monorepo/sgf/test_data/2019.02.15_shuwa_genan_annotated.sgf

(;GM[1]FF[4]SZ[19]AP[SmartGo:3.1.4]CA[utf-8]
GN[shuwa_genan]
PW[Inoue(Genan)Inseki]WR[8P]
PB[Honinbo Shuwa]BR[7P]
DT[1842-5-16-18]
RE[B+6]MU[0]
C[Shuwa became the leading player in the mid-19th century after the previous Honinbo Jowa retired. 
Genan was a strong player of that time who had previously tried to take Honinbo Jowa down from his post as Meijin (leader of the Go community and recognized to be the strongest player). 
Genan's playing style was often brilliant and exciting, but he failed to beat the Honinbos Jowa, Shuwa, and Shusaku in the most important games of his career. This game from 1842 is the second of a three game challenge match in which Genan hoped to show his ability to beat Shuwa with White (no komi), but lost all three games, after which the match was discontinued. ]
MU[0];B[qd];W[oc];B[ec];W[pf]LB[ec:A]
C[The Taisha at White 4 was a move that Genan liked to play. The joseki is well known to be complicated, and Genan had some special variations of his own. Indeed, he caught Shusaku once with a new variation to get a distinct advantage, but could not take it to a win after Shusaku played the ear-reddening move in one of the most famous games of Go history. 
In this game Shuwa was ready for the Taisha joseki, as we see he has played in the UL corner at A, which will be well placed when the usual fighting joseki happens in the UR corner (var for White 6). ]
;B[od]
(;W[nd]
C[Genan did not play the fighting variation here (see var). ];B[oe];W[pc]
(;B[qe]TR[of]
C[Black's bamboo joint was the standard move when the ladder favors White (see var). Playing Black 9 at the marked point is a modern variation of this joseki that was not yet played in the 19th century.]
;W[qc];B[qi];W[de];B[cd];W[ce];B[bd];W[mf]LB[qg:A]
C[White 16 threatens the invasion at A, and prepares to build a moyo on the upper side. ]
;B[of];W[gc]
(;B[fc]
C[Black 19 was not an expected move (see var). Honinbo Shuwa seemed to have a strategy of simplifying the game at the cost of giving up some of his original advantage (Black with no komi), and quite often his games with Black would seem to be about even after the opening. However, he was extremely good at converting the smallest advantage into a win in the middle to endgame. ]
;W[gd];B[ee];W[do]LB[cp:A]
C[With Black's solid group in the upper right, I agree that the LL corner seems to be bigger than the LR corner. I would have played a lower point, maybe at A. ]
(;B[cj]
C[Human pros did not like Black 23, and it was completely off the map when Zen analyzed the game. Zen suggested an invasion to the upper side (see var). ]
;W[di]LB[dq:A][qq:B]
C[I imagine AlphaGo would give White a thumbs-up for this shoulder hit, if only it could. While corner plays such as A or B are surely candidates in this board position, it seems clear that White 24 is the best move for the left side.]
;B[ci];W[dh];B[cm];W[fe]TR[ce]LB[qp:A]
C[White is willing to sac the two marked stones, and is trying to link up onthe outside and expand the moyo on the upper side.]
;B[ef];W[ff];B[eg]
(;W[fg]
C[With 32 Genan seemed to get a bit carried away, and I would call this an overplay. There was an easier way to connect on the outside (see var). ]
;B[eh];W[fh];B[ei];W[dj];B[dk];W[ej];B[fi]LB[fg:A]
C[The fight now favors Black, and I will blame White's overplay at A for this. ]
;W[dl]LB[fg:A]
C[My guess is that this clamp is the move that Genan had in mind when he played the forceful sequence starting with A. ]
(;B[ek]
C[Connecting beneath would have been weak (see var), and naturally Black pushes through. ]
;W[fj];B[gi];W[dd];B[dc];W[ck]
(;B[el]
C[With 47 Black could not cut on the third line (see var). ];W[bk];B[ch]
(;W[cg]TR[dj][bk]LB[dg:A]
C[Covering here is a local sacrifice that gives White time to save the two marked groups. If White had pushed through at A, Black could have captured one or the other (see var). ]
;B[dg];W[fl]LB[gj:A]
C[White 52 was the good shape move, the alternative being to simply push out at A. ]
(;B[fk]LB[em:A]
C[If Black had simply extended at A, that would give White a chance to win the semeai on the left (see var). ]
;W[gj];B[bg];W[cf];B[ed]LB[ed:A][bf:B][hi:C]
C[With A, Black wins the semeai while eliminating any possibility for White to play a forcing move from the center. For example, if Black had played at B, White would have been able to force with C. ]
;W[dm]TR[bk]
C[White saves the marked stones.];B[em];W[dn];B[gl];W[hj];B[hi];W[ij];B[il]
(;W[jf]
C[White's weak group in the center combined with White's loosely surrounded moyo on top gives White a thin shape overall. On the left, a cut on the second line might seem to have potential, but Black can win the semeai (see var).  ]
;B[jh]LB[jg:A]
C[This knight's move from a living group is slack, but there is some merit in keeping the game simple when Black is already ahead. Pros analyzing the game have hesitated to call this move a mistake because we know that Shuwa had an almost perfect record for using such moves to maintain a small lead. I would have played at A without thinking twice, and Zen tells me that my move is better. ]
;W[kj];B[gn]LB[gn:A]
C[While Black A is indeed a vital point in Black's shape, agian I get the feeling that Shuwa is saying that he is confident of a win. ]
;W[oi]LB[ph:A][qh:B]
C[With this capping move Genan is trying to reinforce the center while putting pressure on Black's right side. Zen suggests White A, Black B, and the playing away to one of the corners on the lower side. ]
;B[hg]LB[ie:A]
C[Another tight move, but this is big with the extra value of Black A next. ]
;W[qj];B[rj]
(;W[qh]
C[Throughout the years, this move has been critisized as an overplay, and the computer analysis seems to agree (see var). ]
;B[pj];W[pi];B[qk];W[ri]
C[White has challenged Black to a ko, when it seems that Black should be able to manufacture some ko threats on the upper side to have an advantage. The way Shuwa does this is instructive. ]
;B[ie]LB[rh:A]
C[If Black had immediately cut at A to start a huge ko, chances are that White would have ignored any ko threats towards the upper side. Therefore, Shuwa lets the right side wait, attacking the upper side first. ]
;W[if];B[hf]
(;W[jd]
C[Cutting was dangerous (see var). ];B[hd];W[hc];B[me]TR[od][qd]LB[rh:A]
C[Black is hoping  to make some forceful ko threats on the upper side, also looking for a way to strengthen the marked group. The ko at A is still waiting to happen. ]
(;W[lf]
C[White pulls back to keep the pressure on Black's group on the right. The variation shows how it would have been easier for Black if White had simply pushed through. ]
;B[ne];W[md];B[le];W[kd];B[kf]
C[By cutting White off, Black weakens White's shape in the center and increases the number of ko threats for Black on the upper side. ]
;W[kg];B[ke];W[je];B[jg];W[kh];B[rc];W[rb];B[rh];W[qj];B[ld];W[lc];B[qi]
(;W[bh]
C[From the start, White was counting on these ko threats on the left, but there will not be enough of them now that Black has plenty on the upper side. Maybe White can try to kill Black in the upper right without a ko? No, it would end badly (see var). ]
;B[bi];W[qj];B[mc];W[nc];B[qi];W[bf];B[ah];W[qj];B[kc];W[mb];B[qi];W[af]
;B[cl];W[bm];B[ki]LB[ag:A]
C[This is a point where Black could have won the semeai on the left by playing at A, but took this chance to ignore White's ko threat. Not to resolve the ko just yet, because Black still has plenty of ko threats, and attacks in the center first. ]
