It seems to happen way too often that my opponents get good moyos that seem impervious to attack, and whatever I try I don't get such moyos. I went through this game and the two big mistakes Leela pointed out where the S3 atari that should have been at R2, and the J3 atari that should have been at J5. Other than that, I just seemed to fall slowly behind. Needless to say, I don't understand how or why. By the time I invaded on the left, I had very little chance to win anyway.

The engines don't seem to be crazy about the Kobayashi, thus -- not a game-over move, but may be a candidate to "fall slowly behind".
( ditto for . )

.. Access to Leela can tell if this is bad for .

This move doesn't look right.
vs. Q18 -- what does Leela say ?
About Q18:
- isn't "trapped in the corner"
- does Q18 means have a good winrate ?

Q4 seems natural -- Leela ?
"I figured it would be easier to get an eye on the side."
Easier for or ?
-- if for , strange; no reason to make it easier for to make eyes;
-- if for , also strange; why think about getting eyes for when it is who has to struggle to live here ?

"felt I had to escape to the center..."
Something is vastly off in the evaluation:
why does a live group need to "escape", anywhere ?

o10 seems a natural candidate.

P9 seems natural -- Leela ?

the shapes and variations here are so common IMO the local lessons are worth this game.Leela: J5 -- both Q2 and R8 groups are not alive.

Leela -- take ?

does Leela have a variation where kills 's corner ?

start with B5 hane -- fundamental shape.

A few comments.

I don't know about Leela, but AlphaGo has taught us not to play this hane. Simply extend once more on the left side to secure your group.

White can ignore this extension, which is small, anyway. Just play Q-18 to secure your corner and threaten to invade the top side.

Good.

Fear of monsters.

Surely a solid connection is better, eh?

Edit: Note to Ed. P-09 looks good to me, but Fedya is not a fighter.

I don't know about big trouble, but you have wasted a lot of moves on the right side.

The hane at B-05 is fine.

Maybe so, but the left side and center look big to me. Maybe G-08?

Why not extend? The C-06 stone is not a ladder breaker.

Huh?

Somehow?

Desperation. At least Leela's moves threaten something. And at your level, overcoming an 38% chance of winning (by Leela's reckoning) is not unusual. It's probably more like a 42% or even 45% chance, unless you let your pessimism get the best of you.

Edit:
You do not have to save all of your stones. You can give up your two stones on the 12th rank. Sacrifice your stone on E-09. Atari twice, starting at G-10, and then make a solid connection at G-11. Your stones will be hard to kill.

Edit 2:
Black 119: Play at H-09 and save your stones.

_________________
The Adkins Principle:
At some point, doesn't thinking have to go on?
— Winona Adkins

Visualize whirled peas.

Everything with love. Stay safe.

One of the fun things about Go problems is that your opponent always has a weakness.

_________________
be immersed

I would disagree and say that both players were more or less equal regarding the direction of play, but the game was rather decided by some important tactical details.

One example is the atari in move 33. In the game, white establishes a living group which is splitting your stones and which later made an invasion on H3 (Move 46) "troublesome" for you. The Leela variation (with P2 in move 35!) would let you capture the corner and leave behind three eyeless w stones around O3. In terms of a chess game this would be like a mistake as "I could have captured his rook but instead let him capture mine" (in terms of Go I would rate this as a swing of at least 40+ points).

Also in your analysis, you didn't comment much on the suggested continuation P2 in move , did you really analyze this situation in detail? If have a few more questions on that position (Edit: For both questions you may assume the existence of the white wall around D16, which is not shown in the diagrams...).

Q1:

Q2:



For the other atari at I have to admit that I am unsure in such situations by myself. The problem is, that w didn't play the best move at and thus, the reason for playing J5 instead J3 on move maybe not yet obvious for everyone (it wasn't for me until I looked more in detail into Leela's variations). Did you analyze what would happen in case w played the best continuation?

Another big tactical miss is move 119, where you could have set up a working ladder, but Bill already beat me to it.

Lee Sedol 9p recently beat Shi Yue 9p from a position LZ 157 gave him 15% and Elf v1 gave 2%

Answering several of the questions in rough order of when they were posted:

For , Leela is spending a lot more time looking at Q8 and F5 than Q18. Everything is in the 51-53% winrate, though.

For , I was worried about getting eyes for myself.

Yes, Leela spent almost all its time on P9. I was worried about getting cut at P7.



I thought I did fight in this game.

I assume for you wanted me to block in the other direction at Q17?

