I suppose the point is that black has to do something about his threatened bishop but a straight-forward bishop retreat allows black to play Ra4 attacking the white queen. Then black is having way too much fun.

Instead 31...c4 creates an immediate attack on the white Rd2 and blocks the fourth rank so that Ra4 doesn't have so much impact. It also shifts the c5 pawn which was making the black bishop mostly bad.

Black doesn't mind the temporary pawn sacrifice because his more mobile pieces will sweep up weak white pawns.

The other interesting thing about this position is that white would really like to double his rooks on the seventh rank. White makes it difficult for him to do that by using his queen and bishop to attack both white rooks. Black has to settle for doubling on the d file which is far less attractive because there is not much to attack from there.

click for larger view

Every so often you solve a puzzle not by finding the clever combination, but by simply asking "what would I play here?". Today was such a case.

I tried to find some clever combination that would allow me to attack the pinned Rd3 or fork the stray Ra7, but without success. Finally I just looked at the position and thought that I (black) would have a winning game if I could make passed pawns. When I finally turned my attention to winning pawns, things began to make sense.

The main pawn that needs to go is Pe3, and that's easy: <31...c4 32.bxc4> opens the a7-g1 diagonal allowing the <32...Bb6>, forking of Ra7 and Pe3, while also attacking Pc4.

click for larger view

Here, I guessed that white might play <33.Ra4> to guard Pc4, and then <33...Bxe3>, giving me a passed Pe6.

click for larger view

This is just awful for white, he is practically paralyzed and my bishop can now to work on white's h & g pawns.

Trying to improve for white, I then considered <33.Rd7> (as played), which frees the king from the defense of Rd3. This leaves me with the choice of taking Pe3 or Pc4. I think grabbing white's passed pawn <33...Qxc4> would be best since it represent's white's best chance to cause me trouble:

click for larger view

It's not clear what white plays now, but I'm satisfied that black is much better. White's isolated e-pawn and distant g & h pawns are poorly defended. With white's K so exposed, the pawns are very targetable with queen forks.

White threatens 32.Rxc7.

White's weakest point seems to be the pawn on h2. This suggests 31... c4 32.bxc4 Bb6, attempting Bxe3 and Bg1 while keeping an eye on the c-pawn. For example, 33.Ra1 Bxe3 34.Re1 Qg2+ 35.Kd1 Bf2 36.Rd2 Qf3+ 37.Ree2 Qxc3 38.Rxf2 Qxc4.

c4 is the obvious way to exploit the rook pin but it is neutralized by the p at b3. c4, pxc4, Qxc4 trades the pin on a R for something less. Qg2+ allows white to rearrange his defenses.

i missed that Bb6 takes a tempo away from the white R if played *before* Qxc4. doubling the rooks is clearly a defensive move.

my incomprehension includes 33. ... Qxc4, when 33. ... Bxe3 seems useful to prepare Qg2+.

but there seems to be lots of room for White improvements all along this sequence

I give this one a diluted thumbs down

yes, i'd call

31. ... c4
32. Ra4 b3+

Black manages to temporarily save the e3 pawn, but only at the cost of giving up the exchange and allowing White to target the remaining King-side pawns.

click for larger view

Komodo-10-64bit:

At this level there is no doubt White saw 16...Nc6 so he basically wanted to exchange the Q for 2 Rs. Why not, some players are more confortable with unbalanced material and analysing the resulting position is above my pay grade. But then he should have activated his Rs instead of doing it 13 moves later. It appeared Black's Q+B were too active and posed threats White needed to deal with first.

The pun:
One of the players is Japanese ❌
The game was played in Japan ❌
It is a 7-move miniature ❌
It was played in 1954 ❌
White committed seppuku ❌
Oh I get it, apart form the location the first letters are the same as A R Saleh Salem 's initials.