Filling in the gaps

[amatterof]

I've been a long-time lurker on this forum, but finally decided to sign up so I could create a study journal here. I'm hoping to have a New Year's resolution to be more diligent in my go studies, and this seemed the best way to keep myself honest about that goal, regardless of whether the resolution happens or sticks. (I thought about doing this as a blog, but it seemed that this way I'd have a better chance at feedback.)

Who am I?
Since I'm new to the forum, an introduction is probably in order. I started playing go in high school, and since then have mostly played online. In the past year, however, I've been fortunate enough to find a good local go club where I've enjoyed playing, as time permits.

I've developed a very territorial style of play, and have to resist the temptation to always take a few extra points when it means giving my opponent too much thickness.

Right now, my strength is generally as follows:

• AGA - weak 4d / strong 3d - In casual games, I usually play as a 4d, but my official AGA rating is still 3d
• KGS - weak 3d / strong 2d - My various accounts are either 2d or 3d, but even then I find myself losing to kyus sometimes.
• Tygem - weak 5d / strong 4d - My account tends to move back and forth between the two.

The problem
I've made it to my current level mostly by playing lots of games and watching stronger players play. Only recently have I started paying attention to more lectures and purchased a few books. The result is that I play mostly by intuition, and I feel that I make a lot of mistakes that are more characteristic of weaker ranked players. My hope is that by filling in the gaps in my knowledge, I can reduce the number/severity of those mistakes, and improve my play.

The goal
My goal is to pick up 1.5-2 stones in strength in the course of 2015. In other words, I'd like to make it to a solid AGA 5d / KGS 4d / Tygem 6d during next year.

The plan
I'm going to try to resist the temptation to always just play games, and instead use some of the time I devote to this hobby to studying instead. Whether it's doing tsumego, analyzing pro games, watching lectures, or reading go books, I'm hoping this will help fill in the gaps that I probably don't even know are there. I realize I should probably be more concrete, but I expect my free time during the year will vary considerably, so I would either significantly over- or under-promise with anything more specific. In any case, I'm hoping to record what I study here, both for feedback and in case it's useful to anyone else.

Thanks for following along. Any constructive feedback is greatly appreciated.

[amatterof]

One of the areas of my game where I feel I can improve is life&death. With enough time, I can usually read out all but the most complicated of problems, but "enough time" isn't always available. If I can analyze some of the more common shapes in advance, I can save this reading time during the game for other important things. Here is some basic work I did on the small-L group.

(Edited Dec. 17, 2014 to incorporate response to illluck's helpful comment.)

The basic shape is probably the first major life&death shape I remember learning as a kyu. The discussion below may seem rather elementary, but I think it's worth covering even these things as I work on making more of my intuitive knowledge explicit.

The basic shape is dead
As I remember learning back as a kyu, the basic shape is dead.

(;GM[1]FF[4]CA[UTF-8]SZ[19]AB[ob][oc][od][pd][qd][rd]AW[pb][pc][qc][rc]C[The basic shape is already dead. Regardless what White plays, Black can keep the group dead.];B[tt](;W[ra];B[sb];W[sc];B[pa](;W[qa];B[rb])(;W[rb];B[qa]))(;W[rb];B[qa];W[pa];B[sc];W[sb];B[ra]))

It is worth noting that White does have 3 ko threats here. That leads to an important corrollary: if White already has 1 or 2 in place, the shape becomes unsettled. That is, if White plays first, the group becomes alive, but Black can kill the group if he plays first, as shown above.

Touching the edge (part 1)
What if White's extra move is not inside the shape, but rather is a descent to the edge?

(;GM[1]FF[4]CA[UTF-8]SZ[19]C[With an extra stone at X, White's shape becomes unsettled. If White plays first at 1, she can live.]AB[ob][oc][od][pd][qd][rd]AW[pb][pc][qc][rc][sc]MA[sc];B[tt]C[White can live with plays at 1 or 2. Although 2 looks like it leads to a ko, White has enough outside liberties to prevent trouble.](;W[ra];B[rb];W[pa]LB[qb:A][sb:B]C[White makes miai out of A and B. However, as we'll see later, these are only miai because white has outside liberties.])(;W[rb];B[qa];W[pa];B[ra];W[sa];B[sb]SQ[sa];W[qb];B[sd];W[sa]))

Here, White's shape is unsettled. As shown above, White can live. However, with extra black stones on the outside, the situation can become more complicated, as we'll examine below.

