I am new to twixt. It may be a complete mistake but does it help white?
On 2008-12-16 at 13:31, pallab said:
Sorry simple F17 seems to be wining to black.
On 2008-12-16 at 15:09, Alan Hensel said:
Not bad, Pallab; it's tricky enough that White could have hoped for a blunder. (For example, 33.e1634.f1935.b1936.f1737.c15 ... might end in a draw.)
But there is another good response for Black worth mentioning: 34.b21 for the double threat at d22 and f19. Because of this threat, White cannot leave a gap (if he is sure Black will see 34.b21); he must link to his c21-g19 group on move 33, so that limits his options. I don't see a good one.
My wording in the previous post might be confusing. The lines I showed are still winning for black as far as I know. Sorry.
On 2008-12-16 at 19:28, twixter said:
I'm still working on 21.p17. If black has a win, that probably means that all my fancy footwork on the left beginning with J14 was unnecessary, because black could play
which leads to almost the same position on the right as with 21.p17. In fact, 18.018 looks BETTER for black on the right in some variations. I hope to complete my analysis soon.
Continuing Pallab's i20 variation, this is a good try, and black would lose against it if he plays mechanically with the "standard" responses of H20 or J20.
Variation A 29.h1930.h2031.f2032.e2033.h22 trying to draw. We are under standard rules here with link removal and no link crossing, so one continuation is 34.g2235.e2136.f21 This site uses Little Golem's "PP" ruleset, where links are never removed and your own links are allowed to cross. Under Standard rules, black must not remove the G21/E20 link yet. So the link from F21 to H20 is not added, but white cannot prevent black from adding it later. Meanwhile white's E21 is hammered by F21. Black will get past white on one side or the other.
This way I try to open up a second double thread in vertically on the right side. I have played out many situations - I am sure this opens up many ways to destroy my strategy for white - go ahead!
At any rate, I have changed my mind about 18.J14. I now believe black should play O18 instead, as shown in my post where I say "...allow me to back up as well..."
But if we look at the variation 18.M1819.L19 then my fancy solution with J14 works again. See my above post.
wojtek, there are so many different variations discussed here, I do not know which variation you are talking about. Please provide the first 30 moves for your variation, thanks.
1.s4 2.n7 3.i8 4.j13 5.o14 6.p19 7.m15 8.l17 9.n18 10.m20 11.m21 12.o10 13.l11 14.j7 15.k7 16.k12 17.j12
Black resigned here, but I believe black can win as follows:
18.j14 19.n12 20.m18 21.l19 22.o17 23.k16 24.k19 25.i17 26.g21 27.e20 28.d20 29.g19 30.i20 31.c21 32.h18
Please shoot holes in my analysis!
I am new to twixt. It may be a complete mistake but does it help white?
But there is another good response for Black worth mentioning: 34.b21 for the double threat at d22 and f19. Because of this threat, White cannot leave a gap (if he is sure Black will see 34.b21); he must link to his c21-g19 group on move 33, so that limits his options. I don't see a good one.
Backing up a little, what about 21.p17?
1.s4 2.n7 3.i8 4.j13 5.o14 6.p19 7.m15 8.l17 9.n18 10.m20 11.m21 12.o10 13.l11 14.j7 15.k7 16.k12 17.j12 18.j14 19.n12 20.m18 21.l19 22.o17 23.k16 24.k19 25.i17 26.g21 27.e20 28.d20 29.g19 30.i20 31.c21 32.h18 33.f17 34.i16
and now either
35.f13 36.g13 37.h14 38.i12 39.g11 40.h8 41.g9 42.f9 43.e10 44.l6 45.f8 46.f4
or
35.f12 36.h13 37.e14 38.h8 is similar.
18.o18 19.p17 20.m19
which leads to almost the same position on the right as with 21.p17. In fact, 18.018 looks BETTER for black on the right in some variations. I hope to complete my analysis soon.
Again I may be wrong, but does it help white?
28.i18 and now
Variation A
29.h19 30.h20 31.f20 32.e20 33.h22 trying to draw. We are under standard rules here with link removal and no link crossing, so one continuation is
34.g22 35.e21 36.f21 This site uses Little Golem's "PP" ruleset, where links are never removed and your own links are allowed to cross. Under Standard rules, black must not remove the G21/E20 link yet. So the link from F21 to H20 is not added, but white cannot prevent black from adding it later. Meanwhile white's E21 is hammered by F21. Black will get past white on one side or the other.
