(;GM[1]FF[4]CA[UTF-8]AP[WebGoBoard:0.10.11]ST[2]RU[japanese]SZ[19]KM[6.5]HA[0]TM[1209600]OT[259200 fischer]RE[B+R]DT[2012-05-01]CP[online-go.com]PB[Uberdude]BR[7d]PW[breakfast]WR[6d]PC[OGS: http://online-go.com/game/325246];B[pd];W[dc];B[pq];W[cq];B[de];W[ce];B[dd];W[cd];B[ec];W[db]C[Uberdude: breakfast follows his usual style of playing for territory ];B[df];W[eb];B[fc];W[bg];B[kd]C[Uberdude: I thought about playing chinese opening on the right or an enclosure from the corner like n17, but decided I should extend from my wall to make sure it doesn't become a weak group, and this seemed to make a good balance with q16. d11 is another continuation, but c3 makes the left side low and uninteresting. ];W[qk];B[qm]C[Uberdude: Check from below rather than above to induce making solid territory at the top right: I didn't like r11 r5 r7 p4 p3 o4 n3 q8 r4 q4 r3 s5 n4 p6 which is a sort of joseki for this situation, as white gets a rock solid group on the right side and r11 ends up too close to it and the top right corner being open at 3-3 is annoying. Uberdude: Maybe Yu Changhyeok would play something cool like q10 here. ];W[pg]C[Uberdude: This was an interesting move, high to counteract black thickness on top side. If just r12 then black r14 makes some nice profit whilst keeping the pressure and white probably p12 or p9 jump. ];B[qi]C[Uberdude: Defending at the top with o16 seemed passive, given I already have l16, so I invaded to stop making easy points here. ];W[pi]C[Uberdude: White choose the simple answer instead of double approach the corner which is probably an overplay with l16, but complicated. Or maybe white can try something clever like r16 to try to block r11's connection to corner in sente. ];B[ph]C[Uberdude: I have both ladders so can wedge. ];W[oh];B[qh];W[pj];B[qg];W[og];B[qe]C[Uberdude: This result seems ok for both: black gets some territory and white some thickness which reduces black's top potential. White's thickness doesn't have eyes yet though, and a cutting point at p11, so he has to be careful about using the r5 gap. This kind of thickness facing thickness position I find hard to judge: is white's thickness making black's less useful or vice versa? ];W[jp];B[om]C[Uberdude: Afterwards I thought I should have peeped at o11 first: I don't think white can resist or black just cuts. ];W[nj]C[Uberdude: A solid move, I actually thought white might tenuki to take the big point at d6 and allow the o11 o9 attack, so was happy to get the p7 jump in sente to strengthen the corner aganist the r5 invasion. But now I can't get o11. ];B[lp]C[Uberdude: It was hard to decide where to play now, maybe this move would be better against c3 but I wasn't sure whether to d4 or d6 or what else. My corner is still thin with weak points at o3, r3 and r5 so this doesn't make much real territory, but my plan was if white plays d6 then I play d3 and build a group on the lower side and attack k4, and if k4 is weak then white won't have a chance to invade the lower right and I can encourage him to play moves to help k4 that help my corner. ];W[dn];B[ep]C[Uberdude: But then I changed my mind and didn't play d3 because I had recently seen this move in a pro game and it looked cool and can revert. There was some logic to this over d3: the variation I wanted with d3 was d3 d4 e3 c2 e4 d5 h4 k6 h6. But maybe white switches d5 to e5 which is stronger on the outside but gives black d2 sente. This move prevents white from choosing e5. However, it enables white to choose the e3 cut variation of the game which means k4 becomes strong which turns m4 into a bad move. ];W[dp];B[dq];W[eq]C[Uberdude: White resists my plan to attack k4 ];B[dr];W[er];B[do];W[cp];B[fp];W[gq]C[Uberdude: White could also e6, and then I f2 in sente and can attack k4 but compared to the d3 variation I have less eyespace on the side and white has a bigger stronger corner. ];B[gp]C[Uberdude: This feels like aji keshi (e.g. losing h3 k3 combo and solidifying connection to k4) but I thought I needed it to strengthen my group. Also later if white has played c2 then he can maybe answer at h4. ];W[hq];B[co]C[Uberdude: White could think about cutting at e5 here as I can't capture it. Rather annoyingly, had I got the o11 p11 exchange earlier then the ladder would be good for me! The idea of g3 was that it discourages the cut as I get b5 in sente and then e6 and f6 atari and can then either get g6 atari too before d7, or