What do I have to do just to get to 5k?

[sparky314]

It gets easier with more practice, but...

Even pros miss ataris: https://youtu.be/noCZ5gwRBCY?t=58m54s
(Can't seem to embed youtube to start at a given time? Am I missing something?)

And that isn't even reading a future atari!

[EdLee]

Hi sparky, a nice feature request.

Testing:
[youtube]noCZ5gwRBCY?t=58m54s[/youtube]

In this particular case, Mr. Shida probably read ahead too much:
he didn't miss the atari --
in his mind, probably he had already replied to the atari, in sente,
so he was planning his follow-up sequence;
instead he played it directly... -- it happens; we're human.

What I really like is Mr. Ishida picked up in one scoop four stones:

[Fedya]

Today I may have had the first game in quite some time where i may have just flat out played badly. I blundered with , and compounded the blunder by trying to same all my stones. (Not that sacrificing them seemed like a very good idea either.) But in my other recent losses, I thought back and realized I had played well enough to get into winning positions before blundering. This time I'm not so certain. I was thinking for a while that if I just connected on 66 I would have had an easy victory, but then I looked even closer and have the sneaking suspicion that I was actually a good ways behind after . I'm not at all certain how I should have handled my opponent's moyos, but think in hindsight that the right side attempt was particularly a mistake.

My comments stop after . I did eventually save those three stones which became a whole bunch more, and resigned when I screwed something up in the lower left not long before the game would have gone to counting. By that point, however, I was well behind with no way to make things up. I didn't realize until after the game just how behind I was -- the thing is I didn't have much time to count during the game since I was spending all my time trying to work out how to save my stones.

[DrStraw]

is playable but jumping one further is better.
Attach at R3. bR2, wS3, bQ4, wQ5 should follow.
this is one point too far. It leave L3 behind and does not have a follow up extension. When the opponent makes a move like this you should try to steer the game so that the aji of that invasion becomes a problem or B.
Wrong side. Approach from the other side and then make a position at the top when B replies with this move.
Can still consider O17
Pushing again on the 3rd line is not good. Just hane then you can jump after B connects.
Why don't you want B to play there? It hold no interest at all for B as he cannot do much with it if you reply at N17.
Or you could ignore it altogether and play in the center somewhere.
Way too deep. You cannot expect to connect below. B should prevent that now and the lower group is cut off and eyeless. You should reduce from above and then aim at S8.
You should not have been allowed to get away with that. B's attachment was bad. Attachments are for defense, not attack.
way too close to the black position. Somewhere around M10 looks more intersting.
I don't understand this. Had he blocked you would have had trouble living.
This doesn't do much. How about J5, aiming that that aji and also the center?
E7 would connect in sente.

[Shaddy]

68 @ H13

[Fedya]

For a move like , I know there are several josekis possible, but I don't have any idea why your suggestion would be better than what I played.

And as for , I've always thought that if I played there, Black would respond at Q4 and then split at either R4 or Q5.

For , I thought if Black played there he'd be able to jump in either direction and leave me with two pretty weak groups.

[skydyr]

For a move like , I know there are several josekis possible, but I don't have any idea why your suggestion would be better than what I played.

For , I thought if Black played there he'd be able to jump in either direction and leave me with two pretty weak groups.

For white generally plays a jump (often the two space) when white intends to build strength to counterpincer the pincering stone. Playing a knight's move to lean on the corner stone is a similar strategy. With the black corner stone high in the upper left, however, that will have some influence on the fighting. If you want to just settle fast (not recommended unless black is strong all around) you can attach at R3. You can play the diagonal move at Q6 to keep black separated and possibly start a running fight. You can counterpincer from the left side at various different points, and most of the other jumps have some reasonable lines that come from them.

The two space high pincer to the low approach was one of the common joseki beginnings in the mid-20th century, and so there was a ton of research done on it, if you want to look in a joseki dictionary. It will probably be the one with the most variations. The key, though, is not to just try and run out, but to think of what goal you'd like to accomplish (I want to attack this stone, or I want to get influence in this direction, etc.) and then try to think about moves and a line of play that could help you accomplish it, because there is so much flexibility in this situation.

