This was a pretty position. My first instinct was to check from a1, but it doesn’t take long to see that black has nowhere to go after that to win. After that, there two seemingly active knight moves-Ne5 and Nf4, but Ne5 is more interesting in that it allows a subsequent Ng4 or Nf3 and prevents Rf7+ in some instances. With a bit of trial and error, I think I have the winning line:
1. …..Ne5 2. Rh2
I will cover some alternatives here and the subsequent move later. Continuing:
2. …..Kg3 3. Rg2 Kf3
Now, black has what should be a winning attack- with and without the pawn. So, what if white tries protecting the pawn again?
If white tries a move like 8.Rd1, black replies with Nf3 trapping the white king in the corner, and it is only a matter of a few moves before white mates him there on either the 1st rank or the h-file with the rook. On a move like 8.Rg8/g5, black cuts the rook off from access to g2 with Ng4 and white will have to sacrifice the rook for the knight just to prevent Ra1 being mate (from g4, the knight covers f2!). 8.Rf1, preventing an immediate Nf3 is best, but still losing. Continuing from move 8 above:
8. …..Ra3 (again threatening Nf3) 9. Kg1
White might try a trick here of 9.Rg1 with the idea of playing 10.Rg3 in reply to Nf3 since Kxg3 is stalemate. However, instead, black just plays a waiting move like Rb3 to make white move his rook off of g1. Continuing:
9. …..Kg3 10.Rb1
Or [10.Kh1 Nf3 11.Rd1 Ra2-+]. Continuing:
10. ….Nf3 11.Kf1
Again, 11.Kh1 Ra2 with a deadly mate threat. Continuing:
11. ….Nd2 forks and wins.
Now, let’s work backwards a bit since I have left a few holes in this analysis. At move 10, white put the rook on a square it could be forked at on move 11. Did he have an alternative that was better? No.
This move probably is best in the previous line, too, but I wanted show the different possibilities. Here, 10. …Nf3 isn’t nearly as good as in the last line due the inability to capture with the rook on the second rank with stalemating white. Continuing:
11.Rc3
Black plays Nf3 on any other rook move, too. Continuing:
11. ….Nf3 and white will have to take at f3 just to stop Rf2#.
Now, lets return to white’s 3rd move. From the top:
1. …..Ne5 2. Rh2 Kg3
Now, white is facing the threat of Ra1#, so any rook move long the 2nd rank can’t save him, and 3.Kg1 is still mate after Ra1. So what is left? Only the promotion moves a8Q or a8R:
3. a8Q Ra8 and white still faces the same threats. So 3.Rg2 really was the best move white had at this point. So, now lets back up to move #2.
So, given this line above, and the analysis I showed previously, I can reasonably conclude that white had nothing better than 2.Rh2. Now, I need some more time to consider the line where white doesn’t play 4.Rg7 in the main line above.
This was a pretty position. My first instinct was to check from a1, but it doesn’t take long to see that black has nowhere to go after that to win. After that, there two seemingly active knight moves-Ne5 and Nf4, but Ne5 is more interesting in that it allows a subsequent Ng4 or Nf3 and prevents Rf7+ in some instances. With a bit of trial and error, I think I have the winning line:
1. …..Ne5
2. Rh2
I will cover some alternatives here and the subsequent move later. Continuing:
2. …..Kg3
3. Rg2 Kf3
Now, black has what should be a winning attack- with and without the pawn. So, what if white tries protecting the pawn again?
