Material: B+P for R. The White Kh1 has 1 legal move. Black has a local superiority on the K-side, with Bb8, Bd7, Qg4, and Nh5 attacking both light and dark squares. Only the White Nb6 needs activation. The White has relevant defenders Qe1 and Bg2, with possible aid from Rd1, but Na3 and Nb5 are on another planet, away from the K-side. The White Qe1 defends Pg3, the focus of the Black attack, but the defense is illusory, because Qe1 is overburdened and must defend Rd1 against Qxd1. The candidate is therefore a capture of Pg3. White has no immediate counterattack, because Bd7 covers e8, the entry point for Qe1. In general, Ns are better "local" pieces than Bs, so the first candidate 30…Bxg3 leaves the N on g3. A sample of 30…Bxg3 is unpromising, so we sample 30…Nxg3+ next.

Candidates (30…): Bxg3, Nxg3+

30…Nxg3+

White is best to decline the sacrifice [<according to computer evaluation, acceptance is better>], but then loses a P without compensation and faces a continuing K-side attack in a passive disorganized position. If White accepts the sacrifice, however:

31.hxg3 Bxg3 32.Qd2

The Bd7 no longer needs to guard e8.

[32.Qf1 or 32.Qg1 crowd Kh8 into a mating net; other Q moves abandon Re1 to Qxe1+]

32…Qh4+ 33.<Kg1>

I had a failure of board vision and went for

33…Qh2+ 34.Kf1 Bh3 (threatening 35…Qh1+ 36…Qxg2)

missing that 34…Bxh3 is possible.

With the <last move entered> emphasized, and full computer variations given, so humans can improve at the end (definitely the case below), Toga II 1.3.1 evaluates this line as

[ply 15/45 time 00:32 value (to White) -1.46]

33.<Kg1> Bg4 34.Qd4 Bxd1 35.Qxd1 Qxb4 36.Qd4 Qe1+ 37.Bf1 Qe6 38.Qd3 Qg4 39.Nc2 Qg5 40.Bg2 Qc1+ 41.Bf1 Bf4 42.Nc3 Qb2 43.Bh3

Toga evaluates the game variation as

[ply 15/53 time 01:41 value (to White) -1.33]

33.<Nxc4> Qh4+ 34.Kg1 Bh2+ 35.Kf1 Qxc4+ 36.Qd3 Qxd3+ 37.Rxd3 Bxb5 38.Ke2 Bd6 39.a3 h5 40.Bxd5 b6 41.Bg2 Kg7 42.Ke3 Bxd3 43.Kxd3 h4 44.Kc4 Kf6 45.a4

Hi, <lost in space>. A friend told me this story when I lived in Germany. A German tourist in England told someone, "I become a fish." ("Ich bekomme einen Fisch.") The answer was: "I hope not, sir."

It looked like cg.com fumbled the ball when they said it was "white to move". I guess it's as easy as dropping a baton at a relay race.

:-)

seems that your German is better than my English.

30…Nxg3+

White can decline the sacrifice

(1) 31.Kg1 Ne2+ 32.Kf1

32...Nf4 (threatening 33…Qxg2#, 33…Nxg2, and 33…Qxe1)

33.Rd2

White has lost a P without compensation. Despite approximate material equality, he faces a continuing K-side attack in a passive disorganized position. The variations are complicated, but refusal eventually loses more material than acceptance.

White can accept the sacrifice:

(2) 31.hxg3 Bxg3 32.Qd2

The Bd7 no longer needs to guard e8.

[32.Qf1 or 32.Qg1 crowd Kh8 into a mating net; other Q moves abandon Re1 to Qxe1+]

32…Bf4 (threatening …Be3 and …Qg5, locking the dark-squares down for a mating net)

The game might continue:

33.Qd3 Qg5 34.Nd4 (to bring another defender to plug the h-file)

34.Nf3 Qh5+ 35.Nh2 Bf4 37.Kg1 Bxh2

Candidates: 30 ... Nxg3+, 30 ... Bxg3

A1) 30 ... Nxg3+ 31. Kg1 Ne2+ 32. Kf1 Nf4

A2) 30 ... Nxg3+ 31. hxg3 Bxg3 32. Qd2 Qh4+ 33. Kg1 Qh2+ 34. Kf1 Nc4 ∞

Hi again, <lost in space>. You do not have a chessforum, but I would like to tell you of an experience I had in Germany.

I had a good friend who spoke excellent English. I called him "mein Teufel" (my devil), because he continually teased me about how bad my German was (despite permitting only one person to speak English to me). I remember the one conversation where he broke into English to tell me: "You know, to understand you, I have to translate everything into English and then back into German. Why don't you spare me a step, and just speak English?"

Thanks for showing me his point of view :)

Needless to say, my friend was a tremendous stimulus to my continued improvement in German, and I am very grateful to him. The moment came, however, when I first conversed with his father, and his father asked me, "Are you planning to stay in Germany? Your German is so good!" Out of the corner of my eye, I saw my friend's face fall.

Now, however, there seems to be an ever greater bias. In general the chessgames.com puzzles range from easiest on Monday to insanely difficult on Sunday. So we become programmed to look for concrete wins early in the week and then just good lines later in the week. Again, this is an even worse bias to deal with. We are not only looking for wins, but for "easy" or "hard" wins depending on the day of the week.

