|
< Earlier Kibitzing · PAGE 2 OF 2 ·
Later Kibitzing> |
| Apr-07-10 | | johnlspouge: Wednesday (Medium/Easy)
Van der Sterren vs Timman, 1981 (28...?) Black to play and win.
Material: Down a P. The White Kg2 has 4 legal moves and is vulnerable to Rf5-g5+ or Rf5-f7+. The Black Kg8 is secured from check. Candidates (28...): Rg5+
28…Rg5+ 29.Kh3 [Kh1 Rf1#]
39…Rf6+ 40.Kh4 Rg1 (threatening 42…g4+ 43…Rh3#)
The interposition Rf2-f4-h4 is both infeasible and pointless, so White can only check: 41.Rd8+ [or Re8+] Kf7 42.Rd7+
After the Black Kf7 captures the R on the e-file, 43.Rxh7 prevents mate, but White is down a R for 2P. Toga prefers 42.Rf6+ as a "defense", but White still must drop a R to avoid mate. |
|
| Apr-07-10 | | johnlspouge: < <whiteshark> wrote: <28...Rg5+ 29.Kh3> (else Rf1#) <Rf3+ 30.Kh4 Rg1!!> > They should have had the Qs on the board, just to make your day :) |
|
| Apr-07-10 | | wordfunph: 28...Rg5+ 29.Kh3 Rf3+ 30.Kh4...
then black is winning but it will take some time.. |
|
| Apr-07-10 | | whiteshark: <johnlspouge> Yes, indeed. But then it's <πάντα χωρεῖ καὶ οὐδὲν μένει>, too. <Everything changes and nothing remains still>, that is from one endgame collection into another. :D |
|
| Apr-07-10 | | desiobu: I can't believe I missed Rg1. In retrospect it's obvious. Black wants to play Rh3# so he has to keep the g file, and play to force Kh4. |
|
| Apr-07-10 | | Patriot: I looked at both 28...Kf7 (attempting to remove the guard on e6) and 28...Rg5+. I chose 28...Rg5+ only because black has nothing to lose and a lot to gain if 29.Kh1. It's one of those "just play it and see what happens" moves. But then I calculated 29.Kh3 Rf3+ 30.Kh4 h6 and also considered 30...Rg1 without seeing the mate potential. I would not have calculated all of this in a 30-minute game for sure. The power of 30...Rg1 just didn't seem obvious to me, but I probably would've played the first move or two and hopefully would've seen the finish with the position in front of me. |
|
Apr-07-10
 | | tarek1: Van der Sterren obviously missed this slow motion mating attack ... but it looks like if he plays, say, <28.h4> instead of the blunder <28.Rxe6> he just loses a pawn with no clearcounterplay.
So I wondered if <28.Ra7> may be a better try, because the combination doesn't work now that White has a back rank mate : 30...Rg1 then 31.Rd8
And if Black takes the pawn with 28...Rxe5, doesn't 29.Rdd7 guarantee white a draw ? |
|
Apr-07-10
| | tarek1: Sorry, Rd8 then just Rf8, but with the seventh rank in white's hands it looks hard to make progress for black. |
|
| Apr-07-10 | | sethoflagos: <OrangeBishop: So where did White go wrong? Was it 23. Raxa6, after which he has no good way of protecting the f2 pawn after Rbf7?> White's position after move 25 is okay so 23. Raxa6 wasn't a bad move. Maybe black has a slight edge on pawn structure. 26. Rd2 isn't good. Pe5 is much more important than Pb2. If he just pushes one of his rooks onto the seventh rank or plays 26. Rxe6, he is safe and sound. The howler though is 28. Rxe6. Several moves are available to counter the threat of 28. ... Rg5+ of which the best is probably 28. h4. |
|
| Apr-07-10 | | sethoflagos: <tarek1:>
What about
28. Ra7 Rxe5
29. Rdd7 Rg5+
30. Kh3 Rf3+
31. Kh4 h6
with Rf2 Rxb2 to come? |
|
Apr-07-10
| | tarek1: <sethoflagos>
I see your point, the king will hide on h7 the rook stay on the g file, I don't see a good plan for white against this.