Click Here To Show Diagram Code

[go]$$Bcm35 And does it still work, if there is a white stone instead?
$$ . . . . . . . . . . . . . . . |
$$ . . . . . . . . . . . . X . . |
$$ . . . . . . . . . . . X . . . |
$$ . . . . . X . . . . O O X . . |
$$ . X . . . . . O . O . X O . . |
$$ . . . . . . . . . . . . b . . |
$$ . . . . . . . . . . . . . . . |
$$ ------------------------------[/go]

I assume the point of a black stone at M3 is to allow a peep at N3:

Click Here To Show Diagram Code

[go]$$Bcm33 And does it still work, if there is a white stone instead?
$$ . . . . . . . . . . . . . . . |
$$ . . . . . . . . . . . . X . . |
$$ . . . . . . . . . . . X . . . |
$$ . . . . . X . . . . O O X . . |
$$ . X . . . . . X 7 O . X O 2 . |
$$ . . . . . . . . 6 5 3 . 1 4 . |
$$ . . . . . . . . . . . . . . . |
$$ ------------------------------[/go]

Edit: Leela suggests White not turn at S2, but jump out to O6 instead, with a lot of complex fighting.

Top pros thought certain things, too, until the engine showed them what's what.

Here are my variations on Q1 and Q2. If they are wrong/incomplete I would be glad about corrections.

You were trying to decide between Q-18 and R-12. Q-18 looks rather better to me. I would not have considered R-12. I compared the two. BTW, if you had played Q-18, your invasion at M-17 would have worked out better.



Fedya, mon vieux, it's true that I do not regard you as a fighter. I am not a fighter, either. But I was when I was a DDK. I lost many groups that way. I lost more than one group per game, if not in every game, at least on average. You don't have to be a fighter to enjoy the game, and you don't have to become a fighter to improve. Please tell me, which moves of yours in this game do you consider fighting moves? Thanks.

_________________
The Adkins Principle:
At some point, doesn't thinking have to go on?
— Winona Adkins

Visualize whirled peas.

Everything with love. Stay safe.

The first ones that come to mind are M3, P8, M17, and C10.

P8 is like a pass. You can play that even with the white stone not there and I would play the white stone in the same place anyway.

Thank you. Two pincers and two invasions.

Let me add the slide, , which undercuts White's potential eye space on the right side, and the nose jab, Black 115, a very nice tesuji in the center. Too bad you missed the chance to save your stones with Black 119. Next time.

After you say, "Somehow my opponents get a better result when they invade {than} when I do, despite picking the same invasion point." I think that you are referring to your opponent's invasion at H-03 versus your invasion at M-17. But at first I thought you were referring to his invasion at R-17 and your invasion at C-03. Well, for those invasions one obvious reason he did better is that he kept your C-03 stone separated while you allowed his R-17 stone to connect.

Please go back and look at the board after . If you were White and it was your turn, where would you want to play and why? I'll hide my guess.



In this game, and others, you imagine that your opponent's stones are stronger than they are and that your stones are weaker than they are. Maybe taking a look at the board from your opponent's perspective would help to cure that.

Counting the number of stones may help to give some perspective. After your pincer in the bottom right, Black has three or four stones locally (depending upon what you consider to be local) to White's one stone. After White crosscuts, you decide to sacrifice one Black stone, forcing White to live. Then you decide to escape to the center. Does that make sense to you? You started off with more stones than White, you got three forcing plays for your sacrifice, yet you're in trouble? Yes, you might have played better, but you are good enough that that should not happen. And it didn't. Your corner is alive.

With P-08 you do "escape" to the center. And again, you outnumber White locally. But what happens? White makes shape and you take gote. That should not happen. Yes, if you make the extension to P-09, attacking White's weakness, there is a cut and bad things could happen. But your stones have escaped. Whatever happens, don't let White take the initiative. OK, you decide to protect against the cut, and let White make shape with P-09. Now play somewhere else. Don't take gote. It doesn't make sense.

I'm not sure what you were thinking in the top right corner. I suppose that you were worried about your three stones on the top side, and wanted to save them over anything else. But again, once you sacrificed the stone on the top side, you outnumber White in the corner. Surely it cannot be right to let White make a big corner. In a similar situation in the bottom left corner, White did not let you make a big corner, did he? You worry too much about the safety of your own stones.

_________________
The Adkins Principle:
At some point, doesn't thinking have to go on?
— Winona Adkins

Visualize whirled peas.

Everything with love. Stay safe.