 1d, tartuffe's list seems to make sense. The 5-stone rank units seem to fit with conventional logic.
Alex: Regarding knowing vs. applying, I think you could say that tartuffe's list is approximately correct in its assessment of when most players understand the fundamental idea of the concepts in question, but it's in the next rank bracket that those concepts gradually become integrated in the player's game in a practical way. So although a 20-25k may be able to find the move for a simple ladder or geta when the stones are already on the board (or avoid making a move that will immediately be caught in a ladder or geta), it's in the 15-19k range that the player will start foreseeing a potential ladder or geta arising a few moves in advance and reacting appropriately. Likewise, 1k-5k may understand the idea of light play and why it's good, but one often sees these players starting off making light shape but then allowing it to become heavy by responding incorrectly to peeps etc., whereas attempts to stay light are generally more successful at dan level.
As for what a dan knows, I think it's safe to say that most dans have at least fundamental understanding of all the concepts mentioned here, but it's more a matter of finesse in applying them...
As well as the things like reading, tesuji and life & death that can be improved on at any level, here are a few of the things that I think are still in the process of being absorbed at my low-going-on-mid dan level:
- Positional analysis early in the game - assessing ill-defined moyos, making furikawari, etc.
- Global miai - moves in two separate local positions that are similar in size and can be treated almost like miai
- Probes - aside from the well-known ones like the 2nd line attachment to a keima shimari
- Honte - when it's big, when it's slow
- Timing - when attacking a group is urgent, when it should be delayed to choose the best way to attack, when to play somewhere directly vs. attempting to induce, when to play kikashi, etc.
greenviper (3k) This page very interesting to me, not least because i really ought to know more than i do ... but i suspect Tamsin's "I suspect that the relationship between knowledge and strength is different for every individual player" is entirely true. I'm not convinced you can guarantee 3k has a concept of anything. I am constantly amazed at the lack of reading and the unwillingness to tenuki at my level, but on the other hand i don't really understand anything else. I think this is a lot to do with the difference between knowledge and applying that's been mentioned. I have read a lot here (in the library) about various concepts, and have some vague idea about them, but the only one i'm truthfully think i have the hang of (tenuki) seemed to be pretty much inbuilt for me; improving 'what i know' doesn't seem to help so much. Hence Tamsin's comment ...
See also: