[sgf-full](;GM[1]FF[4]CA[UTF-8]AP[CGoban:3]ST[2]
RU[Japanese]SZ[13]KM[0.00]
PW[W]PB[B]AW[ek][al][bl][cl][dl][fl]AB[bj][dj][ej][fj][bk][dk][gk][gl][il][gm]C[(Apologies for putting the tsumego on the bottom: this way, the coordinates are the same no matter the board size.)
First of all E1 looks like a good place to play. Usually, the wedge-like structure of D2-E3-F2 indicates an extra eye there, that can be poked out by playing at E1. In this case, however...
(see the E1 variations)
---
As can be seen from the variations, B1, D1, and F1 all seem to work as defences for white. Those are, therefore, likely to be variations that work.
(see those there variations)
---
And, strangely, none of those three seem to work. It seems like capturing at E2 is important.
Two lines currently seem the best:
E1 F1 E2 D1 E2 F3
B1 F1 E2 E1 F3 D1
Apparently, E2 is an important point to get. Both of those variations show black being one turn too late: perhaps aiming at it immediately is best?
(see the F3 variations)
---
Hm, that doesn't work either. E2 is auto-atari, and A1 looks plain stupid, so let's try C3.
(see those variations)
---
E2 is the only move not explored so far, so perhaps it is the answer after all?
(see those variations)
---
A1 does not seem worth trying. Therefore, a variation has been missed somewhere...
C1 hasn't been explored much, maybe that's it? B1 looks most promising, though, particularly the ko.
So, erm, there. Apparently, there's no solution I can find, so the problem is left as an excercise to the reader. Hehe.]
(;B[em]
(;W[fm]
(;B[dm]
;W[el]
;B[cm]
;W[bm]C[White is just a little too big.])
(;B[el]
;W[dm]
(;B[el]
;W[em]
;B[fk]C[If white reacted to this, white would die, but E3 is tiny.]
;W[bm])
(;B[fk]
;W[el]C[This is also bad.]))
(;B[cm]C[This is obviously wrong.]
;W[dm]))
(;W[dm]C[This also works. F1 and A1 are miai]
(;B[bm]
;W[fm])
(;B[fm]
;W[bm]))
(;W[el]C[This, on the other hand, does not.]
;B[fm]
;W[dm])
(;W[bm]C[And this, once again, does. All other moves do not, as you can read out if you want.]
;B[dm]
;W[fm]
;B[el]
;W[cm]))
(;B[bm]
(;W[dm]
;B[fm]C[This kills.])
(;W[fm]
(;B[cm]
(;W[dm]
;B[el]
;W[em]
;B[fk]C[So does this.])
(;W[em]
;B[fk]C[And this.])
(;W[el]
;B[dm]C[But this makes a seki! Therefore, we want white to play E1 or D1.])
(;W[fk]
(;B[em]
;W[el]
;B[dm]
;W[am]C[White lives.])
(;B[el]
;W[em]C[Also lives.])))
(;B[el]
;W[em]
(;B[cm]
;W[fk]
;B[dm]
;W[am]C[White lives.])
(;B[fk]
;W[cm]
;B[el]C[This makes a ko -- not very good. However, white cannot connect.]))
(;B[em]C[This is better, but it doesn't stop D1.]
;W[dm]C[From these variations, we see that D1 seems more important than B1.]))
(;W[em]C[This doesn't work at all.]
;B[cm]))
(;B[dm]C[Seeing as B1 failed when white played D1, let's try that.]
;W[em]C[Now, however, E4 makes it feel like D1 should have been at B1 or C1. No moves look right, so I won't read further.
(further sequence added later)]
;B[fm]
;W[cm]
;B[fk]
;W[el]C[No, no, all this really lives. D1 is useless.])
(;B[fm]C[This is the last move that looked correct, so it makes sense to try it.]
;W[em]C[As far as I can see, this makes B1 and C1 miai. However, seeing as capturing E3 and F2 would lead to D1 becoming a false eye, B1 is better.
There's also F3, but playing that simply leads to white connecting.
Now, it once again feels like E1 was the key point... ]
(;B[bm]
;W[cm]C[There is no stone at F3, so white still lives...]
;B[fk]
;W[el])
(;B[cm]
;W[bm])
(;B[fk]
;W[el]C[Same miai for life.]))
(;B[fk]
;W[el]C[White is too big to kill.]
(;B[cm]
;W[bm])
(;B[fm]
;W[em]))
(;B[cm]
(;W[em]C[This doesn't work for white.]
;B[bm]
;W[el])
(;W[fm]C[White can still play this, though.
]
(;B[el]C[A throw-in to get white to play E2.]
;W[em]
;B[bm]
;W[fk]C[White still lives.])
(;B[fk]
;W[el]C[White lives.])
(;B[bm]
;W[dm]C[Only move.]
;B[el]
;W[em]
;B[fk]C[White is dead -- meaning F1 was wrong.]))
(;W[bm]
(;B[em]C[Can't do this due to D1 living.]
;W[dm]
;B[fm])
(;B[dm]
;W[em]C[This lives too, though, and is better for white.])))
(;B[el]
;W[em]C[This clearly lives.]
;B[fm]
;W[cm]
;B[fk]
;W[el]))[/sgf-full]