Saturday, 27 June 2015

Beauty of chess

(Originally posted on Sunday, 4 November 2018)

Since I wrote the post “My opinion about chess” my results have been awesome: 7 wins, 5 draws and only 1 loss. Interesting. Anyway, below there are two examples that show the beauty of chess. The first example is just a position from one of my games.

I was playing black, so flip the board upside-down. Black to play and win:

White has just played 33.Qg6, preventing me from taking the pawn at g3. I played 33...Qxg6 because I was very short on time and I wanted to “simplify the position”. However the chess engine Stockfish 9 (one of the best chess engines ever created) shows that it was a huge blunder and that there was a much better move for black. There were 3 winning moves, but one was clearly the best. Can you find it? It's actually pretty “obvious”, but I missed it in the game. The beauty of chess.

Here's the best continuation:

The second example is a whole game that lasted for 101 moves (202 single moves). It was a blitz game, but with time increment – the clock was set as 3 minutes + 5 seconds per move and it lasted 17 minutes overall!

As always I played a very quiet and safe opening (the London System – the “boring system” – the “old man’s system”). My opponent didn't make any clear blunders (clear for chess amateurs) and he (or she) outplayed me in the middle game because I ended up being a pawn down. My opponent was trying very hard to win the endgame and in fact I made some blunders, but he missed the winning continuations. It was a hard fight, but I managed to get to a basic pawn endgame that I knew was drawn. I just kept blocking his pawn with my king whenever I could or placed my king opposite his king (which is called “taking the opposition”). The crucial rule however is to move your king vertically down when you have to move your king away from the pawn. I was playing white:

Wasn't this very long hard fight beautiful? Sure it was, but only because of the stalemate rule. If a stalemate wasn't a draw then the game wouldn't have lasted 101 moves, but instead I would have resigned already after move 36! Such a game wouldn't be beautiful at all.

Many people say that a stalemate should be a win for the side that made the last move, but I strongly disagree with such opinions.

On the site:
I found a great argument for a stalemate made by someone nicknamed Quickpull:

“If you play a lot of different strategy games, you've probably often found yourself in a position where the game you are playing is already over, yet you are compelled to go through the motions until the game is actually complete. It's really one of the toughest challenges you face when designing a strategy game. How does one define the win conditions so that the game ends at the appropriate time, before the game outstays its welcome?

Chess has what I think it a pretty good answer to this problem. It gives the player that is down an incentive to fight to the very end by offering them a possibility of forcing a draw. It's a possibility that infuses the endgame meaning of its own, rather than simply being the middle game trailing off into boredom. While it's not unreasonable to assert that stalemates being a draw is somewhat unintuitive, I disagree that it is bad game design. On the contrary, its pretty brilliant solution that uses the game mechanics that are already there to make the game engaging from start to finish. Very elegant design.”

A game of chess is often compared to an old-time battle or an old-time war (I prefer the second comparison), so we can think about stalemate in a similar way.

Looking at the history of warfare – the crucial observations about typical old-time battles and wars:
1. Winning a single battle and/or gaining a material advantage didn't usually decide the outcome of a whole war.
2. Old-time battles didn't usually end with a total destruction of an entire army and the losing side managed to covered a retreat of their king (or military leader) to a safe place.
3. To kill or capture a king you had to usually conquer a fortress, not to win a single battle. A small army hadn't been usually able to conquer a capital city with high walls.
4. Some wars/sieges couldn't be won even with a clear advantage of the attacking side – the outnumbered defenders were so stubborn and/or so cunning that the attacking side finally gave up trying to win or made a costly mistake that resulted in a miraculous change of situation.
5. Sometimes even a successful war ended in a peace treaty that gave winners hardly anything!

A stalemate is a perfect analogy to each of the five kinds of historical cases described above. Moreover without a stalemate there would be no endgame finesse at all!

Imagine what would happen without a stalemate – the side with the advantage of a pawn would FORCE a trade of all the remaining pawns and pieces (you can't hide your army without letting your king die) and then would FORCE a stalemate with a lone pawn! Almost every material advantage would be winning! How strange would that be?!

By the way, you can download Stockfish 9 for free here (it's an open source chess engine):

To use this chess engine you have to launch it in a GUI program (GUI = Graphical User Interface). Personally I use the program SCID 4.2.2., but there are several newer versions of this program AND there are also many other GUI programs. The most important thing to me is the fact that SCID 4.2.2. can annotate a whole game AND you can then replay all the correct variations on the screen. That's enough for me. The “Annotate ...” button shows up only after you launch a particular chess engine (at the bottom of the window):

The game with the first position in the present post made me realize that the SCID 4.2.2. analysis “forgets” about blunders that happen right before the end of the game. After I played 33...Qxg6 my opponent either abandoned the game on purpose or lost the Internet connection, because the game ended right there (after 1 minute of “recovery time” used on the website where I play online chess – “”, which is actually the same site as “”). The SCID 4.2.2. analysis at the end of the game showed only the variation AFTER the final move (+6.27 advantage for white), even though it should also report a huge blunder on my side (the score after move 32 was -8.28 advantage for black):

I had noticed that there is something wrong with the analysis and I launched the chess engine again to see what happened. Interestingly if you add just one half-move (34.Bxg6) the annotations are better (they point out the blunder by black), but still not perfect (they don't point out an additional blunder by white). Move-by-move analysis shows that white should have played 33.Rxh5 (-8.67), instead of 33.Qg6 (-18.85). I have never seen such holes in the SCID annotations before, but such flaws are easy to notice anyway:

As you can see in blitz chess games played by amateurs anything can happen – in a span of just 3 moves (precisely 5 half-moves) the position changed from a draw to a win for black, then to an even bigger win for black and then to a win for white. Blunder after blunder after blunder. The beauty of chess.

Current list of the best chess engines is here:

No comments:

Post a Comment