;W[qj];B[lb];W[mc];B[qi];W[ag]
C[The left side is now a step ko that favors White locally, but it will take two more moves for White to completely resolve this ko. Black's plan will be to obtain an advantage on the rest of the board while White must spend two moves on the left. ]
;B[lj];W[qj];B[ic];W[jc];B[qi];W[li];B[ji];W[qj];B[ib];W[id];B[qi];W[mj]
;B[lk];W[qj];B[he];W[jb];B[qi];W[mh];B[si]TR[mh]
C[Finally the ko ends. The marked White group is not alive yet. ];W[oj];B[ok]
;W[mk];B[pq];W[jj]LB[jj:A]
C[White's connection at A has always been a mystery to me, to the extent that I have asked if it was a mistake in the game record. Other commentators have similarly failed to find a reason for this move, and I am relieved to be able to say that the computer does not show it as a candidate. ]
;B[ll];W[ol];B[eq];W[cq];B[mp];W[bh]LB[bh:A][be:B][gq:C]
C[Not to forget that White had this ko to deal with at A. Otherwise, Black could have made a one-step approach ko by playing at B. Could White have invaded at C first? I would like to, and after erasing the lower side White would ahve an advantage in territory, but the ko would be very dangerous. ]
;B[gp];W[nk]LB[nk:A]
C[White A is a point where Black could have played to attack White's eye space in the center. While it was not immediately necessary, it was a large move that takes away Black's threats and reduces Black's potential on the right side. ]
;B[mn]LB[lo:A]
C[Could it be that Shuwa was a bit greedy here? Black is winning, and I would have been happy to play the more conservative A, but maybe that's just because I've seen this game and I know what happens next. ]
;W[gq];B[fq];W[hp];B[hq];W[mq];B[nq];W[lp]
(;B[gr]
C[What a patient and quiet move! I would want to cut again, of course (see var). ]
;W[nr];B[np];W[kq]LB[mr:A][lr:B][lm:C]
C[If Black cuts at A, White offers to play a ko with B. White doesn't have enough ko threats elsewhere, but could use local ko threats such as C for this ko. ]
;B[or]
C[Black declines to play the ko. ];W[lr];B[jr];W[ln];B[lm];W[jq];B[pk];W[nm]
;B[lo];W[ko];B[kn];W[jo];B[ns];W[mr];B[pm]LB[pm:A][gq:B]
C[Shuwa offers no resistance, allowing White to live without even a ko. However, when Black plays at A, it suddenly appears that Black has not lost anything. Black's territory just switched from the lower side to the right. Failing to find anything better for either side, I will say that the sequence from White's invasion at B ended in an even result. ]
;W[on];B[pn];W[bc]LB[cc:A][ai:B][ad:C]
C[By exchanging this move for Black A before finishing the ko with White B, White created a big move at C, gaining some points in the corner. ]
(;B[cc];W[ai]LB[bc:A][cc:B][ad:C]
C[The exchange of A for B was a cleverly timed exchange that set up the connection at C and made the game very close. ]
;B[ir];W[kr];B[cr];W[ad];B[bb];W[br];B[dr];W[ab]
C[Yet another ko starts. Since the life of Black's group is at stake, and White's potential local loss is small, Black will seek a peaceful ending here. ]
;B[ac];W[oo];B[pl];W[bc];B[ca]LB[ac:A][gb:B][aa:C]
C[If White connects at A, Black can live with B. Otherwise, Black can capture two stones in the corner with C. ]
;W[qg];B[og];W[ph];B[rg];W[fb];B[ac];W[op];B[oq];W[bc];B[eb];W[nn];B[gb]
;W[hb];B[fa];W[rd];B[re];W[sc];B[cp];W[bp];B[dp];W[co];B[ip];W[jp];B[in]
;W[fo];B[fn];W[eo];B[ho];W[ac];B[kk];W[qf];B[rf];W[jn];B[jm];W[ps]LB[cs:A]
C[It has been suggested that Genan misread the tsumego in the lower right. However, if one assumes he just played the biggest move at A (maybe with some optional one point forcing moves first), the game would have ended with Shuwa winning by one point. Having already had the experience of losing to Shuwa by the smallest margin, maybe Genan chose to avoid that and take the more exciting ending. ]
;B[os];W[qo];B[po];W[qq];B[pp];W[qr];B[rn];W[ro];B[so]
(;W[cs]
C[Having lost five points in the LR corner, Genan will lose by six points now. He decided not to finish the tsumego, so I will do it for him in a var. ]
;B[ds];W[bs];B[se];W[jk];B[jl];W[gk];B[hl];W[en];B[fm];W[no];B[mo];W[js]
;B[is];W[ks];B[sd];W[rc];B[ms];W[ba];B[aa];W[sp];B[qp];W[ba];B[cb];W[ae]
;B[aa]
