I’ve played a lot of 9x9 and recently started playing more 19x19.
It’s fairly simple to improve on 9x9 boards - 90% of the time it comes to doing more tsumego.
I’m feeling a bit stuck on 19x19 though. Not sure where to focus on improving.
I am also not a “complete” noob for 19x19 - I have taken classes and trained on 19x19 when I was young, and I watch a lot of review videos on 19x19 game plays, so I’m pretty familiar with the basic principles.
You’re totally right - This was a fast game and I didn’t think before Move 94. My instincts just told me my opponent would answer it, but the second I played it I knew that was stupid
Move 232: I was trying to be safe than sorry. Time wasn’t enough for me to see if my opponent has any good moves there, so i decided to just kill that 3 stones, making sure my weak group connects.
But I didn’t see the corner was KO not alive. This was me not being good enough at tsumegos for sure.
The ones I am thinking of I’m sure you could predict your opponent’s answer, but you didn’t correctly judge the resulting position as being worse for you than before it, i.e. you played bad sente moves or “thank you moves” (the opponent says thank you). Several peeps for example: d13, f13, k10.
Pretty sure I need to eliminate it from my game, but genuinely when to play a forcing more or when not to is quite difficult to judge.
It depends very heavily on the local and global situations, and even timing of the move, it can be good now and bad later and vice versa because of it.
For me, one of the most difficult transitions from 9x9 to 19x19 is having to consider the consequences of the local position on whole-board strategy - especially as it applies to the underlying efficiencies that come with the whole corners > sides > middle aspects of the game
As I was ranking up, I spent a LOT of games chasing local fights, only to realize that I handed my opponent 3 corners and left myself with 1 corner and a few inside moyos
YouTuber Dwyrin has some good episodes of the BASICS PLAY series where he talks about zooming out whenever you get sente and seeing where the next most valuable spots are for that stage of the game. Dwyrin usually phrases that as
are you chasing / fighting / killing?
are you building?
Because if you’re spending the entire game chasing and killing, chances are, someone else is reaping the benefits of the building that happens whether you realize it or not. If you’re not aware of Direction of Play - chances are you’re benefiting your opponent.
I guess, in order to make sense of that broad whole-board strategy, one needs to be aware of the Stages of the Game (i.e Opening, Early Midgame, Late Midgame, Endgame, etc) and the strategic priorities for that stage.
I’ve done a bit of writing here if you’re curious. I know I’m just an 8kyu, but skim it and see if some of the pointers there may be useful to you - especially those in games 2 and 3 in that essay
Moving from 9x9 to 19x19, I’d say that is the most likely - and most easily improved - candidate, along with improving on how to handle invasions and how to build and use influence (which, as a concept is almost totally absent from 9x9).
Move 94 is a good example of the “a single point matters” attitude/mindset which one might have in a 9x9, but it should be absent in the middle game of a 19x19 game, until way into the endgame.
The black lines are where your wall is solid.
The arrows are the direction where your wall creates influence.
The red lines is where your wall is thin or your influence is not really strong.
The bluish box is where the opponent’s shape is not totally connected.
Not by accident that is where all the AI suggestions are located, aiming to improve or capitalise on the aforementioned assessments. I’d probably go for the shoulder hit at L4.
But let’s stick to the suggestions. For example, P5 seems almost sente, so you can get that for free (and then pivot to k4/k5) or gain a valuable push.
Let’s follow the AI’s very sensible suggestion on this move and re-evaluate:
The reddish squares seems like solid points now.
The red lines are no longer very important weaknesses because now your main goal is to connect k5 with any of those white stones around it (e.g. with k7) and then all the yellow circle is “garbage area” with mostly no points for both players. Black’s own influence is mostly not worth much.
Now, you want sente?
n16 (the potential atari at q16 might cause some confusion, but it will definitely lead to a response) or q12 (ignoring this might lead to severe cuts).
You want to win some points and give sente?
s11 or h7/h6 or defend at t6. Those might seem “small”, but they are not as small as the atari of move 94.
About forcing moves predilection:
That’s going against the usefulness of multiple choices. Most forcing moves are bad when you are not really committed in a global analysed plan and they are part of it. You are depriving yourself of different options for later, which is more hurting you as what you think you gained. At this point, even when you think it’s the right move to force, the timing has still to be well chosen which is not so easy to decide.
This was really enlightening. I never thought those moves were bad.
My thinking when I played those peeps:
A principle I learned is attacking is not necessarily to kill, but to benefit somewhere (strengthening yourself or making territory). Peeping at d13 and f13 feels like I am benefiting by making the top left group stronger and potentially make some territory, while also destroying any potential black has to make eyes in the area, so I could keep attacking black while making territory in the lower left.
Without those peeps, I felt like the top left white group could be counter attacked and black could potentially make eyes there.
Now I see the AI reviews, those peeps are definitely losing moves. I can see I am helping black strengthening its connections in a way, but I didn’t think it was a big deal as black is pretty connected with one space jumps anyways.
Can you share how you’re able to tell those moves would lead to worse positions for white? Seems this could be an area I really need to improve on. Thanks again!
That’s a good idea. I also tried to think this way and seems to still be wrong a lot of times.
For example, as @Uberdude mentioned d13, f13, and k10 are bad sentes. I thought about it and wasn’t going to try to cut there anyways since I thought one space jumps are pretty well connected (I could be totally wrong), and those sentes kinda help me strengthen my own white groups and territory, so I thought they were good forcing moves.
Thanks! Really good description of the differences there and I totally agree.
I actually started from 19x19 and had a fair amount of training (not on OGS) before moving to 9x9. I moved to 9x9 mostly because I was really bad at fights on 19x19 and thought 9x9 would be great exercise for fighting.
So I am not sure if this applies to my situation, but I agree this is definitely true for people who started on 9x9 and moved to 19x19.
Great advice. I should definitely play longer games and think more before I move.
This should be obvious but somehow I have this bad habit of moving too fast before reading clearly (probably a side effect of playing fast games). Definitely a major barrier for becoming better.
One advice I’ve read for this situation is “take your hand off the mouse while you think”. It means that after you click, you take your hand off your mouse until you’re ready to place the next stone…
