White to move and win! No computer please!
Lamoss, 1968
8/4kpPr/2P4n/4PR2/8/6K1/7P/8 w – – 0 1
Chess Daily News from Susan Polgar
- About Us
- Chess Improvement
- Chess Puzzles
- Chess Research
- College Chess
- General News
- Home
- Major Tournaments
- News
- Polgar Events
- Privacy Policy
- Scholastic Chess
- SPICE / Webster
- Susan’s Personal Blog
- Track your order
- USA Chess
- Videos
- Women’s Chess
- Contact Us
- Daily News
- My Account
- Terms & Conditions
- Privacy Policy
1. g8Q Nxg8 2. c7 Kd7 3. Rxf7+ Rxf7 4. e6+ Kxe6 (4. … Kxc7 5. exf7 wins for white) 5. c8Q+ 1-0
1 g8=N+ Nxg8
2 c7 and the c pawn is unstoppable
My first thought is to go ahead and queen the pawn:
1. g8Q Ng8
2. c7! Kd7 (or Rg7 first?)
3. Rf7 Rf7
4. e6 Ke6 (Kc7 5.ef7+-)
5. c8Q
Now, I checked this position in a tablebase to save time, and it tells me black draws with 5. …Kf6, but the position already looked drawn to me due to white’s pawn being on the edge and black’s pieces already near their king.
I have a sinking feeling this is a busted composition.
fsimple but fun
1 g8 Ng8
2 c7 Rg7
3 Kh3 Rh7
4 Kg2 Rg7
5 Kh1 Kd7
6 Rf7 Rf7
7 e6 Ke6
8 C8=Q any
9 Qg8
1. g7=Q NxQ
2. c7 Kd7
3. Rxf7+ RxR
4. e6+ Kxc7
5. e6xR
and White wins.
1. c7 Kd7 (…Nxf5+ 2. Kf4 wins) 2. Rxf7+ Nxf7 3. c8 queens Kxc8 4. g8 queen+ and wins
Doesn’t seem that hard to me, and I’m not that brilliant.
1. Rxf7!+
If 1…Nxf7, 2. g8=Q wins.
If 1…Kxf7, 2. c7 and black cannot stop the c-pawn from queening.
If 1…Kd8 (preventing c-pawn from queening), 2. g8=Q+ Nxg8 3. Rxh7 wins for white.
I took a long look at this problem this morning, and my conclusion is that it is flawed. My line from last night is a bit inaccurate, but the key position is identical. Unless I really am missing completely the right solution, this position is a draw with best play. Below is the line I analyzed last night, but I can now see it is a bit flawed:
1. g8Q Ng8 (Rg7 follows)
2. c7 Kd7
3. Rf7 Rf7
4. e6 Ke6 (Kc7 5.ef7+-)
5. c8Q Kf6
And I consulted a tablebase for this position and it confirmed my suspicion- this is a drawn with best play. The knight, of course, is immune to capture here due to the skewer from g7. While a number of such material positions are wins for white, the rook’s pawn here is not decisive (nor are a number of other nearly identical positions where white has the pawn on other files).
Now, in this line, black has what may be a clearer drawing line:
1. g8Q Rg7
2. Qg7 Nf5
3. Kg4 Ng7
4. h4
If I am missing the win in this line for white, I simply am not seeing it. Continuing:
4. …..Ne6 and this is drawn even if black doesn’t have the f-pawn.
I have been unable to find this problem online, so I will assume until someone can demonstrate otherwise that the author of the problem had the following line as the solution:
1. g8N
The idea is that this forces black to take with the knight at g8 rather than the line I discussed above since white has promoted with check. Continuing:
1. …..Ng8
2. c7 Kd7
3. Rf7 Rf7
4. e6 Ke6
5. c8Q Kf6 which is the identical and drawn position in the 1.g8Q line.
Now, are there other potential first moves for white? White can try to protect the g-pawn:
1. Rg5 Rg7 (Ng8 probably ok, too)
2. Kf4
If white takes at g7, black again forks from f5 to draw. Continuing:
2. …..Rg5
3. Kg5 Ng8
Now, this looks more promising than the previous line where white had the three pawns vs the knight and pawn, but the tablebase tells me this is drawn even when black doesn’t have a pawn on f7! It is surely drawn with the pawn on f7.
My conclusion from last night stands- this is drawn.
I think I have found the solution, but not sure it is a win for white. After managing to queen the c-pawn, with check, White is unable to take the N because the rook can fork K and Q. So then it’s Q + RP vs N+R and I was able to hold this against a computer playing White.
An alternative to g8 as first move:
1. c7, Kd7
2. Rxf7+, Nxf7
3. g8=Q, Kxc7
4. Qxa7, with clear superiority for White
Happy New Year to Susan and everyone!
Now for the loose ends suggested by the other commenters:
1. Rf7?? Kf7! (this or lose)
2. c7 Rg7 (or Nf5 wins too)
3. Kany Rg8 and the c-pawn is stopped.
And finally
1. c7 Rg7
Here, Nf5 might draw, too- it leads to another Q and pawn/s vs R+N ending, but one that is even less clear to me, so is riskier than the simple capture at g7. Continuing:
2. Kf3!
Tempting to try Kf4, but after Rg4+ followed by Rc4, black has the upper hand, though I doubt it is decisive. Continuing:
2. …..Kd7
3. Rf6
White can’t even play Rxf7 now since black can capture with the knight now. Continuing:
3. …..Ng4 (Rh7 looks good too)
And this is surely drawn.
This composition is indeed cooked.
The intention 1. g8=N+ Nxg8 2. c7 fails due to 2… Kd7!=: 3. Rxf7+ Rxf7 4. e6+ Kxe6 5. c8=Q+ Kf6 6. Qxg8 Rg7+!=.
2… Rg7+? loses though: 3. Kh3 Rh7+ 4. Kg2 Rg7+ 5. Kh1! Kd7 6. Rxf7+ Rxf7 7. e6+ Kxe6 8. c8=Q+ Kf6 9. Qxg8+-.
1. g8=Q? Rg7+ 2. Qxg7 Nxf5+
what is the solution?
Anonymous at 1:57 shows a draw result for white which seems to be the best result. His last line explaining that g8=Q actually loses for white is correct as well.
1.g8Q does not lose for white- the two passed pawns gives white enough counterplay to eventually win the f-pawn for a draw.
Yancey Ward: you were as always right. This exercise is a recap from 2008, but in that year, Mario Garcia published the bust 2…Kd7 in EG 174, 10.2008)