Kasparov, Deep Junior draw game four

X3D World

FIDE

Reports

Schedule

Live coverage

Score

	1	2	3	4	5	6	total
Garry Kasparov	1	½	0	½			2
Deep Junior	0	½	1	½			2

Junior Shows What It Knows in Game Four

Two-thirds through the "First Official Man-Machine World Chess Championship" (FOMMWOCC?) the match between Garry Kasparov and Deep Junior is tied at 2-2 with a win apiece and two draws. Kasparov held off Junior with the black pieces today, although not in the effortless way he might have hoped. After a brutally abrupt loss in game three Kasparov no doubt wished for an easy draw before his white in game five on Wednesday.

Garry Kasparov brought Deep Junior into a closed, maneuvering position in game four. The ..e6 Sicilian from game two turned into Hedgehog, one of the few systems in the chess encyclopedia that is fully comprehensible from its name. (The Giuoco Piano is rarely quiet, the Queen's Gambit often isn't one, and the Anti-Meran has nothing to do with King Philip II in 1196.)

In typical Hedgehog fashion Kasparov put all his pieces and pawns on the first three ranks. He also tossed in an odd knight hop on move eight in order to get Junior out of its opening library and thinking on its own. From the thousands of games that had reached the position after 8.Na3, no one had ever played 8...Nd7, which is remarkable. In the post-game press conference Kasparov joked that it was a "brilliant move" before continuing, "Okay it's a lousy move but it got the computer out of book." In hindsight that might not have been such a great idea.

In the maneuvering session that followed, Junior only occasionally exhibited the lack of comprehension of the position that you might expect. The board was full of pawns and the position required artifice and planning, not brute force and tactical wizardry. In the olden days this would have been dark clouds on the horizon for the computer player.

With open lines and free piece play a program's powers of calculation are in full effect. But with a dozen or more pawns on the board and no clear targets machines tend to drift, either making useless waiting moves or embarking on Quixotic adventures that can easily backfire.

Junior did a little of both of these things from moves nine to 22, but most of the time it played competent, intelligent chess. Kasparov wasn't exactly burning with ambition either, although that's exactly how you have to handle the Hedgehog on both sides. Black curls up and lies in ambush in case White overpresses or lets down his guard. White proceeds with caution, snipping off the points of the quills one by one until Black is left defenseless against a breakthrough or is crushed under the pressure.

A few of Deep Junior's moves drew laughter from the Grandmasters because they were clear signs that the program had no idea how to continue. In particular, 12.h3 and 18.a3 were the sort of moves a beginner makes because they don't mess anything up. They also make chess programmers wince visibly. Computers can't plan, of course. They must see every move in sequence, beginning from the position on the board. They cannot visualize strategic goals or fantasize about how to achieve them.

A human can look at a position and think, "Wouldn't it be nice if I could get my knight to the b4 square? It would be very strong there, how can I make that happen?" A computer thinks, "I go there he goes there I go there he goes there I go there he goes there I go there and my evaluation goes up to +0.21." If that analysis reaches as far as achieving the human's visualization, the same end can occur by very different means.

Top programs like Junior can perform that analysis so fast and evaluate the results so well that they can simulate human planning. This is really put to the test in closed positions in which human visualization is superior to computer calculation. In game four Junior showed us that this gap is getting smaller all the time.

While Junior did drift considerably, Kasparov was not interested in challenging the machine by attempting to wrest the initiative with black. He preferred to wait and see what the Israeli program would come up with against his solid but spiky defensive wall. Unfortunately for Kasparov, Junior came up with enough to gain a serious, perhaps decisive advantage.

The crisis finally came with (diagram) 24.a5!? bxa5 25.b5. This maneuver is illustrative of Junior's uncomputerlike disdain for material. It gave up a pawn in order to plant another on the advanced b6 square where it inhibited Black's pieces and created various tactical possibilities. And since I have been unable to duplicate 24.a5 with any of the various top programs (including Deep Junior 7) it also shows what a special beast this Junior is. World #3 Vishy Anand was following along at Playchess.com and he immediately said that he had been wondering if this ambitious push was possible. But no one would expect this from a computer, at least not one that plays like a computer!

When a program evaluates a position almost all of the weight comes from material balance. "Who has more stuff?" the computer asks after calculating each move. This sounds a little brain-dead, but it asks this question a few million times per second! Along with that fundamental calculation, programs add in pieces of knowledge, the sort of thing humans use: space, king safety, passed pawns, etc.

So when a program plays a move that involves giving up material for positional considerations it is betraying its very nature. Junior does this so regularly, and so well, that game after game you can see that co-programmer Shay Bushinsky said quite a lot when he said in a press conference this week that the difference between Junior and other top programs was that Junior gave less importance to material.

Before we go too far we should point out that just because a move isn't computer-like doesn't mean it's the best move! But Junior is consistently good at justifying these curious decisions. Its 24.a5 in game four was just such a move. The commentators soon turned against Junior's aggressive decision, saying that the passed b-pawn would eventually be surrounded and chopped down, leaving Black with the advantage. 35 moves later when the draw was agreed the pawn still stood on b6 like a statue in the Louvre that had seen the French Revolution take place around it. If Junior wasn't completely right, it wasn't wrong either.

Not that Junior played flawlessly. It was hasty in entering an endgame that looked good but that most GMs would quickly decree a theoretical draw. Kasparov admitted after the game that had Junior kept the light-squared bishops his position was probably lost. Instead, Junior traded the bishops and took 20 moves to realize that its extra pawn was worthless. (Actually it never realized this at all, but operator Shay Bushinsky acknowledged that White was not making any progress and offered the draw.) Kasparov showed he could also sacrifice pawns by forcing the queens off the board with 47...R5c7 instead of the prosaic 47...Rb8, which might have let Junior keep all the heavy pieces on the board.

It was a strange and complex game. Commentators Ashley and Seirawan took turns admitting they had no idea what was going on and no idea what the players were doing. Several guest star analysts like Joel Benjamin, Jennifer and Greg Shahade, and Susan Polgar were also brought up to the stage to confess that they had no idea either. They often wondered if the players had any idea themselves! Several redundant-looking knight maneuvers were later declared essential while others stayed redundant. Kasparov was happy that he had managed not to blunder in a tricky position with under 10 minutes on his clock before move 40.

Junior showed that it could make something from nothing instead of foundering when there were no targets to aim at. It made a few pointless pawn moves only to follow them up with useful ones. Most importantly, it showed that Kasparov couldn't just set up a defensive wall and take a nap. Junior pressed him for almost six hours, although the last 30 minutes were purely academic.

Four games have passed and all four have been rich in chess, a tremendous ratio. As Kasparov said after the game, flashing his poetic side, "I wanted to play normal chess, not anti-computer chess. It's like normal chess with a few small modifications. It's a game with the same ball and the same net, but you have to compensate for the wind!"

He will have the white pieces for the final time on Wednesday the 5th. He has had Junior under intense pressure in just a dozen moves in both games with white so far, so he has every reason to be optimistic. But Amir Ban pointed out that they have their only win with black, so they also have grounds for optimism.

Replay the game here