All players here should strive to eliminate ALL bias in their play. Forget what day it is. Set up the position, put 15 or 20 minutes on the clock, and try to solve as much as you can under simulated tournament conditions.

And we should especially avoid ALL arguments as to whether this is a perfect puzzle for the day or not. Concentrate on the position. Examine the variations. Strive to discover the truth about the position of the day. That's what we are here for. To learn chess.

Respectfully, many of us have to plan our time ahead.

<Strive to discover the truth about the position of the day. That's what we are here for. To learn chess.>

Ahem, I have to disagree. After plugging the position into my evaluation algorithm, it came out as average Saturday difficulty (among the tractable Sat. puzzles). Note that about half of Saturdays are intractable (in the sense that the Hiarcs engine at highest setting has no clear idea of the best line within 10 seconds).

Another data point arguing for Sat difficulty: I saw the threat of Nxg3 when looking at the initially mis-posted puzzle (one move before the real puzzle position, white to move) fairly easily. Usually I have to spend more time on Saturdays to even come up with a plausible first move.

<DavidD: One of the biggest problems in solving chess puzzles is the very fact that we know it is critical position with either White or Black to play and win.>

Yup, I 100% agree. Perhaps you'd like to lend support to my modest proposal (see my profile for details) for making the puzzles more resembling the real life?

While the yesterday's mistake in the puzzle presentation was a clear SNAFU, I found it quite satisfactory nevertheless. I think anyone who diagnosed the mistaken postion to be won for black, not for white, deserves a good portion of the credit.

However, Black's follow-up must be conducted precisely to secure the win. For example:

If 32. Qf1, Black gets a mating attack with 32...Qh4+ 33. Bh3 Bxh3

If 33. Nxc4!?, White puts up more resistance but Black gets a won endgame after 33...Qh4+ 34. Kg1 Bh2+ 35. Kf1 Qxc4+ 36. Qd3 (36. Qe2 Qf4+ 37. Qf3 Bxb5+ 38. Kf2 Qh4+ 39. Ke3 Qxb4 40. Qf6 Qa3+ 41. Kd2 Bd6 ) 36... Qxd3+ 37.Rxd3 Bxb5 38. Ke2 Bd6 39. Bxd5 Bxd3+ 40. Kxd3 Bxb4 41. Bxb7 f5 .

you are right, so far I have no chess forum. The reason is easy: I am a rookie on CG, just started a few months ago (after a long, painfull and frustrating debate with fritz) and my biggest problem is the time (especially during the week). But I am sure, sooner or later I will estabish my own forum.

I hope you are still in contact with mein Teufel. He can much better explain his point of view than I can.

Here is the Toga evaluation, under my usual conditions:

[ply 15/50 time 00:40 value (to White) -4.52]

29...<Nxg3+> 30.hxg3 Bxg3 31.Qd2 Qh5+ 32.Kg1 Bg4 33.Kf1 Bxd1 34.Qe3 Bh2 35.N7b5 Qf5+ 36.Ke1 Bg4 37.Nd4 Qf6 38.Bf3 Bxf3 39.Nxf3 Qa1+ 40.Kd2 Qxa2+ 41.Kc1

It might be basically the same as a game variation I gave below, but with Nb5-d4-f3-h2 to rescue the White K no longer available. I still think this is a tough position, however, and I do not entirely trust my Mom-and-Pop computation.

My first thought with Black to play was Nxg3, but I like Bxg3 better because it is trickier and if White goes with impulsive Q x B, then ...Q x R mate.

<Jason Frost> You overlooked the White Rook at d1. Look -> 30...Nxg3+, 31. hxg3 Bxg3, 32. Qc3??? (Blunder) Qxd1+, 33. Qe1 Qxe1+, 34. Bf1 Qxf1#. 0-1

32. Qc3 is a blunder so, White's only move is 32. Qd2.

That was so for this puzzle. I figure the puzzle masters are trying to find games that do have a fairly straight line to advantage or victory.

The knight was a fairly easy start. It took out a defender while opening the H file for the queen. White takes the pawn.

What to do with that dangerous white queen? Chase it away with the dark bishop.

Now came a dilemma. Should I check the king now? or am I missing something.

Ah! of course, the white queen is inhibiting black's dark bishop. Any checkmate web would need more working space.

First I thought of the white's dark bishop. That proved a dud right off the bat. I hadn't first noticed how neatly the bishop wards the white queen off protecting it's king.

Other candidate, only the night way up there on the B6.

Down it comes to C4. The white queen offers a trade. I paid no attention, and checked the white king with my black queen. The only place to do that without the trade, is to go to H5. King moves to G1.

The black bishop needs to keep sentinel. I had to think this one out a long time. I knew I could check the white king with the queen, but then what? There is that pesky white bishop. I could not think of a way to continue the attack.

Of course the white king is now at F1. The only square he could flee to.

The white square bishop was poised to join the fray. Ah! The exchange made brings the black queen back into checking the king.

What caught my eye was the possibility of Ba7, controlling g1 with a Greco mate (Nxg3+ and Qh5+). White's chained Ns control a7. But Black can undermine the front N with 30..Nc4.