36.o3

|36.o3 - what if 37.r5 ?

38.r4 39.s7 40.q2 41.t5

38.t4 39.q7 40.q2 41.s3

38.q7 39.t6 40.r8 41.t9

38.r8 39.q7
?

36.o3 37.r5 38.s3

|36.o3 37.r5 38.s3 39.t6

Yep 34.p5 is a loser. What about 35.p4?

Yep 34.p5 is a loser. What about 34.p4?

Yep 34.p5 is a loser. What about 34.o4?

I think |34.o4 does not sufficiently protect against the latent threat 35.n9. For example, 36.m3 37.j4. Then again, n9 would seem to work in the regular line as well (|34.p5 35.n9) and that should have been simpler than what actually happened... unless I am missing something.

It's interesting, because as of |24.p13, it looks like Black has the game won. Nobody would have blinked if White resigned that position.

What would you have done as black for #26? 26.r13?

Not sure it works, but |26.q15 looks best to me.

27.s15 28.r17 29.t17 30.s21 31.r18 32.q20 33.p19 34.o19 35.n18 36.m20 37.g19 38.j17 or maybe 33.p17 34.m20 ... but 31.q20 and 31.r19 also have to be checked.

|26.r13 looks equivalent after 27.s12 28.q15 29.s15, but if the path across the bottom doesn't work, this is worse across the top.

Thanks Alan :-)

Years ago, Klaus Hussmanns showed me the resource |26.q15 27.s15 28.r17 29.t17 30.s21 31.r19 which I believe is best for white in this variation. For example
32.q20 33.t20 34.u20 35.v19 or
32.t18 33.u19 34.u22 35.v21 36.w21 37.t20 38.p20 39.q19 40.o22 41.o18 42.m19 43.l17 44.j18 45.d19
Maybe 28.r19 would be better for black?

Lemme follow your Klaus line up to

|26.q15 27.s15 28.r17 29.t17 30.s21 31.r19 32.t18 33.u19 34.u22 35.v21 36.w21 37.t20 38.p20 39.q19

and then 40.q20 would seem to be winning for Black... or 40.n21, then 41.p18 42.m19 43.l15 44.g19 or 41.q19 42.q20 43.s18 44.m19 45.l15 46.g19 which also looks good for Black.

I totally screwed up the cut-and-paste on that latter line. n21 was supposed to follow o20:

|26.q15 27.s15 28.r17 29.t17 30.s21 31.r19 32.t18 33.u19 34.u22 35.v21 36.w21 37.t20 38.p20 39.o20 40.n21

then 41.p18 42.m19 43.l15 44.g19 or 41.q19 42.q20 43.s18 44.m19 45.l15 46.g19

eh, maybe not.... |26.q15 27.s15 28.r17 29.t17 30.s21 31.r19 32.t18 33.u19 34.u22 35.v21 36.w21 37.t20 38.p20 39.o20 40.n21 41.q19 42.q20 43.p18 44.m19 45.l17

|26.q15 27.s15 28.r17 29.t17 30.s21 31.r19 32.t18 33.u19 34.u22 35.v21 36.w21 37.t20 38.p20 39.q19 40.q20 41.p18 42.n21 43.l17 so I don't follow your complaint against q19. At any rate, isn't 28.r19 better for black?

I should include the xariation 26.q15 27.s15 28.r17 29.t17 30.s21 31.r19 32.t18 33.u19 34.u22 35.v21 36.w21 37.t20 38.p20 39.q19 40.q20 41.p18 42.o19 43.n19 44.m20 45.l20 46.k21 47.d19 48.l19 49.l15

Sorry my mistake, |26.q15 27.s15 28.r17 29.t17 30.s21 31.r19 32.t18 33.u19 34.u22 35.v21 36.w21 37.t20 38.p20 39.q19 40.q20 41.p18 42.o19 43.n19 44.m20 45.l20 46.k21 47.d19 48.j17

Ok so it would appear white wins after |26.q15 27.s15 28.r17 29.t17 30.s21 31.r19 32.t18 33.u19 34.u22 35.v21 36.w21 37.t20 38.p20 39.o20 40.n21 41.q19 42.q20 43.p18 44.m19 45.l17 46.n17 47.o17 48.m16 49.n16 or 46.n18 47.m17 or 44.l20 45.k18

So my Klaus move seems to work after all. But I still believe 28.r19 works for black.