When attempting to kill this shape, Black must be careful.

(;GM[1]FF[4]CA[UTF-8]SZ[19]AB[ob][oc][od][pd][qd][rd]AW[pb][pc][qc][rc][sc]C[To kill this shape, Black should first reduce White's eye space by playing at 1, rather than playing inside. Although playing inside still kills, it is somewhat vulgar.](;B[pa]C[By playing here, Black forces White to make the standard 5-space shape that is easily killed.];W[qa];B[rb])(;B[ra];W[rb];B[qa];W[pa];B[sa]C[In this variation, black creates a bent-four-in-the-corner shape. This is still dead in most scoring systems, but others require it to be played out.])(;B[rb];W[ra]C[If Black plays inside first, then playing at 1 next is a tempting mistake. Black must play at 2 to keep the white shape dead. This risk could be avoided by playing the proper initial move.] (;B[pa];W[sb]C[White lives.])(;B[sb];W[pa];B[qb])))

Touching the edge (part 2)
What if White's descent is on the other side?

(;GM[1]FF[4]CA[UTF-8]SZ[19]AB[ob][oc][od][pd][rd][qd]AW[pa][pb][pc][qc][rc]C[With an extra stone at this marked X, White's shape again becomes unsettled.];B[tt]C[White can live by playing at 1 or 2.](;W[ra];B[sb];W[sc];B[rb];W[qb]C[White lives. However, the importance of outside liberties is again becoming clear.])(;W[rb];B[ra];W[qa];B[sb];W[sc]C[White can also live this way.]))

Again, White's shape is unsettled, and White can live if she plays here first. If Black plays here first, he can of course kill.

(;GM[1]FF[4]CA[UTF-8]SZ[19]AB[ob][oc][od][pd][qd][rd]AW[pa][pb][pc][qc][rc]C[To kill, Black should again follow the technique of first reducing white's eye space from the outside. Although inside moves again can kill, Black must be careful.](;B[sc];W[sb];B[ra]C[White remains dead.])(;B[rb](;W[sb];B[ra])(;W[ra];B[sb];W[sc];B[qb]))(;B[ra];W[rb];B[sc];W[sb];B[qa]))

Supporting stones outside
One important comment should be made about the fact that, when White touches the edge, the shapes become unsettled. This means that, if White had supporting stones nearby, she could use the threat of connecting with them to make the basic shape live.

(;GM[1]FF[4]CA[UTF-8]SZ[19]AB[ob][oc][od][pd][qd][rd][eb][ec][ed][dd][bd][cd]AW[pb][pc][qc][rc][rf][qg][rh][bc][cc][dc][db][gb][hc][ib]MA[da][sc]C[White has the basic shape in both corners. However, because moves at X create miai to either connect or live, these basic shapes are no longer dead if White plays first. It is worth noting that White could create a ko in either shape by playing at O, but in these situations they are unnecessary.]CR[sb][ca])

When liberties matter
Finally, we should look at the situation when White has descended on both sides. Here, White is unconditionally alive provided she has enough liberties.

(;GM[1]FF[4]CA[UTF-8]SZ[19]AB[ob][oc][od][pd][qd][rd]AW[pa][pb][pc][qc][rc][sc]C[White is unconditionally alive. Black's attempts to kill fail.](;B[ra];W[rb](;B[qa];W[sa];B[sb]SQ[sa];W[qb];B[sd];W[sa])(;B[sa];W[qa]))(;B[rb];W[ra](;B[sb];W[qb])(;B[qb];W[sb])))

When White has exactly one liberty the result is a ko, because White can no longer squeeze the black stones:

(;GM[1]FF[4]CA[UTF-8]SZ[19]AB[ob][oc][od][pd][qd][rd][sd]AW[pa][pb][pc][qc][sc][rc]C[When White has exactly one liberty, Black can create a ko by playing at 1. Playing at 2 is a mistake.](;B[ra];W[rb]C[Black plays at 1 to create the ko. 2 is an obvious mistake.](;B[qa];W[sa]C[White must play at T19. If Black is permitted to play T19, the result becomes a bent-four-in-the-corner shape, which is dead.];B[sb]LB[qb:A]C[Unlike when White has 2 or more outside liberties, with exactly 1 liberty White can no longer play at A. Instead, White must fight the ko.])(;B[sa];W[qa]))(;B[rb];W[ra](;B[sb];W[qb])(;B[qb];W[sb])))

When Black has removed all of White's outside liberties, he can kill White.

