After playing 400 games for last month’s Western Server Challenge, I’ve got some clear insights about my strengths and weaknesses as a go player. My 9x9 rating is 3.7d and my 13x13 rating is 3.3d, but my 19x19 rating is 0.6k. One could question the accuracy of these ratings given the WSC time controls and the different player fields and all that, but I think it’s more or less correct. The 0.6k rating also reflects a couple dozen games with slower time controls that I’ve played since the event.
I would like to get my 19x19 rating closer to my small board ratings, but I’m not sure how to go about it. In the past when I wanted to improve I would do lots of tsumego to improve my reading and fighting abilities, but if reading and fighting were holding me back then I would be doing poorly on small boards, since that’s basically all there is.
So that suggests that the problem has something to do with my opening. I at least try to play simple and solid moves in the opening - no complicated joseki, if I can avoid it - but I do find myself in bad positions in the early middle game a lot. And in a lot of my games I end up trading away territory for big influence, but I often get beaten in the ensuing center reduction fights. I think the core problem is either that I don’t know what good shape looks like in 19x19, or that my positional judgement is bad and I go for more than I need to win.
Any advice in this scenario? Anyone else better at small boards and struggling to improve in 19x19?
So the difference between ranks becomes much smaller on the smaller board sizes. So you can beat a 3D on 9x9 but not on 19x19. My wife (12kyu) can beat me (2D) on a 6x6 lol
It is usually not the opening that causes players to get stuck at 2k and 1k. It is a sum of small things. The difference between kyu and dan is small is some sense, but requires a not small amount of effort to make up the small difference.
A LOT of it is discipline. This comes in many aspects. Are you taking every single ranked game seriously? Are you resigning early because you are behind? Are you finishing a game with a loss instead of doing an all or nothing move to change the result? (Only when you are down 40+)
A big thing I see at kyu level is the lack of real reading. Real reading is actually reading the sequence instead of assuming it. A good example to know if you do this is with Go problems. If you are playing out the first thing you see before picturing the entire result and trying to counter it, then you are doing “lazy reading” or rather not sharp reading. The is a large difference in half reading something and playing vs knowing the result 100% before you go down a path. I’m even talking about 5 move or less sequences. If you see it doesn’t work, or it’s going to die, then move on and come up with a new plan. Don’t invest. Don’t panic. Just re-evaluate the board. Dead stones are future aji.
Another thing I see is that kyu players don’t play enough to improve. As a guess, let’s say it takes 50 games to move one rank if you are efficient. Some players will play 100 games a year and wonder why they are not improving. You have to play the game to improve
I see a lot of mental issues as well, which I am not an expect on helping, but if you find yourself getting upset after losses or putting a ton of pressure on yourself to improve, this will really hinder your progress. A small way to combat this is to try out a different opening or style and pick something that sounds fun. Play Tengen if you want, but enjoy playing the game and have fun.
Don’t be lazy with even one move. It is very difficult to hold yourself to this standard, but at Dan level, Go is kind of just work lol. If you don’t read it, you lose. This includes endgame.
Opening, pick something that sounds fun to try.
Middle game, don’t be lazy and READ the cutting points. Don’t miss the cutting points. Fix the cutting points that work. Read the freaking cutting points!
Endgame, don’t be lazy. Find all the sente every single move. Double check every turn. Use your periods here to find all of the next sente moves. If you have to play an obvious move, that is the perfect time to look around the board for 20 seconds. (Unless you are using Fischer)
Finally, play games and ask for reviews. Don’t have to for every game, but learn from your losses.
Without actually looking at your games, my first thought is that you’ll benefit from studying some professional games. If possible, get or make a printed diagram and lay the stones out on a physical board. This will improve your instincts for shape and direction of play.
Is it worth picking out one or two 19x19 games that you’ve played and posting them here for review?
Lots of good advice there, thanks for writing it all out. And yeah, I think laziness is a factor - sometimes I play when I’m too tired and I can feel myself being too sloppy or hand-wavy with reading and counting. That pretty consistently leads to bad outcomes
That said, I’m still worried about my openings. Or maybe it’s early middle game, rather than the opening. I have two pieces of evidence.
First, I’m finding that in my losses the problems start quite a bit earlier than I realize during the game. I’ll attribute the bad position to making a reading error in the fighting or something like that, but when I review with the engine it turns out that my problems started much earlier, by making poor shape early on or playing a move that’s too slow. It’s not uncommon for me to be behind by 10 points within the first 25 moves without knowing it, even against players around my rating.
