ArticlesThe first time I saw a computer playing chess, I was amazed. My parents were visiting some friends, and one of them was in the middle of a game against one of the early consumer electronic games. His being in the middle of a game didn’t really distract him from our visit – the computer took over 10 hours per move!
Years later, my mother bought one in Hong Kong. With about 20 buttons and a 5-character LCD display, it wasn’t bad at first. However, over time, it started to get a little senile. It stopped conceding defeat when it was checkmated. Later, it stopped really paying any attention to your moves. It finally had its life support removed after my most memorable – and novel – opening gambit. My opening move was to move my Queen’s pawn forward seven spaces – taking the black King. The computer considered this rather aggressive, and I thought, unanswerable move for a few seconds, before responding by moving one of its pawns forward two spaces.
My uncle bought a much more up-market model. Shaped like a thick chess board, it had pressure sensors under each position, and red LED to indicate the move it wanted. What’s more, it spoke! A pre-recorded voice would say “Queen takes pawn!” and confusingly “Clarify pawn!” when you hit the wrong button. The technology had clearly improved, because when it got older, it got crabbier. It would get confused and randomly shout “Illegal move!” at inappropriate moments. When it got to the stage where you could ask it for a hint, and it would both recommend and disallow the same move, it was time to move on.
The Commodore 64 provided a more general-purpose platform for chess games.
The canonical Grandmaster Chess would lock up after a couple of moves on my machine – with just the timer ticking over – claiming the computer was still thinking, and fooling me for hours.
The developers of another Commodore 64 masterpiece saved themselves hours of effort by not bothering to validate the human’s moves. Perhaps other humans were trustworthy, but I had fun using my King to take all of my pieces, and then all of the computers pieces, one-by-one, until I picked off the computer’s King, and it conceded.
Still, if you were willing to play fair, it would play fair too – unless, of course, you “castled”. To my horror, the first time I castled, the computer moved my rook, but copied my King. Left with two different Kings on the board to defend, I soon conceded this challenge as too much for me.
But soon it was 1989 and there was a huge leap in the state-of-the-art in computer chess. Battle Chess on a 4.77 MHz PC XT left the others behind. Battle Chess had a gimmick – a 3D board, with hilarious animations of pieces moving and being taken.
Battle Chess posed an interesting chess puzzle: produce a board arrangement where it is white to move, and white can choose any of six different moves to mate-in-1, where each of the moves uses a different type of piece. One you have found that, you can watch each of the different checkmate animations in Battle Chess at your leisure.
My father wasn’t interested in gimmicks – he wanted a simple chess game, and that’s where Battle Chess fell down. Oh sure, it had a simple 2D mode. However, it chose the slightly unconventional colours of red against blue. On my monochrome Hercules graphics card, that was rendered as mid-amber versus mid-amber!
At that point I gave up, and never played chess against a computer again. By 1997, Garry Kasparov wished he had made the same resolve.
Comment by John Y. on March 3, 2010
I wonder if I am setting some kind of record for what Stack Overflow calls necromancy. (Though I must admit, I’m aided by my comment feed and StartTags.com.)
My actual comment regards this:
This is surprising in quite a few ways, giving rise to the following possibilities:
1. Perhaps you said seven because “the pawn starts one row up from the bottom, and the opponent’s king is on the eighth row” but really meant six (classic off-by-one error, perhaps due to too much Python on the brain).
2. The kings face each other at the start, on the same column, so perhaps you meant you moved your king’s pawn.
3. You moved your queen’s pawn, but perhaps you felt it not worth mentioning the one-square lateral portion of the move because pawns normally capture from the adjacent column anyway.
4. The game’s interface was such that you specified moves in terms of displacement, rather than destination, and thus you issued the command to really move your queen’s pawn seven squares forward (only) and through array bound overflow or other logic error, it calculated the pawn’s new position as the starting position of the black king, thus capturing it.
5. The game actually provided an incorrect starting layout, such that the black king faced the white queen.
6. Not a possibility, but rather a musing, since you didn’t mention whether your pawn got promoted: I wonder if the game was designed to handle promotions. This spawns further musings (which I’ll not bother numbering), such as whether it always promoted to queen, or provided some kind of interface to choose your promotion. And if you chose your own promotion, whether you could choose pawn, or perhaps even a piece of the other color, etc.
But even without the conundrum of Julian’s Gambit (Ignored), this post brings back memories. Also, you are the only person I am aware of who had both a Commodore 64 as well as a PC with a Hercules Graphics Card. (I had various 8-bit Atari computers, as well as a Hercules-equipped PC, but never a C64. And I never witnessed monochrome Battle Chess, as I played the native Amiga version.)
Comment by Julian on March 3, 2010
Is there a statute of limitations on blog errors?
I meant King’s pawn, 6 spaces. I didn’t think it through enough 5 years ago. You would think that I would recall the incident more clearly when I wrote this post. Back then, the incident was probably only about 20 years earlier. Now it is closer to 25.
As for the C64/Herc card combo, I think ownership of a Hercules card (and able to recall that) puts us in a pretty rare group already.
In fact, even remembering monochrome screens at all…