C[The result is recorded as Black winning by six points, when White can win the final ko to get a 5 point difference. Black A (253) should have been played at B to get the correct result, and I think this could be a mistake in the game record. ]
LB[hl:A][hk:B])
(;W[sp];B[qp];W[rp];B[sr];W[rr];B[sq];W[ss];B[qs];W[pr];B[rs];W[ss];B[rs]
;W[rq];B[sn];W[ss];B[sr]LB[ss:A][sq:B]
C[Some extra moves have been added to simplify this for my explanation. This is called the bent four in the corner, and White has no viable local moves, but it is not a seki. Afterfilling all outside liberties, Black can play A, the B, and after White takes the four stones it will be a ko in which Black takes first. Since Black can choose the timing of this ko, this shape is ruled dead in Japan. ]
))
(;B[ad];W[cc];B[bb];W[ac];B[cb];W[dr];B[bg];W[aj];B[bj];W[ak];B[bl];W[cn]
;B[df];W[al]TR[bf]LB[bc:A][ad:B][cc:C]
C[If Black answers White A at B, the cut at C gives White's marked group an extra liberty and makes the threatened step ko more distant. As this var shows, White can even win the semeai without a ko. ]
))
(;B[lq];W[mr];B[lr];W[np];B[mo];W[oq];B[nr];W[or];B[ms];W[go];B[gr];W[pp]
LB[lq:A][fo:B]
C[If Black cuts at A, White can trade to the LR corner, with the endgame move at B as a bonus. This makes the game close.  ]
))
(;W[rg];B[si];W[pg];B[mh];W[ng];B[og];W[nh];B[oh];W[ph];B[ni];W[mi];B[nj]
;W[lh];B[mj];W[li];B[qf]LB[mh:A]
C[After Black A, White has no way to hold things together. An easy win for Black. ]
))
(;W[ne];B[nf];W[lf];B[nh]LB[ke:A][nc:B]
C[Black's group on the right has escaped into the center, and the aji at A and B is troublesome for White already. ]
))
(;W[he];B[id];W[kd];B[ib];W[hb];B[ic];W[kb];B[ge];W[gf];B[gb];W[fb];B[ga]
;W[ha];B[ia];W[fa];B[gb];W[hc];B[fd];W[jc];B[hd];W[ge];B[gh]
C[Black has an advantage in this fight on the upper side, and in this variation wins by one move. ]
))
(;W[pj];B[rk];W[qh];B[rh];W[pi];B[ri];W[eq]LB[pj:A][ie:B]
C[Pulling back at A was the normal move for White, and White has succeeded in reinforcing the center while pushing Black down to the second line. I think Black still has a slight lead with Black B next. ]
))
(;W[bh];B[bi];W[bf];B[ah];W[af];B[bj];W[bm];B[ag];W[aj];B[ae]LB[bj:A]
C[With the forcing move at A, Black gets an extra liberty to win the semeai. ]
))
(;B[em];W[bh];B[cf];W[bg];B[gl];W[gj];B[hj];W[bj];B[ke]LB[em:A][bh:B]
C[When Black extends at A, White can win the semeai with B. Actually, this is not bad for Black, but it does not seem to be Shuwa's style to give White territory on the left when he can win the semai as in the game sequence. ]
))
(;W[dg];B[cg];W[df];B[ed];W[dm];B[em];W[dn];B[gl];W[bf];B[be];W[cf];B[bg]
;W[ag];B[ah];W[af];B[gj];W[ai];B[fk];W[bh];B[ae]LB[dg:A][ed:B][dm:C][gl:D]
C[If White pushes through at A, Black can crawl once and simply connect at B. After White plays C, Black can capture the other side group with D as shown, so the two points were miai to give Black a win. ]
))
(;B[cl];W[bk];B[el];W[ch]LB[cl:A]
C[Cutting at A is bad when the ladder favors White. ]))
(;B[ck];W[ek]LB[dl:A][ck:B]
C[The exchange of A for B was a big kikashi for White, and White has an advantage. ]
))
(;W[dd];B[dc];W[gh]LB[ed:A]
C[Since White is threatening to cut Black off at A, Black cannot break through in the center. This board position is close to even, without needing any komi.]
))
(;B[ke];W[kc];B[me];W[md];B[ne]LB[ke:A][me:B]
C[When Black invades at A, White will have trouble attacking because of the weakness at B. Black can play B immediately as shown, or play away after the original exchange, which is a kikashi for Black. ]
))
(;B[fd];W[gd];B[fe];W[ge];B[ff];W[hg]LB[fd:A]
C[Black A was the natural looking move. While White gets a nice moyo on the upper side, Black still has a distinct advantage when there is no komi. ]
))
(;B[qc];W[pd];B[pe];W[qe]LB[of:A]
C[When the ladder at A favors White, this is generally considered to be bad for Black. ]
))
(;W[pd];B[pe];W[pc];B[qe];W[oe];B[nd];W[qc];B[qf];W[of];B[qh];W[mb];B[ld]
;W[oh];B[pj];W[lg];B[jd]LB[ec:A][of:B][ne:C][pg:D]
C[This variation shows a basic variation of the Taisha joseki. Black A is well placed in relation to the Black group on the upper side, and this board position is good for Black. 
A major diverging point in this joseki was White's connection at B, where White could have pushed at C or extended at D, with a multitude of complicated variations following. ]
))