Second, my win rate against stronger players isn’t bad when I am able to hang on in the opening. Within the past month I’ve managed to pick up several games against 2d’s and 3d’s, I timed out a winning position against a 4d, and I held a 5d to half a point. What most of these games had in common is that I held my own in the opening and then got some sort of advantage in the fighting. In previous eras of my go journey, victories on 19x19 against players 2 or 3 stones stronger were extremely rare.
Possibly all that is a red herring, but I can’t shake the sense that I’m just missing some fundamentals about the early stages of the game.
Here’s a couple games where I felt (and still feel) like I’m just missing something basic:
There’s a couple of missed tactical opportunities in both games. Maybe it’s just that simple. But in both cases I really misjudged my position more fundamentally than that.
I’m the opposite, stronger at 19x19 as I hardly play small boards. First impressions from your games: your don’t like tenuki enough because they’re isn’t really tenuki on 9s, though it is a bit hard when your opp is a brawler. Look for ways to get sente to get to big points even if it means losing something locally.
This comment resonates with how I feel during a lot of my games. I’m in a big fight, and I’m not really getting anything out of it, so I want to just go play elsewhere, but then I spot a nasty local followup for my opponent and lose my nerve.
If it’s not too much trouble, can you give a couple of examples from my games above where I can tenuki and take a big point?
I realized I gave advice targeting 3k-1k and you are 1D.
Well, what you wrote tracked. I’ve had a couple extremely discouraging losing streaks recently, and after reviewing and reflecting I’d say half the losses were just due to the sort of laziness that you describe - hand waving at cuts instead of reading them, being sloppy about sente in the endgame, etc. I think if I were more consistent and disciplined then my rating would improve a stone or so. But beyond that, I feel like I’m missing something.
From 1D-8D do the following almost every day for 1 year.
Play 10+ games a day. Mostly 5 min games on Fox and a couple on OGS. Review losses with AI for 5 minutes or so.
Do 100-200 sdk problems a day.
This is obviously a big time commitment. Probably the natural response is “duh, improving at the dan level takes work!” Before trying to take it on: how quickly should I expect to start seeing improvement? OGS 5d would be great, but right now I’m less motivated by the numerical value of my rank, and more by the sensation that there’s some fundamental insight about the game that, say, most 2d’s have but I don’t.
The best way I can explain it is: when I lose a 13x13 game to a 5d, it’s usually for a concrete reason that I can understand, like I needed to find a specific tesuji, or I miscounted the score and needed to go for more. Sometimes it’s like that when I lose 19x19 games too, but often when I review with the AI it ends up being something more nebulous and seemingly random like “you thought that exchange that you made in the opening was balanced, but actually it sucked and you were just lost the whole time”.
So maybe I’m really asking about how to improve positional judgement specifically, or something like that, so that I’m at least aware of when things are going off the rails. For instance, maybe is there some way I could be using the AI more effectively during review, within the scope of your proposed training regimen?
I think there is more in doing life and death problems as just development of the reading and perseverance to solve. Not so that you’ll find similar position but that the perception of status and available space will get more acute and so gives better global decisions.
I get this impression too, and also a feeling that sometimes you’re too focussed locally, even when fighting locally between groups, and that there is something in the whole-board view of games that you miss considering or holding in mind, (perhaps from more experience on the small boards ?).
There were global moves in both games which relied on the life and death/connection status of groups in a bigger way, even in ‘‘local’’ fighting between groups, which you sometimes missed.
I think that this is a lot less common in 13x13 (which I think you’d mentioned is what you mostly play ?), for example, to have quite as much global space involved in which cutting and connecting on as grand a scale are critical (or at least as frequently as it can be on 19x19 ?), or if not, perhaps it’s just an area which you might overlook ?
But in 19x19 fights, it’s very often important to consider bigger, global moves and implications in the fighting too, when choosing where to respond (above the local tesuji etc.)
I think there are also generally more options/directions to hold in mind when groups are entangled (and more groups/cutting aji/global issues in general), given the larger board, on 19 compared to on 13.
(for example, in the first game, you missed H2 to connect your groups around move 105, and O12 around move 169.
And in the second, I think you missed the elephant-eye peep to cut at P13, around move 129. Or something like exchanging Q15 for R15 first, then P13 – I left the variation in the game. it’s miai to push through in two places, and then the big eyeless White group looks to be in big trouble)
I think these types of global connect-cut moves are very intuitive to me as a mid-high dan, whose focus is fighting, on a 19x19 board, and many other players 5-6d+, and it feels to me your fighting in these games maybe lacks that feeling/constant awareness a bit even though the very local moves/shapes/responses also seem ok and rather solid.