(;GM[1]FF[4]CA[UTF-8]SZ[19]AB[oa][ob][oc][od][pd][qd][rd][sd]AW[pa][pb][pc][qc][rc][sc]C[Playing at 1 is the killing move. Playing at 2 creates an unnecessary ko.](;B[rb];W[ra];B[sb]LB[qb:A]C[Because White has no outside liberties, she can no longer play at A to save this group.])(;B[ra];W[rb];B[qa];W[sa];B[sb]C[White can now save the group by winning the ko.]SQ[sa]))

Black's killing technique here should also reveal why the outside liberties mattered in previous examples of unsettled shapes.

I plan to explore related shapes, such as those with one extra white stone on the second line, later. But for now, hopefully I won't have to spend as much time thinking about this shape's variations in games.

[illluck]

Nice and thorough analysis of that shape. Just wondering, for the last corner situation where liberties matter, did you forget to include what happens when white has exactly 1 liberty?

[amatterof]

Nice and thorough analysis of that shape. Just wondering, for the last corner situation where liberties matter, did you forget to include what happens when white has exactly 1 liberty?

Ah! This is such an important version of the shape. I can't believe I forgot to include it. Thanks for pointing this out!

I've edited the post to include the variation now.

[Kirby]

I think this kind of thorough shape study is an excellent thing to study.

I hope that you continue in this thread with several more shapes!

[paK0]

Yes, really interesting so far. I only knew about the basic shape, but not what happens with all the extra liberties and stuff.

Keep it up ^_^

[amatterof]

So many interesting things happen in the corners of a goban. Today, I want to look at a common situation in which one player attempts to invade an opponent's corner. As with most of my studies, the goal here is to research these variations ahead of time, so that I am not scrambling to untangle the various options under time pressure. (Studying this position was inspired by a recent question from paK0.)

(Edited Dec. 17, 2014 to fix errors and incorporate comments from Uberdude and Dragonfist)
(Edited Dec. 20, 2014 to fix error in response C)

$$B Corner
$$ ------------------
$$ . . . . . . . . . |
$$ . . . . . . . . . |
$$ . . . 3 . a W . . |
$$ . . . . . 1 e c . |
$$ . . . . . . 5 b . |
$$ . . . . . 6 2 d . |
$$ . . . . . . . . . |
$$ . . . . . . . . . |
$$ . . . . . . . . . |
$$ . . . . . 4 . . . |

Click Here To Show Diagram Code

[go]$$B Corner
$$ ------------------
$$ . . . . . . . . . |
$$ . . . . . . . . . |
$$ . . . 3 . a W . . |
$$ . . . . . 1 e c . |
$$ . . . . . . 5 b . |
$$ . . . . . 6 2 d . |
$$ . . . . . . . . . |
$$ . . . . . . . . . |
$$ . . . . . . . . . |
$$ . . . . . 4 . . . |[/go]

Black has a number of options for his next play, and his response will vary depending on the overall board situation. In particular, the Black should consider the weakness of the white stones on the side and the remaining strength of his corner group if it is invaded.

• A is the peaceful continuation, allowing White to reduce the black corner by connecting the invading stone back to the outside stones.
• B and C attempt to split White's invasion and force it to live inside.
• D and E are other attempts to force White to live inside, but are less commonly seen.

Let's explore each.

(;GM[1]FF[4]CA[UTF-8]SZ[19]AB[pd][qe][nc]AW[qf][pf][pj][qc]LB[pc:A]C[A Black response at A proposes a peaceful solution to the invasion. Black allows White to reduce his corner, but keeps many of his points.];B[pc]C[White's proper follow-up is at 1.

