A few days ago,
I presented this then-current Chess Cafe endgame study (
now archived). One V. Kovalenko is the author, and it's White to move and allegedly win.
The intended solution is
1.Ne7 Rh5 2.f7 Rh8 3.Ng8 Rh1+ 4.Ke2 Rh2+ 5.Kf3 and wins, but as I realized when trying to solve it, Black draws very easily with
1...Ra5. At first I hoped this could be fixed by some simple expedient like a black pawn on a6, but not even that will work. When I checked my refutation with the computer, it found a second cook with 1...Rc4. In short, the study, assuming it was properly represented on the CC website, was a real lemon.
My attempts to find the original in my own sources didn't work, and my call to all of you to look for me came up empty as well. That's too bad, but on the plus side, "Yet Another Chess Blogger" came up with
his own version of the Kovalenko study:
Again it's White to move and win, and this one works. The key is pretty easy to find, as are the winning variations, but it's still a nice job. For one thing, as already mentioned, it works. Second, the different variations force White to find the only winning play on move two as well. Third, the variations display an impressive diversity of winning ideas. Maybe it's not a masterpiece, but it's an excellent job, especially for a quick "fix"! (One could offer a similar but slightly less attractive patch with the White king on d3, d4, e4 or f4.)
A disappointment I have with the Kovalenko idea (ameliorated somewhat by YACB's fix, which I like and respect more and more) is that one "smells" Saavedra themes in the air, but that expectation isn't fulfilled. For those unfamiliar with the famous Saavedra position, here it is (White to move and win):
The position has a long history (see
here and
here) and Saavedra was the last link in the chain and/or the man with the best press agent, so it's named after him. Try to solve the position first if you don't already know it (be patient!), and then continue.
The first move is obvious:
1.c7, and now Black must start checking with
1...Rd6+. The next few moves can be easily found by a process of elimination: White can't go to the a-file because of Rc6, can't play Kb7 because of Rd7, and can't play Kc5 because of ...Rd1 followed by ...Rc1(+). So White plays
2.Kb5, Black has nothing better than checking with
2...Rd5+, and the next few moves reiterate the same ideas:
3.Kb4 Rd4+ 4.Kb3 Rd3+ 5.Kc2. Now black's rook lacks room to get under the white king, but there's another nice trick:
5...Rd4! Now 6.c8Q walks into 6...Rc4+! 7.Qxc4 stalemate, and 6.Kb3 Rd3+ is a repetition of moves. If you saw this, you did well, but it's not the last trick. White should promote, but to a rook:
6.c8R! White threatens Ra8, mating, and after the forced
6...Ra4, the double attack
7.Kb3 wins.
This solution, first published in 1895, has found plenty of followers. Here are three: two from Troitzky (of "Troitzky line" fame in the two knights vs. pawn ending) and one from Liburkin.
Troitzky 1911
Troitzky 1912
Liburkin 1931
It's White to move and win in each case. Again, when you're ready to see solutions and discussions, proceed to the next layer.
The first two studies are a pair - the second is just an elongation of the first - and the last offers its own motifs and even transposes into the conclusion of the original Saavedra.
So: Troitzky 1911 is solved as follows:
1.e7 Rh6+ 2.Kf5 Rh5+ 3.Kf4 Rh4+ 4.Kf3 Rh3+ 5.Kf2 [5.Ke2 is faster, a correction he made in his published solution to the 1912 study]
5...Rh2+ 6.Ke1 Rh1+ 7.Kd2 Rh2+ 8.Kd3 Rh3+ 9.Kd4 Rh4+ 10.Kd5 Rh5+ 11.Kd6 Rh6+ 12.Nf6! Rxf6+ 13.Kd5 Rf5+ 14.Kd4 Rf4+ 15.Kd3 Rf3+ 16.Ke2 and wins, as Q vs. R is a theoretical win.
The up-and-down motif adds a certain humor and charm to the study, and it also has the virtue (one lacking in Kovalenko's version and its fixes) of getting rid of the superfluous knight (superfluous, that is, from the standpoint of the final position - one doesn't need both the queen
and the knight to defeat the rook). The second study is like unto it:
1.dxe6! Rc1+ 2.Kf2 Rxh1 3.e7 Rh2+ 4.Kf3 Rh3+ 5.Kf4 Rh4+ 6.Kf5 Rh5+ 6.Kxf6 Rh6+ and now we've reached the position after move 1 of the previous study. The solutions are the same from here, except that (as mentioned above) Troitzky doesn't waste a tempo on Kf3-f2-e1-d2-d3, opting instead of the direct Kf3-e2-d3.
Finally, Liburkin's study has three neat points. First, that it offers an attractive prologue to the finale of the Saavedra position. Second, that its variations utilize different underpromotions. And third, there are some very nice false leads. Here are the lines:
1.Nc1! and now:
(A)
1...Rd5+ 2.Kc2! (2.Nd3? is a thematic move - see variation (B) - but it doesn't work here after 2...Rxd3+ 3.Kc2 Rd5! lets Black sneak out with a draw)
2...Rc5+ 3.Kd3! (3.Kd2? Rxb5 4.c7 Rb2+ 5.Kd1 Rc2!! 6.Kxc2 stalemate)
3...Rxb5 4.c7 Rb8! 5.cxb8B!! wins.
(B)
1...Rxb5 2.c7 Rd5+ 3.Nd3!! Rxd3+ and now we've returned to Saavedra:
4.Kc2 Rd4! 5.c8R! Ra4 6.Kb3 and wins.
Aren't endgame studies wonderful? Perhaps it's clear now why I find the Kovalenko study so disappointing. There are no brilliant moves, no sacrifices, no humorous turns (like the king traipsing up and down the board), and no brilliant false leads. Add to it the superfluous material and two cooks (i.e. two fatal flaws in the solution) and one is inclined to wonder about this guy's bona fides as a study composer!
So I'm going to conclude this spectacularly over-long post with several attractive studies by Kovalenko, and I won't offer their solutions until later.
Kovalenko 1995a: White to move and win
Kovalenko 1995b: White to move and win
Kovalenko 1998: White to move and win
Kovalenko 1999: White to move and draw
Meanwhile,
here's the gigantic collection in replayable format. Phew!
