Albanian Coffeehouse Chess

In many countries, you can find men in cafés, in a park or on the street playing chess. In the Balkan more frequently than in Western Europe. When I was staying in Shkodra, Northern Albania, and walking to the ruins of the fortress of Rozafa (the outlook from there is pictured below), I passed by a group of chessplayers in front of a store who had gathered around two tables with boards.

After having done my sightseeing duties and feeling reminded about this splendid book, I walked back and the same crowd was still busy. They were men of different ages, though definitely over 30, wearing once stylish, but now worn out shoes, suits and jackets. It dawned on me that they were playing real coffeehouse chess although technically, the place where they were playing was no coffeehouse.

But what makes a game of chess a coffeehouse game, apart from the fact that usually, none of the players is organized in a club and the level is not too high? Wikipedia characterizes it as

Adjective used to describe a move, player, or style of play characterized by risky, positionally dubious play that sets traps for the opponent.

My own characterization diverges a bit and can be captured by the following three golden rules:

1. Every game has to begin with 1. e2-e4 e7-e5.
2. The only purpose of the game is to attack, and players are asked to launch their charge as soon as possible.
3. Technical conversion of a material advantage is not an acceptable way of playing.

In addition, some conventions apply. Fingering on the board in order to facilitate calculation is acceptable as long as pieces are not moved. Kiebitzers may suggest moves or reveal threats. Placing a piece on a square with some noise just underlines the strength of that move. Same for insulting the opponent (“This guy is a bit stupid!”).

The above rules shall now be illustrated in some fragments.

White and Black: Two Albanian Coffeehouse Players

1. e4 e5 2. Nf3 Nc6 3. Be2 This is a bit timid for the proper coffeehouse style. Black, however, compensates by means of his response. 3. ...h6!? 4. 0-0 g5!? 5. c3 g4 6. Ne1 h5.

White now remembered about Rule #2, too, and after the moves 7. d4 Lh6 8. de5: Se5: 9. f4, a tense battle developed where Black emerged ultimately victorious.

In another game, the following position appeared on the board:

White and Black: Two Albanian Coffeehouse Players

White, who was actually a player able to calculate three moves deep, rejected the tempting 31. Qg4 because of 31. ...Qe3+. However, after 32. Kd1, White would develop a winning attack by either Qc8+ or Re4. I guess that he saw 31. Rf3, forcing Black to exchange queens sooner or later, but after all, a true coffeehouse player stays faithful to Rule #3 and does not try to win an endgame with an exchange up!

Instead, White played 31. c3?, and after 31. ...bc3: 32. bc3: b4!, Black was back in the game.

In the following fragment, White proved to be extraordinarily inventive in the attack.

White and Black: Two Albanian Coffeehouse Players

The situation looks rather gloomy for White, who is just a piece down with no visible way to reinforce the attack. It is rather Black's attack that seems to prevail. However, perhaps thinking of the old rule to improve the position of one's worst piece---or rather of Rule #2 to attack at all costs---, White found the spirited move

27. Nh5!! [sic]

Nobody noticed, neither the Black player nor the bystanders, that this move was plain illegal. Mezza voce, I remarked that the knight was coming from the square f3. Wouldn't they be able to tell a knight from a bishop?. The White player only shook his head laughingly and pointed to the f6-pawn which was now hanging. Gamesmanship? Or did he fall prey to the same illusion as everybody else?

Black thought deeply and played 27. ...Qc5!. This would have been an excellent move, were it not for the available alternative to declare White's last move illegal. After 28. Nf6:+ Kd8 29. Ne8: (29. d4 Qb5 does not change matters) 29. ...Qe3+ 30. Kc2 Qe2+ 31. Kc1 Qe1+ 32. Kc2 a draw was agreed. The winning attempt 32. ...Qb1+ 33. Kd2 Qb2:+ 34. Ke3 would only have been dangerous for Black.

The final fragment contains some beautiful lines which unfortunately remained behind the scenes.

White and Black: Two Albanian Coffeehouse Players, visibly struggling with the position

Admittedly, this has not really been a coffeehouse game. First, this position emerged from a Sicilian Najdorf with heterogenous castling, violating Rule #1. Second, there was no particularly violent attacking play involved, violating Rule #2. Third, both players played the last moves before the diagram in a materialistic style, violating Rule #3.

Moreover, the position is definitely too complex to be handled with coffeehouse technique---also as a bystander, I did not find it easy. White now did something very natural and decided to push his a-pawn: 45. a4. Black responded in the same fashion and soon, a race emerged. 45. ...e4 46. a6 e3 47. a7 e2.

Perhaps it now dawned on White that the principled 48. Qb8 would be answered by 48. ...Qc5:!, and White cannot take the queen (or the rook) because Black has e1D+ at some point. Checking on h1 first does not win either: 48. Rh1+ Kg8 (48. ...Kg7?? 49. Ne6+!) 49. Qb8 Qc5: 50. a8D Qc2+ 51. Ka1 Qc3+ is a draw by perpetual.

So White played 48. Qb7. 48. ...Qc5: 49. Qf7:+ Kh6 50. Qf4+ now looks quite scary---White threatens to come to e5 or h4 with check---and so Black responded 48. ...Rc5:. White approached the black king with 49. Qf7:+, and now, Black played a move that wins in all lines but in one:

49. ...Qg7?? 50. Rh1 mate.

On the way back home, enjoying an excellent espresso macchiato in a proper coffeehouse, I asked myself how to evaluate the position after 49. ...Kh6 50. Rh1+ (what else?) Kg5. White can never promote his queen, e.g., with 51. a8Q, because of 51. ...Rb5+ and mate is approaching. And 51. Qe7+ Kf5 52. Qc5: Qc5: 53. a8Q Qb4+ followed by e1Q leads to a superior endgame for Black. So it looks like a draw after the perpetual check 51. Qe7+ Kf5 52. Qf7+ Kg5 53. Qe7+. Can any side deviate? The computer says no and displays the usual “0.00”. If Black tries to run out of the checks, he gets mated by White's heavy pieces.

This raises the question of whether White missed a win somewhere. The position is given as equal with the pawns on a7 and e2 (the second diagram). But the engine shows that Qb7 was winning in any earlier move. The pointe is that with the Black pawn on a different square, White has Qf7:+ followed by Qf8+, winning the rook on c5. (In the game, this is not possible because Black threatens promotion on e1.) This is actually the only working winning idea. I conclude with some sample lines:

45. Qb7! Rf8 46. Ne6 Qd3+ 47. Kb2. The checks will soon be over, and then f7 will fall.
45. Qb7! Rc5: 46. Qf7:+ Kh6 47. Qf8+! (Rh1+ does not lead no mate unless the Black pawn is on e3!) Kh5 48. Qc5:+-

After all this analysis, I feel like playing in a coffeehouse again. In a nice ambiance, with coffee that is as great as in Albania, and with 1. e2-e4!