If White tries resist Black's peaceful offer and attempts to live in the corner with 2, the best result she can hope for is a ko, while her outside position will be significantly undermined.

White 3 is not recommended, but may be possible.](;W[re]C[Black usually responds at either 1 or 2.

If Black strongly desires sente, he may consider playing at 3. This results in a loss compared to 1 or 2.](;B[rd];W[qd];B[rc];W[pe];B[qb]C[Black can later capture the two white stones in a small endgame move.

His corner is now a solid 12 points, and depending on the surrounding situation, possibly more.

White has expanded her side by 2 points, but this is offset by the two white stones in atari. More importantly, White has sente.])(;B[qd]C[White may consider playing elsewhere (tenuki) at this point. If white can later fix the weakness at S14, she can later play S17 to make points in the corner.

Otherwise, White continues at S16.];W[rd];B[rc]MA[sd]C[White can continue reducing the corner with 1 or 2. Depending on the global situation, however, white may wish to play elsewhere.

This continuation is fairly big endgame. This is because it is somewhat of a double-sente situation. White 1 ends in sente for white, while a black play at X would almost certainly gain a white response.

Playing at 2 gives White a slightly larger reduction, but in gote.](;W[sc];B[rb];W[sb];B[qb]C[Black's corner is now reduced to 9 points. However, White has sente.])(;W[rb];B[qb];W[sc]C[Black's corner is now only 6 points, and may be even less if White is successful in winning the ko. However, Black does have sente now.]))(;B[qb]C[If White does not respond, her invasion will come to almost nothing and will have served to merely strengthen Black and remove aji.];W[rd];B[rc];W[qd]MA[rb]C[Later in the endgame, White can still play X in sente.]))(;W[rd];B[re];W[pb]C[Black should respond at 1. Playing at 2 is generally a mistake.](;B[ob];W[rb];B[pa];W[qa];B[qb]MA[pb]C[White must find the first threat, and later Black can play A.]LB[rg:A])(;B[rb];W[qb];B[ob];W[ra];B[sb];W[se];B[rf];W[pa];B[sa];W[rc];B[qa];W[sc]C[White lives.]))(;W[qd];B[re]C[White must play at 1 to live.](;W[rb]C[Black can play 1 to force White to live in the corner, or sacrifice the two stones with 2.](;B[pe];W[pb];B[ob];W[qa];B[sc];W[sb];B[sd];W[rc])(;B[qb]C[White's proper response is a 1.](;W[rc];B[pb];W[pe];B[rd];W[rf];B[ra];W[sd])(;W[pb];B[qa];W[pa];B[rc];W[ra];B[qb];W[qa];B[rd]C[White dies.])))(;W[pb];B[rc];W[rb];B[rd];W[qb];B[ob];W[pe];B[sb])))

(;GM[1]FF[4]CA[UTF-8]SZ[19]AB[pd][nc][qe]AW[pf][qf][pj][qc]LB[re:A][re:B]C[A Black response at B is probably the most commonly seen answer to the White invasion. It aims to separate the invading stone, while limiting the number of forcing moves that White can make in an effort to live.];B[re];W[ob]C[Black will normally play 1, trying to keep White low and minimize eye space. Attempting to seal White in with 2 will usually let White live in the corner.](;B[pc];W[pb];B[rc];W[rb];B[qd];W[qb]C[If Black does not fear the cut at O16, he can kill White by playing at 1. Otherwise, White lives when Black retreats with 2.](;B[nb];W[oc];B[od];W[nd])(;B[md];W[nb];B[mb];W[sb]))(;B[nb];W[oc];B[od];W[rc];B[ne];W[pc]C[White is alive.]))