else leave the aji of the ladder and aim for a ladder breaker like n10. ];W[bo]C[Uberdude: So white opts for territory. ];B[cn];W[bn];B[cm]C[Uberdude: I now thought I was losing as white got a big corner in sente and connected to k4, making m4 thinness a problem, and my thickness isn't that great and could even be attacked later. ];W[dh]C[Uberdude: And white gets to play this nice point in a good rhythm ];B[hc]C[Uberdude: Giving white d12 in sente was painful (as if white had played elsewhere I wanted to d11) and this move feels rather slow and submissive (I could be greedier and play f18 for better yose, but white can't actually h19 monkey jump here due to f19 and then c14 and c17 shortage of liberties). But if I don't defend then after d12 a white invasion of the top is quite severe so I thought better sit tight and be patient. ];W[nq]C[Uberdude: And now the thin m4 problem comes back to haunt me. White is in control. ];B[np];W[qq]C[Uberdude: If white just n3 and then p3 or m3 and connected to the left and I got corner I would have been reasonably content, so he tries for more. ];B[qr]C[Uberdude: r4 seemed too slack territorially and made it easy for white to settle. ];W[pp];B[oq];W[rr]C[Uberdude: Classic double hane technique for sabaki. ];B[rq];W[qp];B[rs];W[op];B[sr];W[nr];B[or];W[pn];B[pm];W[nn];B[mp];W[no]C[Uberdude: It was very hard for me to judge this result (which is what I expected when I played r2) and I still don't know who it was good for: on the one hand black has kept the corner territory and white has made no points (compared to say living in the corner with a tripod group which was one variation I thought might happen here if he starts at r3 and I r4 hane), but black has lost all the points in q6 area and connecting to r7 on the 2nd line is miserable. Uberdude: My original plan was for this sort of result but that black would get sente to play o16, but as white hadn't yet exchanged l3 for m2 I worried about white pulling out his 2 stones at m2 and attacking m4: it's pretty risky for both but is possible because I gave white a stone at h3: more downside of g4. ];B[ln]C[Uberdude: Originally I was going to play this at s5 directly and then expected white to jump here or o7 (with l3 first) and then I get sente for o16. Uberdude: In fact maybe I can o16 now: if white blocks at s4 black can come out at o7 and maybe the fight is possible with o16 also making the white group at o10 weaker. But very dangerous. ];W[nm];B[ro];W[kq];B[lr];W[kr];B[mr];W[ob]C[Uberdude: White getting sente for this made me pretty depressed as I had hoped for sente to close the territory at o16. But at least with m6 I have some aji at o8 ];B[pc]C[Uberdude: Not q18 to avoid giving p17 sente. ];W[mc];B[lc];W[md];B[ne]C[Uberdude: Maybe I should just jump at l14, but I thought that made it too easy for white to connect. ];W[ke];B[je];W[le]C[Uberdude: Not exchanging m16 for l17 left me with some shape problems at the top. ];B[jf]C[Uberdude: But on the flip side this is a nice nobi. ];W[lg];B[nl]C[Uberdude: If I can't make something of this cut aji then I expected to lose. In hindsight this attack went rather well and I think is what turned the game around, so perhaps the result in the lower right corner was ok for me as although white reduced my points a lot he made a heavy group which I could attack later. ];W[ml];B[mk];W[nk];B[ol];W[lm];B[lk];W[km];B[jk];W[im];B[li]C[Uberdude: Getting this in sente was possible because of o15 making white's connection thin. ];W[mh];B[gm]C[Uberdude: It was nice to get this cap in a nice rhythm to bulk up the eyespace of my e4 group. ];W[hl]C[Uberdude: With all these moves I am trying to force white to submissively connect on dame at l6 (or else I cut at k5). ];B[fk];W[hj];B[mn];W[mm];B[jh];W[fi];B[cj]C[Uberdude: I also considered g13 to close the top, but decided this group was weaker so I should help it (even though it doesn't make much territory with b7 open) and this aims at c12 and other weaknesses in white's knight's moves and stops white make a few points and base here. ];W[gf]C[Uberdude: This move surprised me, I didn't expect white to go in so deep, just h12 honte or maybe g13. I guess he counted and decided they weren't enough: I wasn't sure. ];B[mi]C[Uberdude: Part time tesuji, but also this might not be sente later with the g12 fight and in some variations it is needed to keep the k9 k12 connection. ];W[ni];B[gh]C[Uberdude: Maybe we