4. Rg7 Ra1
5. Rg1
Here 5.Kh2 loses quickly: [5.Kh2 Ng4 6.Rg4 (6.Kh3 Rh1#) Kg4-+]. Continuing:
5. …..Ra7
6. Rf1 Kg3
7. Rg1 Kh3
8. Rf1
If white tries a move like 8.Rd1, black replies with Nf3 trapping the white king in the corner, and it is only a matter of a few moves before white mates him there on either the 1st rank or the h-file with the rook. On a move like 8.Rg8/g5, black cuts the rook off from access to g2 with Ng4 and white will have to sacrifice the rook for the knight just to prevent Ra1 being mate (from g4, the knight covers f2!). 8.Rf1, preventing an immediate Nf3 is best, but still losing. Continuing from move 8 above:
8. …..Ra3 (again threatening Nf3)
9. Kg1
White might try a trick here of 9.Rg1 with the idea of playing 10.Rg3 in reply to Nf3 since Kxg3 is stalemate. However, instead, black just plays a waiting move like Rb3 to make white move his rook off of g1. Continuing:
9. …..Kg3
10.Rb1
Or [10.Kh1 Nf3 11.Rd1 Ra2-+]. Continuing:
10. ….Nf3
11.Kf1
Again, 11.Kh1 Ra2 with a deadly mate threat. Continuing:
11. ….Nd2 forks and wins.
Now, let’s work backwards a bit since I have left a few holes in this analysis. At move 10, white put the rook on a square it could be forked at on move 11. Did he have an alternative that was better? No.
10.Rd1 Nf3
11.Kf1 Ra2 (threatening Rf2#)
12.Rd2 Nd2! (Rd2 stalemates) wins.
Or:
10.Rc1 Ra2 (threatens Nf3+)
This move probably is best in the previous line, too, but I wanted show the different possibilities. Here, 10. …Nf3 isn’t nearly as good as in the last line due the inability to capture with the rook on the second rank with stalemating white. Continuing:
11.Rc3
Black plays Nf3 on any other rook move, too. Continuing:
11. ….Nf3 and white will have to take at f3 just to stop Rf2#.
Now, lets return to white’s 3rd move. From the top:
1. …..Ne5
2. Rh2 Kg3
Now, white is facing the threat of Ra1#, so any rook move long the 2nd rank can’t save him, and 3.Kg1 is still mate after Ra1. So what is left? Only the promotion moves a8Q or a8R:
3. a8Q Ra8 and white still faces the same threats. So 3.Rg2 really was the best move white had at this point. So, now lets back up to move #2.
1. …..Ne5
2. Rh5 Ra1
3. Kh2 Nf3
4. Kh3 Rg1! (threatening Rg3#)
5. Rg5 Rg5! (white king immobile)
6. a8Q Rg3#
So, given this line above, and the analysis I showed previously, I can reasonably conclude that white had nothing better than 2.Rh2. Now, I need some more time to consider the line where white doesn’t play 4.Rg7 in the main line above.
I managed to calculate the following lines:
1. Ne5 Kh2 (alternatives below)
2. Nf3+ Kh3 (Kh1 Ra1#)
3. Rg5! a8=Q
4. Rg3#
1. Ne5 Rh8
2. Ra1+ (this is why Ne1 wasn’t good)
2. … Kh2
3. Nf3+ Kh3
4. Rg1! a8=Q
5. Rg3#
1. Ne5 Rf7+
2. Nf3 a8=Q (Rh7 Ra1#)
3. Rh5#
1. Ne5 a8=Q
2. Rxa8 Kh2
3. Rg8!
Not Nf3+? Kh3 Rg5 Rg7! leading to stalemate!!
Black has now nothing against Nf3 and Rg3#/Rg1#.
But I am still unshure about:
1. Ne5 Rh2+!
2. Kf1 black rook move in 2nd rank.
I think perhaps I have a line with Rh2+ here:
1. … Ne5
2. Rh2+ Kf1
3. Rd2 Rh5+!
4. Rh2 Rg5!
5. Ra2 Nf3
6. Ra1+ Ke2
7. Ra2+ Ke3
8. Ra3+ Kf4
9. Ra4+ Kg3!
10. any Rh5#
Ne1 if Rf7Nf3 & Rh1# if Kh2Nf3 Kh3Ng5 & Nh7 if Rh2Kg3 & Nf3
simple Nf4 wins
The pawn at a7 can be taken by black but winning with an extra knight is not going to be easy
1. … Ne5
2.Rh2+ Kg3
3.Rg2+ Kf3
4.Rg7 Ng4
5.Rf7+ Kg3
6.Kg1 Ra1+
7.Rf1 Rxa7
8.Rf8 Ra1+
9.Rf1 Ra2
10.Rf8