(;GM[1]FF[4]CA[UTF-8]SZ[19]AB[pd][qe][nc]AW[pf][qf][pj][qc]LB[rd:A][rd:B][rd:C]C[A Black response at C is an aggressive response seen more commonly in amateur games rather than professional play. White must be careful to avoid his invasion being killed entirely. ];B[rd]C[White continues at 1 or 2.](;W[pc]C[Black responds aggressively at 1 if he does not fear the cut, or at 2.](;B[oc];W[rc]C[Black plays at 1 to create a ko in the corner, or at 2 to allow White to live, while attack the right side group more severely.](;B[qd];W[ob];B[nb];W[oa];B[sc];W[sb];B[ra];W[qa];B[qb];W[pb];B[pa])(;B[rf];W[ob];B[nb];W[sd]C[Black plays at 1 or 2, but White lives in either case.](;B[qd];W[pa]C[White is alive.](;B[sc]C[White must answer at 1, not 2.](;W[rb])(;W[sb];B[ra]))(;B[ra];W[rb];B[na];W[qa];B[pb];W[sa]LB[oa:A][sc:B]C[White has miai at A nd B to live.]))(;B[se];W[sc];B[pb](;W[qd];B[pe];W[re])(;W[pe]))))(;B[od];W[rc];B[rf];W[ob];B[nb];W[ra]C[White is alive, but her right side group will come under attack.]))(;W[rc]C[Black continues at 1 or 2. A play at 3 is a mistake.](;B[pc](;W[qd];B[re]C[White must play at 1 to create additional liberties before attempting to cut off the Black stones. Cutting now (with 2) is a mistake.](;W[pb];B[ob];W[qa]C[Black's proper response is at 2. Attempting to kill at 1 is an overplay now that White has more liberties.](;B[rb];W[sb];B[ra];W[pe];B[oa];W[qb];B[pa];W[sa];B[sd];W[rf])(;B[pe];W[rb]C[Black can now attack the left side group; that is the price for White's invasion.]))(;W[pe];B[rb]C[Black S18 is a tesuji to win the capturing race.](;W[rf];B[qb])(;W[qb];B[sc];W[sb];B[sa];W[qa];B[pb];W[rf];B[pa];W[se];B[ra])))(;W[pb];B[ob](;W[qd];B[re];W[qa]C[Black should connect at 1. Attempting to kill with 2 is an overplay.](;B[pe];W[rb]C[White lives, while Black can continue to attack the left side group.])(;B[rb];W[sb];B[ra];W[pe];B[oa];W[qb];B[pa];W[sa];B[rf];W[rg]C[White has captured the Black stones.](;B[se];W[sg];B[sc];W[sf])(;B[sc];W[sf];B[rb];W[sd])))(;W[qa]C[Black should play at 1. Attempting to kill (with 2) is an overplay.](;B[rf]C[White must spend another move in the corner to live.];W[rb])(;B[rb];W[sb];B[ra];W[qd];B[re];W[pe];B[oa];W[qb]C[White has captured the corner and the entire side.]))))(;B[qd];W[ob]C[If Black is weak in the area, he should play at 1. If he does not fear the cut at O16, then 2 is an aggressive response.](;B[oc];W[nb];B[mb];W[pc];B[md];W[ra]MA[rf]C[White is alive, but Black still has a powerful move around S14 to attack White's side group.])(;B[nb]C[If Black is strong in the area, and not afraid of a cut at O16 (for instance because of a stone around N15), he might try this aggressive approach to killing White.];W[oc]AB[me];B[od]C[White plays at 1 to create a ko. A play at 2 is a mistake.](;W[pc];B[rb];W[sb];B[qb];W[ra];B[qa]C[White plays at 1 to create a Ten Thousand Year ko, or at 2 to start the ko directly.](;W[oa]C[This is the classic Ten-Thousand-Year Ko shape. If White begins filling in the liberties at Q18 or Q19, Black can start the ko, and filling the ko now creates a dead shape.

If White plays elsewhere, Black can fill the outside liberties, and then elect to begin the ko. He then has the option of creating a seki, or attempting to kill the entire White group.])(;W[pa];B[sa]))(;W[ra];B[oa]C[The proverb says "There is death in the hane."];W[sb];B[qb](;W[pb];B[pa];W[qa];B[pc])(;W[pc];B[pa];W[qa];B[pb])(;W[qa];B[pa]))))(;B[pb];W[qd];B[re];W[pe];B[rf];W[rg](;B[pc];W[sf];B[rb];W[sd])(;B[rb];W[sd];B[pc];W[sf]))))