That would be consistent with this too, in the sense of global positional judgement.
There are some other things too, but I can’t quite put my finger on them at the moment.
For my part, I was able to get from 2d-6d KGS within less than a year (less than a few months from first reaching KGS 2d to 5d, then some months later to 6d) by playing many games in which I took my time at each point to really consider each move, the global situation, etc. and try to improve at every point
(actually rather slow games, anywhere from 30-60 minutes main time, but usually at least 45. I had the time for this, though.)
I suspect that engaging in fighting and situations which made me less clear or sure what was going on, created complications, etc. helped me get an experiential feel for them and what can happen, and thus developing my intuition.
But I would say that taking my time and really trying hard to read as much/well as I could, and consider options and play the best move, also gave me a lot of experience that I wouldn’t otherwise have had, had I put less into that.
(At the time, there was not strong AI, so I reviewed all of my games manually, searching for even just a few points that I didn’t understand, and going over variations which I wasn’t sure of or could have gone differently, to explore the possibilities.)
If there are situations like that now, I think that you could use the AI, especially if you have the real-time supporter AI on OGS, which suggests further branches in real-time as you play more stones, to get an idea of what it might suggest.
(for example if you aren’t sure where to play etc.)
But it does also suggest moves which might be more difficult for a human to handle, so if you don’t understand a continuation down one branch or find it harder than a more understandable-to-you choice, often it can be better to play what you can at least somewhat grasp or follow-up on (unless you are doing it to learn and as a challenge), than an AI tesuji which doesn’t make sense to you.
In terms of fighting and opening and midgame, I see in those two games that you played based on influence a lot – it feels fine in the opening to me (relatively) before your opponent invasion/reductions and the fighting begins, but it feels like the falling behind in the midgame was due to using the influence inefficiently and letting the opponent build light groups, or run through your potential.
How about looking over some pro games and/or high dan games more often to get a better sense of where they tenuki or play differently ?
I’m not sure if that would help you, but I remember that having seen that helped me a lot when I was developing my global sense.
High dan and pro games also helped me see moves like the type of global tactics that you missed in the fighting, a lot more easily (those moves tend to be played and taken advantage of very quickly in high dan games, especially weaknesses like the elephant-eye peep, if they are left unfixed by the opponent – probably easier to understand than pro games in that regard as it occurs more frequently)
I used to watch them when they were more common on KGS many years ago, although I’m sure that trying hard to find the best move at each move, also helped a lot in that regard.
(It’s possible to see tesuji etc. or style, then not try to find them in one’s own games when one plays, of course)
I suspect that if there is some issue with understanding midgame/opening positional judgement, that could help, or being exposed to stronger players discussing/commenting their games and how they think about global positions.
(I was always stronger in fighting before, so also had some feeling like my opening/positional judgement/choices may not be as strong, and both helped me a lot along the way to have seen.)
I think others might have better recommendations, but things like this ?
Other players also recommended this in another thread.
In terms of tsumego and tesuji, actually I had done none in the period when I went from 2-6d+ (I felt the deep reading and doing my best in many games was enough), and when talking with a pro about it it was something they didn’t feel was necessary if playing a lot either.
However, I think I have found it helpful to have seen stronger play and techniques before (not necessarily local tesuji, although those also, but things like global fighting, opening or midgame tactiques that stronger players/pros use)
And I’ve since done some tsumego and tesuji problems (and other types of problems - whole-board, etc.) which I’ve found very helpful, especially some common shapes/key ideas like handling 3-3 invasions or the 8-space shapes in the corner, or conceptual global/whole-board or direction of play ones in discussion with stronger players about the possible options/answers and all of the merits or drawbacks.
(we have a group session each month in our club which does the latter from real game positions, and usually there are a lot of global possibilities and interesting discussions)
Maybe you could also ask for a few pointers from your dan opponents in general, as they might have more insight and perspective into your play after having played you ?
It’s usually possible to get a feel for your play after having played an entire game with you, and they might see useful things, or have helpful positional judgement analysis.
If you find someone willing to review/discuss together with you in more detail after the game, that’s often great for learning, too.
(especially if your main experience/speciality isn’t on the 19x19 and you might not have been exposed as much to how other dan players think – it may be less nebulous and easier to understand when reviewing with a human who can discuss and explain the concepts too, as compared to the AI )
It might or might not help directly with what the positional judgement issue is you’re feeling, but you can also learn to count more often (for example, at least once in the opening/beginning of midgame, later in midgame, before taking an important global decision, and in the yose), if you know how to do that.