So it's already tough for white even if he saw the combination. |
|
| Apr-07-10 | | 310metaltrader: One of the keys to this puzzle is how the pawns and rooks are in each other's way. in the middle game, there is really no good way to rearrange pawns and rooks, but in the endgame, you can do so and to great effect. |
|
| Apr-07-10 | | eblunt: < wordfunph: 28...Rg5+ 29.Kh3 Rf3+ 30.Kh4... then black is winning but it will take some time.. > That's not what Van der Sterren thought! |
|
| Apr-07-10 | | kevin86: I answered this one-but still can't figure why... |
|
| Apr-07-10 | | mworld: i didn't see this as winning since I did not see Rg1. The resignation seems premature because its easily conceivable that black could have played Rf5. |
|
| Apr-07-10 | | this is a sign: I got this too. But didn't see the winning combination. It just seemed like the only move :) |
|
Apr-07-10
 | | Jimfromprovidence: <sethoflagos> <The howler though is 28. Rxe6. Several moves are available to counter the threat of 28. ... Rg5+ of which the best is probably 28. h4.> I found 28 h3, Kh1, Kg3, b4 and even Rdd6. |
|
Apr-07-10
 | | chrisowen: White folds, the win going to Jan. Stirring the pair of rooks gives white the victim man. Fit checking the plight of the monarch after Rg5+. The dovetail plan makes sure he leaves the nest.. Kh3 Rf3+ Kh4. Rg1 would have then gone down and crumbles his attack is past. Rye is the hawk (isn't it wry?) who shoots down the flank. You cooked up a nice dutch treat, thanks CG |
|
Apr-07-10
 | | al wazir: As black I never would have played 23...Rbf6, because 23...Rxb3 looks straightforward and winning. (If 24. Rxe6, then 24...Rxb2, threatening mate and the f-♙; if 24. f3, then 24...exf3 25. gxf3 Rxf3.) |
|
| Apr-07-10 | | YouRang: The first few moves were pretty intuitive (my apologies to the intuition-challenged). But 28...Rg5+ is called for because (1) it is forcing, and (2) it pushes the white king up against a wall (where it is particularly vulnerable to mate, especially in the presence of rooks). Of course 29.Kh3 is the only choice (29.Kh1?? Rf1#). Again, intuition suggests 29...Rf3+, if for no other reason than it's check (which always feels kinda good), it's forcing, and it attacks a pawn. Granted, these reasons alone are not the stuff of puzzle solutions, but they do catch the eye. The only reply: 30.Kh4. Now what? Well, our rook on g5 is in take and guarding it with ...h6 doesn't look promising. It does has one safe move where it nicely maintains it's control of the g-file: 30...Rg1. Is there anything promising about this? Yes! the king can't leave h4 lest ...Rh3#, but we threaten to force it from h4 with ...g5! (note that this threat was only became possible when we vacated g5 with 30...Rg1). Black can give check, but curiously it has no satifying way to stop the threat. |
|
| Apr-07-10 | | sethoflagos: <<Jimfromprovidence>: I found 28 h3, Kh1, Kg3, b4 and even Rdd6.> .. Rd3, Rd7, Kg1, Kh3, Ra1, Ra7, Rd1 ... :-) |
|
| Apr-07-10 | | wals: Rybka 3 1-cpu : 3071mb hash: depth 16 :
White's fatal error :-
28.Rxe6 (-7.75), better was h4 (-0.83), or h3 (-0.89). A quickie calculation by Rybka has Black mating 40...Rf4# |
|
| Apr-07-10 | | jackpawn: I hope I was just off form today, but I didn't get it. I missed the Rg1 move entirely. It's a pattern to remember. |
|
| Apr-08-10 | | JG27Pyth: I had Rg5+ Kh3 Rg1... I think this move order still works (although a bit slower)... or did I muff it? |
|
| Apr-08-10 | | johnlspouge: < <JG27Pyth> wrote: I had Rg5+ Kh3 Rg1... I think this move order still works (although a bit slower)... or did I muff it? > 30.Rg2 saves White. |
|
 |
|
< Earlier Kibitzing · PAGE 2 OF 2 ·
Later Kibitzing> |
Kibitzing