(;GM[1]FF[4]CA[UTF-8]SZ[19]AB[pd][qe][nc]AW[qf][pf][pj][qc]C[Black D splits White's invading stone off from the outside. However, it leaves White a clear path to life in the corner, in exchange for Black's ability to weaken the outside stones somewhat more considerably.]LB[rf:C][rf:D];B[rf]C[White plays at 1 to live in the corner, or 2 to solidify the right side group in sente.](;W[rd];B[re];W[ob]C[White makes significant eye-space to live in the corner. The continuations from this point will depend on the outside situation.](;B[oc];W[nb];B[mb];W[pc];B[md]MA[rb]C[White is alive, but must be careful about the remaining aji around X.];W[tt];B[rb];W[rc];B[na]C[1 is correct. White must give up the 2 stones to live.

2 and 3 are mistakes.](;W[qb](;B[pa];W[sb];B[ra];W[qa];B[sd];W[pb])(;B[oa];W[ra]LB[pa:A][sb:B]C[A and B are now miai for life.]))(;W[pb];B[qa];W[sb];B[sd];W[ra];B[sa])(;W[pa];B[qb];W[pb];B[sd];W[sb];B[ra]))(;B[nb]C[If Black is not afraid of the cut at O16 (for instance because he has another stone around N15), the Black might choose to block this way.];W[oc]AB[me];B[od];W[rb]MA[qh]C[Black now has a powerful play around the marked spot. This is the price White must pay for the invasion.]))(;W[rg]C[Black usually responds at 1. Playing at 2 leaves considerable aji in the corner.](;B[pc];W[re];B[rd];W[sf];B[rc]C[White has solidified his own territory and strengthened the right side group in sente.])(;B[re];W[ob])))

(;GM[1]FF[4]CA[UTF-8]SZ[19]AB[nc][pd][qe]AW[qf][pf][pj][qc]LB[qd:A][qd:B][qd:C][qd:D][qd:E]C[Black E is also sometimes seen.];B[qd]SQ[qe][qf][pf]C[In many ways, Black E attempts to treat the White invasion as if the marked stones did not exist.

If the marked stones were not there, the result is a ko for White to live in the corner.];W[pc]C[Black's most common response is 1 or 3.

2 is an attempt to limit White's eye space, but Black must inevitably fix his thin shape.](;B[od];W[rc]C[White threatens to either live in the corner with 1, or connect underneath with 2.

Attempting to connect underneath requires White to be strong enough to capture Black's attempt to keep the groups separated.];B[ob](;W[rd];B[re];W[pb]C[If Black does not fear the cut at O15, he can kill White's invasion by playing at 1. Otherwise, White lives when Black fixes his shape at 2.

In either instance, Black has a follow-up attack on the White stones at X.]MA[rf][rg](;B[sd];W[oc];B[nb];W[ra];B[sb];W[sc];B[pa]C[White is dead.])(;B[ne];W[ra]C[White lives.]))(;W[re];B[rd];W[sd];B[rf];W[se];B[rg];W[qh];B[qg];W[pg];B[rh];W[qi];B[ri];W[rj]))(;B[rc];W[oc];B[nd];W[rb];B[sd];W[nb];B[mb];W[na];B[od];W[sb])(;B[oc];W[rc];B[rd];W[ob];B[nb];W[oa]MA[od]C[If Black does not fear a cut at X (for instance because of a stone around N15), then he can create a ko in the corner.]AB[me];B[sc];W[sb];B[ra];W[qa];B[qb];W[pb];B[pa]))

I probably missed some important continuations and variations here. Please feel free to point them out, or any errors I might have made.

[Uberdude]

Something you missed is that after black q17 s15 r16 it is often better for white to tenuki as if white gets to defend the cutting point, either by black peeping it or making hanging connection herself for eyeshape, then white can later connect to the 3-3 stone with descent.

With the 4th variation of hane, white can also just block rather than living in the corner, black then q17, and white captures the stone in sente and black keeps a big corner.

And in the last sgf, if white pushes (kosumi is also possible), then p17 block is normal, extend seems rather soft.

Also I think you have given too much prominence to C: D and E are more common: the order from ps.wlatheri.net is B, D, E, A, C. The most common beginner mistake I see is playing the wimpish A too much, though in pak0's case it would probably have been prudent .