(to simplify the method often used, counting the solid territories of each player, if you assume all forcing moves around the edges/boundaries of it are played)
This is 11, for example,
And this is 4.
It can help a lot in general (even if you play with influence, to get an idea of how much your influence has to give you in value, if you give away more territory to your opponent, for example) to be aware of, or to spot some areas of positional misjudgement, for example if you find yourself 15-20+ points behind in early midgame.
(although your opening didn’t seem bad in either of those games, it was more a matter of things like midgame fighting/direction, and not tracking important connections which severely affect the life and death/groups balance on the whole board)
(especially in the first one for your groups, and in the second missing an opportunity later, and also leaving a bit of an exploitable shape with aji/cutting points at K4 (move 51).)
White has spearheads (strong 3rd line stones) at F17 and G3. Those make the upper and lower sides relatively uninteresting to develop (for either side) compared to the right side that has a wide space between 4th line stones (Q14 and Q4) open to knight’s move approaches at C or D.
Yet you end up playing at A and B developing towards white’s spearheads, allowing white to develop the right side.
To avoid that sort of outcome with black, I’d probably have played on the right side at C, or perhaps D if I wanted to avoid early fighting.
To avoid that development with white, I might have played kick-and-jump instead of Q14, to continue with a territorial amashi strategy that leaves black fewer opportunities to develop quickly.
I don’t think I would have played that R8 hane, because I would be too afraid of white’s immediate cut at Q8. White’s corner group is already alive and white’s upper right group is relatively safe, so I think that cut would be pretty severe for black. White responded at S8 in the game, but I feel that move was way too timid. If white didn’t want to fight with the Q8 cut, a checking move at R10 or R11 to strengthen white’s upper right group still seems much better than strengtening white’s strongest group on the board.
Caveat: I didn’t check with AI if my moves are actually better than yours (I don’t expect a big evaluation difference either way). I just wanted to share considerations I use in the opening that you might not commonly use, but which might help to reach less akward positions.
I’ve been taking my time parsing this excellent answer, and I’m going to need more - thank you for writing it all out. I have a couple follow-up questions for now, if you’ll continue to indulge me.
I think that this is a lot less common in 13x13 (which I think you’d mentioned is what you mostly play ?), for example, to have quite as much global space involved in which cutting and connecting on as grand a scale are critical (or at least as frequently as it can be on 19x19 ?), or if not, perhaps it’s just an area which you might overlook ?
But in 19x19 fights, it’s very often important to consider bigger, global moves and implications in the fighting too, when choosing where to respond (above the local tesuji etc.)
Yes, this is a very accurate assessment. In 13x13 the consequences of your opening choices are nearly immediate, and it’s more or less possible to directly read out the interactions between local fights. In 19x19 I think the global interactions between local fights are much more difficult (or impossible?) to fully read out, and so I assume one is forced to rely more on intuition and heuristics. And yes, I spent several years playing primarily 13x13 games, so it wouldn’t be surprising if my skill in this area is underdeveloped compared to the rest of my game.
I’m going to follow your subsequent advice: think more about the global situation during games, and review carefully afterward. But ideally I would like to be able to learn by drilling this stuff. Do you know any source of tsumego that focus on navigating whole-board fights?
In terms of fighting and opening and midgame, I see in those two games that you played based on influence a lot – it feels fine in the opening to me (relatively) before your opponent invasion/reductions and the fighting begins, but it feels like the falling behind in the midgame was due to using the influence inefficiently and letting the opponent build light groups, or run through your potential.
How about looking over some pro games and/or high dan games more often to get a better sense of where they tenuki or play differently ?
I do indeed review pro games, and it certainly helps my game overall, but I find it to be a difficult way to learn the sort of global decision making that we’re talking about here. I think the reason is that there are subtle issues of timing involved in using influence correctly - sometimes pros will wait longer than I expect to start an attack, or invade sooner than I expected, or stop an attack even though the opponent’s stones still look weak. Often I can sort of figure out the thinking behind those choices, but I find it hard to tell whether my intuition was wrong or just different stylistically.
I suspect that if there is some issue with understanding midgame/opening positional judgement, that could help, or being exposed to stronger players discussing/commenting their games and how they think about global positions.
Yes, this does help - sometimes I watch streamers like dwyrin, clossius, and zchen talk about their games, and I have watched several of Nick Sibicky’s YT videos. This has helped some things gel in my head; for instance, I learned from watching dwyrin that when you exchange influence for territory you should count how much territory you gave up to gauge how much compensation you need when using the influence to attack or make territory. So I’m going to keep this up. Are there any other content creators that you recommend, beyond the ones already mentioned?