For , if you play at N17, and black jumps in to the top side, first, black can't make a two space extension in either direction. Second, N16 is a big move for white that has a side effect of hurting black's stone, possibly in sente. Third, white's groups on both sides are strong, so black's stone makes an easy target for attack. What would black do if you just capped? Black might live inside, but white's outside strength would make black's moyo on the right quickly disappear, and likely keep the left from growing unmanageable too.

[Fedya]

I guess I have to do more studying. I still have difficulty seeing what makes some of the joseki so much better than others.

And I know you and Bill Spight have talked quite a bit about capping moves, but I still find it hard to see how they (and other reducing moves like the shoulder hit) work well in practice.

[jeromie]

The recently published Encyclopedia of Go Principles has a short section on using the capping move that I thought was pretty helpful. It's a very easy read; while not as in depth as many other books, it provides a useful overview of many important ideas.

[Fedya]

Nobody said it was going to be easy.

After a long series of wins, I've had a number of reverses in succession, and KGS has demoted me back to 7k. It's been yet more games where I've had to fight, finally get a good position out of the fighting, and then make some mistakes that cost me an entire group.

What's probably more frustrating is that I have to engage in the fighting in the first place. Now, I'm not necessarily opposed to fighting; I find the sort of style I've faced in my limited forays onto the Korean servers to be bracing, like a cold shower. It's as though they say, "Lets fight with , because, well, why not? (OK, maybe it's not that severe.) But the sort of fighting I'm talking about isn't that. Instead, it's that I keep falling behind in the beginning and get to the point where fighting is the only attempt to get out of the holes I dig for myself. And I still can't figure out where I'm going wrong early on.

In this particular game, I spent a lot of time trying to figure out where my opponent's weaknesses were, and not having any idea. And again, I had weaknesses I didn't know about that my opponent had no difficulty exploiting, leading to my trailing badly and having to fight.

[dfan]

Just looking at the first 50 moves or so (take all opinions with a grain of salt, given that I am only a few stones stronger):

The first move I have any real quibble with is (though I probably would have played one line lower because as played it can easily be dug under from the right). My immediate instinct is that White has to hane; the only question is which direction. After ten seconds I would probably flip a coin. K3 isn't even necessarily more of an attacking move than H3, since you're likely to sacrifice it to make a nice solid position if he crosscuts.

Don't be worried about the idea of a crosscut! If there is some concrete variation to be worried about, okay, but you can't just be worried in the abstract about all local fighting.

looks wrong. A stronger player could tell us what move would be better (I would probably either play E17 or just tenuki, and ladder or squeeze if he cut), but the played move makes the stone at G17 look stupid.

One move later: if had been at E17, you could play at B14 with very few worries, so that's one thing.

seems pretty optimistic given that Black has support at C7. I think R18 is the biggest point on the board, though stronger players may disagree. I would probably tenuki with rather than save two stones.

: don't fear crosscuts in the abstract! You are locally strong, fighting will be fun.

: unless he threatens to connect (I haven't read it out, but I don't think he is), I think it is more important to cover him with Q10, denying him liberties and keeping him confined to the side. Turning at Q10 is big for him.

[skydyr]

Here are some comments on the game:


In general, there are a couple deficiencies I can think of:

First, white tended to build thickness without a plan for how to use it, but just because it was something to do, and to play overly defensive moves to keep everything connected without considering what might have an effect on black.

Second, when you are white and black goes straight for territory and quite tightly, as he does in this game, it's fine to compete on territory instead of trying to go for the outside. Komi should be fine to keep you in the game if not win it.

[Fedya]

Comments to skydyr:

I think I avoided the possibility of a diagonal fuseki mostly because I really feel like I don't know what I'm groping blindly in it. I've said something to the effect regarding opponents playing moves in a joseki that I don't realize that in such cases I'm reduced to playing from first prinicples, and I find it hard to come up with good moves that way. With the diagonal fuseki, as opposed to playing (or playing against) the Chinese, sanrensei, Kobayashi, or what not, I really feel quickly reduced to playing from first principles.