Also it's important to say black's choice is strongly influenced by how weak white's outside group is.

[tentano]

This is really good stuff. I can't really contribute but it's a good read.

[amatterof]

Thanks for the very helpful comments @Uberdude. I'm going to work on integrating these into the sgfs in the post.

I think my focus on variation C is probably due to playing on Tygem recently, where the aggressive nature of many players probably explains why I see it played against me more than is called for.

[Dragonfist]

Hi, I'm curious about a variation.

After black Q17 after the 3-3 invasion, how about wR16? I think it's possible, but rarely good. However if the side is strong W can take the corner like this later. If black just connects at q15 white will get the corner with s15. Feels like black must play s15 in response to r16, but then white will play w q18, bp18, ws18. If black does anything but connect at q15 white will kill the 2 stones, so black must connect and white will live with r19. What are your thoughts on these variations? Looks like an option to me, but have never heard of it being possible, although I believe I saw one or 2 pro games where it was played but I'm unable to find them now. In addition to black breaking the side living in the corner like this will also end in gote for white, as opposed to the usual sente variations. Am I missing something?

[amatterof]

I've made a bunch of edits to the 3-3 corner invasions post to incorporate Uberdude's helpful comments. Also, in working to respond to Dragonfist's question, I discovered some serious errors in my discussion of Black's response at C. I've updated everything to hopefully be more correct, but I encourage anyone else to point out other errors/omissions.

Hi, I'm curious about a variation.

After black Q17 after the 3-3 invasion, how about wR16? I think it's possible, but rarely good. However if the side is strong W can take the corner like this later. If black just connects at q15 white will get the corner with s15. Feels like black must play s15 in response to r16, but then white will play w q18, bp18, ws18.

I don't think Black would respond to W q18 at p18, but rather at s17. However, let me know if I'm missing something about why you think the p18 response is necessary.

[Dragonfist]

You're right, I put the wrong order. Instead of Q18 white should immediately play S18. Now W has miai to cut of the 2 stones or live in the corner.

Not sure how to get the embedding to work but anyway like this: http://eidogo.com/#2kfrK74ln

[amatterof]

You're right, I put the wrong order. Instead of Q18 white should immediately play S18. Now W has miai to cut of the 2 stones or live in the corner.

Interesting! S18 is quite the vital point. I've fixed this in the post. Thanks for pointing this idea out. I'd never seen it before.

[amatterof]

I spent some time watching the first few of Haylee L's streaming games. This wasn't the most serious of study, but everyone needs a day off now and then. Still, I feel I came away from the games with something useful. Mostly, I was impressed by how calm she seemed to be in the face of moyos built by her opponents, particularly in Games 1, 2, and 7. I need to be better in my own games about resisting the temptation to immediately deeply invade after my opponent builds a framework. With patience and proper play, it's possible to build up enough strength on the outside that the moyo can be reduced naturally, as she showed in her games.

Here are the games I watched:

• Game 1 - An eventual invasion of a Kobayashi framework leads to the opponent's collapse
• Game 2 - The opponent opens with 5-4 and an extension, and Haylee is happy to give thickness, pointing our the flawed shapes. Eventually, Haylee builds her own moyo and kills the resulting invasion.
• Game 3 - An error in a complicated joseki in the lower right leads to an early resignation
• Game 4 - Lots of patience shown here. Little victories around the board eventually create an insurmountable lead in points.
• Game 5 - A good example of how to deal with the high-chinese opening without panicking.
• Game 6 - This game ended early on time, so there wasn't much to it. I do need to add the lower right joseki to my list of things to study.
• Game 7 - More patience in the face of a black framework, but also shows just how hard it is to kill a strong player's stones when that is all they have to worry about.

You can find the remainder of her games at her channel: https://www.youtube.com/channel/UCTji1kQNoWIH85dB_Vxka9g.

Note: Streaming games while thinking out loud seems to be a growing trend in the online go community. I've found these streams to be of varying quality, but Haylee L's videos seem to be particularly good, if only because of the player's strength. (The consensus seems to be that she is a professional, which would make her the first English-speaking professional I'm aware of who has done something like this.)