(especially if your main experience/speciality isn’t on the 19x19 and you might not have been exposed as much to how other dan players think – it may be less nebulous and easier to understand when reviewing with a human who can discuss and explain the concepts too, as compared to the AI )
This is good advice - somehow or another I probably need to find a way to get people to review my games.
I’m curious about this, too. Some of the whole-board problems I have done were from a friend in training/being taught just under insei level, and their school/dojo had whole-board problems of all sorts, along with ratings that came with how you played, e.g. in response to a fuseki problem (and I think, discussion with their pro teachers about the problem).
There were some very interesting attacking direction of play problems there too. Others, I have seen from some pros. Or from problems created for discussion by other strong players in our club/from questions asked about games played.
Interestingly, I don’t recall seeing or running into this sort of problem as often myself, save for when others post them from old Go World or journal issues.
(although I do see a lot of life-and-death and tesuji problems)
I will ask and try to find some available collections.
That makes sense, I was thinking just after writing this reply that higher dan players might be easier to understand than pros, as the play is pretty hard to see as ‘visibly’ as it gets stronger.
But I do recall that seeing even mid to high dans fight on KGS before, and the shapes/tactics I had seen once from playing a 6d KGS in simul, when I was still in the DDK/SDK threshold, had helped me later on with that (probably more so than pro games for this particular aspect of things).
Cool ! I actually learnt before content and streaming became very popular, so I’m not as familiar with it, having restarted Go after a long break of nearly a decade, recently in mid 2022.
I’ve watched and enjoyed some videos by Yeonwoo and Eunkyo Do and Takumi Go Academy, the latter of which a lot of players say is great – although not necessarily specifically on the topic of midgame/opening positional judgement.
I know of other great teaching resources on influence, using your thickness/strong groups etc., but they are in French (some Masterclasses by Stones In The Shell/Motoki Noguchi and various others/commentaries), so unless you can understand French too they would probably not be so helpful.
I always find this great, even with players who are at a weaker rank, I find I learn a lot from the discussion and interaction, and clarifying my own view through it.
Regardless of the strength, they can see things that I didn’t, or I learn from how they consider the board, or something I missed. ^^
It’s nice to initiate a discussion/review a bit after the game if your opponent is willing, too.
(My experience is that if I review and share a bit of my review/analysis/thoughts after the game in the chat/comments, often opponents will at least engage a bit (or come back to review/comment it too after they see the comments), and sometimes I have great reviews with some players who like to review together too after correspondence games, or they leave interesting commentary/review for me in return in future games.)
Reviewing with other players, if you can find good partners to play and/or review with, has been something I’ve always enjoyed and learnt a lot from. ^^
I have started to study this book twice (once in 2013, once in 2024 (still on my bedside table… maybe I’ll get back to it tonight?)), but both times have been waylaid by non-Go life before getting through all the pro game reviews and on to the whole-board problems.
Ooh, I’ve never read that one, I’ll have to give it a try ! ^^
And thanks to your post, I just remembered that some books in that series, like “Fight Like a Pro* – The Secrets of *Kiai” have nice whole-board direction of play problems. I’d just begun reading that recently and enjoyed it so far.
Edit : There seems to also be “Attacking and Defending Weak Groups Mastering the Basics, Volume 12” in that series, though I’ve yet to read it myself.
It seems to have problems in it too, though, and from the description, might be relevant to your positional judgement, especially if you like to play based on influence + fighting, like in the 2 games you posted. ^^
(I’ve done some whole-board problems on attacking/running fight/defending groups recently which really helped me with my own direction of play whole-board judgements, and I have a style which involves attacking, and sometimes a fair bit of influence, too.)
When evaluating a position, besides determining the balance of territories, an important consideration is the identification of weak groups. Positions often arise in professional games where one side seems to have secured a sizable amount of iron-clad territory, while the other side has little or even no area of the board that he can count on as territory. However, if the side with all the territory has a weak group, the other side can rectify this territorial imbalance by attacking that group. The purpose of the attack is not to capture the weak group, but to harass it and, in the process, build influence that will negatively affect the opponent’s groups elsewhere on the board. Even when the territorial balance is relatively even, one side can gain an advantage by attacking a weak group. On the other hand, failure to reinforce a weak group can result in the disruption of the territorial balance.
This book covers all the techniques of attacking and defending weak groups. Each of the nine chapters starts with a few examples of the technique under study, then continues with a few problems showing how that particular technique was used in a professional game. The tenth chapter presents additional problems whose solutions draw upon the techniques studied in the preceding nine chapters.
)
