Life In 19x19

In my "road to 3 dan" I'm psyching myself up and have started playing 3 dan opponents. While solidifying the 2d rank - which is still not achieved - I have limited myself to 2d and 1d. Against 2d I'm probably about 50-50 (edit: that's right, 7-6 in the past months on OGS). Against 1d I would almost claim to have an easy game (edit: 9-2), if I'm careful enough not to make mistakes way below my level of understanding, but against 3 dans the mission seems almost impossible today (0-3). Here I feel like I have to consistently play on my level of understanding and frequently play beyond that. Even when pulling off a tactical victory in a local battle, opponents are very good at controlling the damage. Usually it is me who's on the receiving end of tactical superiority with lesser capability to dodge.

Those few occasions where I outsmart a 3d do give hope. It's a matter of consistently play at that level and also see these opportunities early enough so as to take any alternative plans into account. The level of Go played at 3d is just a higher standard - and that's something to enjoy.

When you've been enjoying a period of success against weaker players, it takes courage to face stronger opponents more regularly. Kudos to you!

Knotwilg wrote:...if I'm careful enough not to make mistakes way below my level of understanding...

This is huge, and I honestly think it's the most important next step in my own development. Weaknesses still abound in my game, but I think I (and most kyu-level players, at least) would gain at least one stone by simply not making mistakes that are way below my level.

jeromie wrote:This is huge, and I honestly think it's the most important next step in my own development. Weaknesses still abound in my game, but I think I (and most kyu-level players, at least) would gain at least one stone by simply not making mistakes that are way below my level.

Yes. When I go over a game of someone who is, say, 10k, I concentrate much less on "here's where you could have played like a 5k" moments and much more on the "here's where you played like a 15k" moments. For one thing, fixing these mistakes will have a greater effect than other changes (it only takes one blunder to lose a game; playing a great move doesn't guarantee you'll know how to follow it up), and for another, it's usually easier to do than acquiring new skills.

Of course you can't improve forever this way, but I do think there is almost always low-hanging fruit to be had.

dfan wrote:

jeromie wrote:This is huge, and I honestly think it's the most important next step in my own development. Weaknesses still abound in my game, but I think I (and most kyu-level players, at least) would gain at least one stone by simply not making mistakes that are way below my level.

Yes. When I go over a game of someone who is, say, 10k, I concentrate much less on "here's where you could have played like a 5k" moments and much more on the "here's where you played like a 15k" moments. For one thing, fixing these mistakes will have a greater effect than other changes (it only takes one blunder to lose a game; playing a great move doesn't guarantee you'll know how to follow it up), and for another, it's usually easier to do than acquiring new skills.

Of course you can't improve forever this way, but I do think there is almost always low-hanging fruit to be had.

This partly came out of my "2021 mistakes" and "2022 mistakes" series, trying to find patterns in big mistakes. Some of these pointed to missing information, like "avoid slow connections" or "surround on a large scale (but check if it is in sente, otherwise consider tenuki)" ... Some other mistakes were plain to see and a result of not thinking or having a momentary lapse of reason. The question is then: WHY am I making these stupid mistakes. And that mostly comes down to what I mistakenly called "gamesmanship" and consists of 1) fighting spirit 2) proper time management 3) checking liberties and overall be disciplined in actually selecting candidates and reading a few move deeps to choose the best one. And an important prerequisite is to play games when I'm fit, as a deliberate act and not just a time killer or a late night closure of the day (which is what I often do).

So, improving at Go needs sacrifice of the best time I have, not the killer time. And that also means I'm voiding myself of an easy excuse for losing: it wasn't serious anyway.

While on KGS I can hold my own against a 4d - perhaps incidental - and hold a stable 2d rank, flirting with 3d, on OGS I'm dropping down, losing badly these days against 1d.

OGS is interesting that way. I think there is more variance in strength for a given rank.

But how are the mistakes in the losses to 1d? (Hint: you could post a game). I think focussing too much on rank can be a distraction from making progress in learning.

dust wrote:But how are the mistakes in the losses to 1d? (Hint: you could post a game). I think focussing too much on rank can be a distraction from making progress in learning.

When I get back to my home laptop (travelling) I'll make some analyses. You are right and for 1,5 year I have been training and reviewing without caring about rank. Now I wanted to get proof points that I'm improving. I am solving problems I couldn't before and I have had a "feeling" I understand more. But still, there's a yummy pudding in front of me, which is playing and possibly beating higher ranked players. Eating it, that kind of proof is not happening to the extent I hoped, as noted above. Intrinsic motivation only goes midway for me.

From the OGS interface analysis, moves 48 to 54 were all mistakes, committing suicide to my invasion. Not a mistake, but a choice leading to the potential of making these mistakes, was the reinforcement in the lower right, allowing Black to build up his moyo. Invasions are definitely a weak point, which is also disappointing given the amount of tsumego I do.

Thanks for posting the game . I think this is just one to put behind you and not read too much into it.

It feels like you got persuaded against better judgement into one of those horrible risk management situations on the left side (with 42 and following) where the risk is all on your side and the potential loss is huge. Would be interested in other views though.

I don't know what AI says, but to me, 32 was the first move that struck me as odd. Black is so strong on the left side both above and below that giving way with a one point jump to E11 seems much more natural than digging in, and also more consistent with the keima before.