Part of approaching L17 from the left instead of the right is that I have more strength on the right since I've already approached Black down in the lower right. After and , I thought I had a reasonable result. It was only later that things started to go wrong, I felt.

and I've had this vague sense that C5 is the point to be aiming at, but I have to admit I don't know what to do with it once I've played there. That's probably why I played a suboptimal line after .

I wouldn't know which stones to sacrifice.

I implied, but didn't say outright with my comment, that none of the several moves I was considering looked particularly better than any of the others. And this is a move I actually spent a fair amount of time on, at least by the standards of online go, where most people seem to play blitz no matter the time settings.

I played F17 since I thought it was the right move for the endgame: when Black tries E18, then F19 is sente. Plus, as you said, the F18 clamp is a pain, and F17 prevents that.

Yes, I was already worried about my groups having been split. was played to prevent Black from playing under my stones and connecting. (See Black's later play of F2.)

I may have more later, as well as a reply to dfan.

[mitsun]

I have a few lengthy comments on three particular mistakes, hopefully with some guidance on how to think about similar positions in the future.

Strategic error #1

You should know that after the double-wing extension to C13, B can and should invade at the 3-3 point and live inside. If W wants to keep the corner, C15 is the maximum extension, while C16 is safer (leaving less aji); both are good moves. The question then is whether a wider extension and subsequent B invasion can give W a better result? The difference in corner territory will over 20 points. So can W make 20+ points of useful outside thickness by confining B to a small life? On an empty board that might be possible, but here the B stones at C10 and L17 make that unlikely. After the corner is settled, W will not be able to make an efficient extension in either direction, and there will be no way to use the W thickness to attack any weak B stones. The conclusion is that the extension to C13 is wrong; it is better to secure the corner territory.

Shape error

C6 is not a good shape move. The shape move is to draw back to D5 in sente, letting B connect underneath. W can be satisfied with this exchange and switch to the next large point. The problem with C6 is that by plastering these stones tightly against the B formation, W leaves behind a cutting point and shortens his own liberties. D8 has exactly the same problem. The result here is not good for W, certainly not worth giving up sente. You lost nearly a full move here.

Strategic error #2

After B invades at R9, W needs to come up with a strategy. Playing softly on top at Q10, letting B get a comfortable life along the side, would be reasonable if W could then secure a large center territory. This seems a little unlikely, but if you made that your strategy and carried through consistently, it could be a reasonable game plan. The other choice is to block B from making easy life on the side and find a way to profit from an attack. You apparently went for this plan, but you blocked on the wrong side.

Here are a few ways to decide which side to block, or equivalently which direction to attack.

1) Block B from moving into the larger open space. This is partly because he will find it easier to make a base in that direction, and partly because you would like it to become your territory. Where else are you going to profit from this attack?
2) Block on the side that forces B to run into your existing thickness. Here the two stones at R5-R6 are more of an immediate threat to B (if you block at R10) than are the two distant stones at R16-Q14 (if you block at R8).
3) Consider the possibility of a leaning attack, as a means of strengthening your position in the direction where B will be running. On the bottom, you have several really strong forcing moves like P4 and N4. This means that the Q5 stones are really much stronger than they may appear. On the top, you have only O14, and that move is not very severe.

[Fedya]

First, white tended to build thickness without a plan for how to use it, but just because it was something to do, and to play overly defensive moves to keep everything connected without considering what might have an effect on black.

Well, to be honest, I don't know what to do with thickness. Technically, I do know that strong players say not to use it as one part of a wall surrounding territory, but to use it to attack elsewhere and profit from the attack someplace else on the board. The thing is, I just don't see how this ever works in practice. At least, it never seems to work in my games.

So why did I allow my opponent to take territory and give me thickness? For a long time, I had the attitude that thickness never worked for me despite what strong players say about it being OK, and felt like getting thickness in exchange for territory was a recipe for disaster. No wonder I could never get anything good out of thickness! But if I want to get better I'm going to have to figure out what to do with it in those situations when I can't avoid it (or when letting my opponent get the small territory is the right thing to do).