Variation B
29.g18 30.h20 31.e20 32.d22 33.d23 34.f21 black wins.
It looks like white wins after Alan's suggested P17. Here is what looks like a crucial line.
1.s4 2.n7 3.i8 4.j13 5.o14 6.p19 7.m15 8.l17 9.n18 10.m20 11.m21 12.o10 13.l11 14.j7 15.k7 16.k12 17.j12 18.j14 19.n12 20.m18 21.p17 22.o19 23.t18 24.r17 25.q19 26.r20 27.u20 28.t21 29.t22 30.s19 31.s15 32.r16 33.u14 34.q14 35.u10 36.t8 37.v7 38.v14 39.w15 40.t13 41.v13 42.w10 43.w11 44.s6 45.s9 46.p12 47.q8 48.q5 49.o7
So if you will allow me to back up as well, I still think black could have won with O18 instead of J14.
18.o18 19.p17 20.m19 21.t18 22.r17 23.q19 24.p16 threatening a double link to M19
25.o16 26.r20 27.u20 28.t21 29.t22 30.s19 31.s15 32.q14 33.t13 34.t8 35.r12 36.p12
And it looks to me like black prevails.
The puzzle I originally intended can still be saved if we enter this line:
1.s4 2.n7 3.i8 4.j13 5.o14 6.p19 7.m15 8.l17 9.n18 10.m20 11.m21 12.o10 13.l11 14.j7 15.k7 16.k12 17.j12 18.m18 19.l19
This is black's last chance to play J14 which transposes in to the winning lines I explored above. Here is what looks like the crucial variation:
20.j14 21.n12 22.o17 23.k16 24.k19 25.i17 26.g21 27.e20 28.d20 29.g19 30.i20 31.d22 32.h18 33.f17 34.i16 35.f13 36.g13 37.h14 38.i12 39.g11 40.h8 41.g9 42.f9 43.e10 44.l6 45.f8 46.f4 47.g6 48.h3 49.i5 50.k4
Fire away!
The situation as it was played originally
1.s4 2.n7 3.i8 4.j13 5.o14 6.p19 7.m15 8.l17 9.n18 10.m20 11.m21 12.o10 13.l11 14.j7 15.k7 16.k12 17.j12
This is the point where David suggested his smart move:
18.j14
But this is a suggestion where white would win:
19.s14 20.m11 21.q9 22.r7 23.s8 24.t8 25.u9 26.v9 27.o8 28.p6 29.m7 30.l6 31.g9 32.k8 33.m9 34.m18 35.l19 36.o17 37.o16 38.p15 39.q15 40.k19 41.s19 42.r20 43.u20 44.s18 45.u17 46.v17 47.w16 48.t18 49.v18 50.u16 51.u15
This way I try to open up a second double thread in vertically on the right side. I have played out many situations - I am sure this opens up many ways to destroy my strategy for white - go ahead!
Cheers,
Tom.
PS: Great analysis tool, Alan!
At any rate, I have changed my mind about 18.J14. I now believe black should play O18 instead, as shown in my post where I say "...allow me to back up as well..."
But if we look at the variation 18.M18 19.L19 then my fancy solution with J14 works again. See my above post.
How about this (still continueing 18.j18):
19.q15 20.m11 21.s10 22.q11 23.t12 24.s8 25.t6 26.u9 27.q9 28.q7 29.o8 30.o6 31.m9
18.j14 19.q15 20.m11 21.s10 22.t10 now what?
I tried to fix my j14 "puzzle move" by entering the variation
1.s4 2.n7 3.i8 4.j13 5.o14 6.p19 7.m15 8.l17 9.n18 10.m20 11.m21 12.o10 13.l11 14.j7 15.k7 16.k12 17.j12 18.m18 19.l19
But I missed this win for white:
20.j14 21.n12 22.o17 23.k16 24.k19 25.i17 26.g21 27.e20 28.d20 29.g19 30.i20 31.d22 32.h18 33.f17 34.i16 35.f12 36.h13 37.e14 38.h8 39.f8 40.g10 41.e10 42.f4 43.h7 44.h3 45.j6 46.k4 47.k3
So I would have to modify the position along the top in order for my puzzle to work.