can call this the winning move, punishing the g14 overplay. ];W[fg]C[Uberdude: I didn't actually read this answer, h12 was the fight I spent time reading and it just about seemed to work for me. ];B[fh];W[eh];B[gi];W[gj];B[fj]C[Uberdude: And with this cut white has to defend against k5. ];W[in]C[Uberdude: This was a nice move, I only expected l6 but this is better as it has more effect on f6 area. ];B[eg]C[Uberdude: I get some nice profit ];W[dg];B[ef];W[ei]C[Uberdude: But this is sente for h12 and my shape is pretty sucky. ];B[ih]C[Uberdude: Maybe j11 is better (it makes some eyes in k11 area), but it leaves a troublesome double peep at j13 (coupled with k16 aji). If I had found a time to get l14 in sente then it's safer. Perhaps I can l14 now and offer a trade. ];W[dl]C[Uberdude: This was a very sharp move and gave me a big headache as suddenly I worry about the life of this group. I only expected white d9 to which I planned this d8 (example of my opponent's key point is mine proverb) and sacrifice c10 for sente for big yose at n18. ];B[dm];W[jc]C[Uberdude: I expected white to continue with sente reduction with d8, reducing that group to just 2 points. Uberdude: I had forgotten about this move: when I played k14 I knew white had it but then it was only yose so irrelevant, now it is very big. ];B[kc];W[jb];B[jd];W[lb];B[kb];W[ka];B[mb];W[la];B[pb]C[Uberdude: I have this or j17 in sente. As I thought white jumping into the corner at r18 was the biggest I chose this, but maybe j17 is bigger. At this time I didn't anticipate the g14 area eyeshape getting so low and j17 would make spare eyes on the edge as well as stopping white points there. ];W[nb];B[cf];W[bf];B[bi];W[bh];B[bm]C[Uberdude: I was happy to turn my shajy 2 point group into a safe 10 point one. Uberdude: *shaky ];W[ic];B[id];W[gb]C[Uberdude: Resisting at j18 didn't seem much better and I wanted sente. ];B[gc];W[hb];B[dj]C[Uberdude: So I spent 2 gotes to make some points here, and white 2 gotes to make points on the top. The white points/eyes on top did mean he could allow the n13 cut later without dying. Uberdude: q14 is also rather big and good aji. At the time I counted this as bigger, but white played better yose than I expected over there so maybe it is better. ];W[gl];B[fl];W[rj];B[ri];W[pf];B[pe];W[qn];B[rn]C[Uberdude: Probably s7 is better: it allows white a point and eye in sente at q5, but stops s8 sente. ];W[rl];B[on];W[oo];B[qo];W[po];B[rm];W[rp];B[sp];W[qf];B[rf];W[rg]C[Uberdude: Pro power yose! ];B[sg];W[re];B[rh];W[rd];B[rc];W[pa];B[qa];W[oa];B[fq];W[fr];B[ks]C[Uberdude: Better make sure I get this sente. ];W[js];B[ls];W[ir];B[bp]C[Uberdude: Best yose for ko threats, though we never had a ko this game. Uberdude: In fact this is wrong in this position due to aji in the corner: black can c2 b2 b3 b4 a2 a3 b1 to set up a common ko shape. It's not actually a ko as white has an extra liberty so can connect at b3, but then black hanes at a6 in sente and gets a semedori that is 2 points better than the game. ];W[bq];B[qb];W[sf];B[se];W[sd];B[sf];W[si];B[ho]C[Uberdude: j11 instead to make an 2 points/eyes at k11 and l10 would avoid the semedori headaches of the game. But this stops white's g6 which is big sente. ];W[jj];B[kj];W[ji];B[kh];W[fe]C[Uberdude: I knew this group had bad shape, but was only expecting/hoping for white to get c13 in sente once f18 liberty gone so I have to defend at h15 which makes an eye at j15. ];B[ed];W[he];B[hd];W[kg];B[jg];W[mj];B[lh];W[sj];B[sc];W[sh];B[qd];W[ah]C[Uberdude: I forgot this move was better than a11 as has b10 follow up. ];B[mg]C[Uberdude: Fighting spirit said I couldn't allow a12 in sente and this cut is quite valuable too. ];W[nh];B[mf];W[me]C[Uberdude: White doesn't miss this. ];B[nf];W[ej];B[ok];W[oj];B[sl];W[sk];B[sm];W[ql];B[io];W[jo];B[kn];W[jn];B[jl];W[jm];B[ek];W[fm]C[Uberdude: Is this hoping for the mistake of e7 allowing white g6? Black gets the same amount of points as if white just played h7 on the outside. ];B[gn];W[dk];B[em];W[bj];B[bk];W[aj];B[an]C[Uberdude: Stops a7 and a5 ko. ];W[ap];B[ak];W[ai];B[ci];W[so];B[sn];W[qj];B[rg];W[ip]C[Uberdude: So h13 or h14 to set up semedori seems to be 1 point better, but I think black still wins by 0.5 that way. Uberdude: Maybe white was hoping black would forget about the semedori and just plays normal yose and then white unleashes it and wins. ];B[hg]C[Uberdude: Now black wins by 1.5 ];W[hi];B[hh])