Harleqin wrote:I don't know what AI says, but to me, 32 was the first move that struck me as odd. Black is so strong on the left side both above and below that giving way with a one point jump to E11 seems much more natural than digging in, and also more consistent with the keima before.

AI finds a way to live. So it doesn't scorn this move.

For me it was too tough, like dust said and I had multiple chances to take it easy: hane at 26 not play in the corner, reduce at 28 not invade, one space jump at 30 not keima, sacrifice at 32 not fight ...

In this game my opponent didn't give me a hard time, since he was overplaying all the time and I only needed to be mindful of my own groups to see his fall into my arms as ripe apples.

Still I want to highlight some moves that came intuitive to me but weren't on KataGo's radar at all or the other way around.

50 seems odd / inconsistent with the atari you played. If you played the other atari then at least you are threatening p13 (ladder doesn't work, but with p15 peep maybe something does, though black can always give up 1 stone if it does) or o12. Whereas the atari you played aims at m17 sheanigans, but means you have less good follow ups on outside so n12 is even worse than if you didn't do that atari.

I did a more lengthy analysis with KG on move 48-50

Click Here To Show Diagram Code

[go]$$ Position at move 47
$$ ---------------------------------------
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . X O . . |
$$ | . . a . X O . . O . X . . . . X O . . |
$$ | . . . X . O . . . , . . . . . X O . . |
$$ | . . . . . . . . . O X X c . . . X O . |
$$ | . . . X . . . . . O O O X . . . X O . |
$$ | . X . . . . . . . . . . O X . . X O . |
$$ | . . O . . . . . . . . . d . . . X . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . , . . . . . , . . . . . , . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . X . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . O . . . . . . . . . . . . . O . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . O , X X . . . b . . . . . O . . . |
$$ | . . . O O X . . . X . . . X O . . . . |
$$ | . . . . . . . . . . . . . X X O . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ ---------------------------------------[/go]

With 39K and 26K playouts respectively, KG is ambivalent about A (+5,9) and B (+5,8).
With considerably less playouts (1,5K) the atari at C (+5,6) is only slightly worse while stretching at D (+4,6) is probably already a minor mistake.

Click Here To Show Diagram Code

[go]$$W Atari
$$ ---------------------------------------
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . X O . . |
$$ | . . a . X O . . O . X e . . . X O . . |
$$ | . . . X . O . . . , . . . . . X O . . |
$$ | . d . . . . . . . O X X 1 . . . X O . |
$$ | . . . X . . . . . O O O X 2 . . X O . |
$$ | . X . . . . . . . . . . O X . . X O . |
$$ | . . O . . . . . . . . . . . . . X . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . , . . . . . , . . . . . , . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . X . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . O . . . . . . . . . . . c . O . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . O , X X . . . b . . . . . O . . . |
$$ | . . . O O X . . . X . . . X O . . . . |
$$ | . . . . . . . . . . . . . X X O . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ ---------------------------------------[/go]

After the atari, C is the blue move (16K, +5,0) so the higher number of playouts reduces the value by 0,6. A, B, D and E are candidates to improve on C.

Click Here To Show Diagram Code

[go]$$W Atari & Stretch
$$ ---------------------------------------
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . 4 . . . . . . . . . X O . . |
$$ | . . a . X O . . O . X e . . . X O . . |
$$ | . . . X . O . . . , . . . . . X O . . |
$$ | . d . . . . . . . O X X 1 . . . X O . |
$$ | . . . X . . . . . O O O X 2 . . X O . |
$$ | . X . . . . . . . . . . O X . . X O . |
$$ | . . O . . . . . . . . . 3 . . . X . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . , . . . . . , . . . . . , . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . X . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . O . . . . . . . . . . . c . O . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . O , X X . . . b . . . . . O . . . |
$$ | . . . O O X . . . X . . . X O . . . . |
$$ | . . . . . . . . . . . . . X X O . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ ---------------------------------------[/go]

When White stretches, KG drops White's lead to 2,3 provided Black secures the corner with (16K)

Click Here To Show Diagram Code

[go]$$W Stretch first
$$ ---------------------------------------
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . 1 . . . . . . . . . X O . . |
$$ | . . a . X O . . O . X e . . . X O . . |
$$ | . . . X . O . . . , . . . . . X O . . |
$$ | . d . . . . . . . O X X . . . . X O . |
$$ | . . . X . . . . . O O O X . . . X O . |
$$ | . X . . . . . . . . . . O X . . X O . |
$$ | . . O . . . . . . . . . 1 . . . X . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . , . . . . . , . . . . . , . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . X . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . O . . . . . . . . . . . c . O . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . O , X X . . . b . . . . . O . . . |
$$ | . . . O O X . . . X . . . X O . . . . |
$$ | . . . . . . . . . . . . . X X O . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ ---------------------------------------[/go]

When White stretches, the lead drops to 4 pts with 30K playouts.

So overall, making central shape in gote loses 1,9 points. Making the bad atari before loses an additional 1,7 points.
The bad atari is probably a more stable value while losing gote depends largely on the available sente.

The most important goal for White is to play at A.

B is only recommended because White retains the privilege to play at A after playing B first.

A is an important move for White because it weakens the top left black group and helps to exploit this weakness for Whites advantage.

Strike while the iron is hot.