The title may be a little extreme, but I got very curious about the concept. I chose to use ICCF games to get a solid pool of computer-enhanced game results. The arrival of multi-core computers and ever-stronger software like Rybka and Houdini have definitely made an impact on the outcome of correspondence chess games. Consider the following data:

Draw percentages by year on ICCF:

Year - games - draws - white wins - black wins - white performance

2006 - 35790 - 39% - 35% - 26% - 55%
2007 - 39152 - 44% - 33% - 23% - 55%
2008 - 42735 - 45% - 32% - 23% - 55%
2009 - 41350 - 47% - 31% - 22% - 55%
2010 - 39430 - 47% - 31% - 22% - 54%
2011 - 12535 - 49% - 29% - 22% - 54%

As you can see, just in the past five years computers have increased the drawing percentage by an astounding 10%! White's expected performance went from an upper 55% to a near flat 54% in that time. Also note how 2007's arrival of multi-core desktop PCs jumped the drawing rate by 5% on the spot.

Aside from all this, I spent many hours reviewing games to get a sense of white's roughly four to five percent performance expectancy. I found this was largely in part due to black's mindset of 'trying to win' rather than going for a draw. Risky opening lines that gave black chances to win also gave white chances to win. Putting it bluntly, on ICCF with computer use, one can practically force a draw from the outset if they so choose to.

I decided to give the theory of forcing a draw from the outset a try myself. I challenged a person from England that had been going on a terror in a small centaur chess league with his new 'beast' PC, sporting an overclocked hex core that was outperforming quad-cores by a factor of 4 to 1. Not only was he undefeated, but he was also winning EVERY game in the tournament they were holding. People snickered when I challenged him, thinking I'd get destroyed. My plan was to simply play the safest opening as black against his 1.e4 that would nullify how powerful his computer was: The Petroff. It worked like a charm; we're in a dead equal position with no life in the game for either side. As white, I chose to play my favorite 4.f3 Nimzo-Indian variation. He used his beast to calculate overnight on each move in the opening. Despite going well past 30 ply, the 'beast' could not give him a winning solution to the opening. I've got a slight advantage in a comfortable position.

Anyway, I feel there are two aspects to the data from ICCF:

1. Computers and software are getting ever so closer to rendering chess a draw at it's ultimate end.

2. Black can almost literally force a draw with the mindset of doing so from the outset in a centaur CC game.

Hmmmm. Maybe that would be good thing, as it would kill centaur chess.

I imagine you'd be plenty relieved as it would remove that gun held to your head. There are many games/hobbies I can't stand, but I'm not about to claim it would be a good thing to end them.

I don't much doubt that achieving "you can always draw" status is on the horizon, though that's a pretty weak sense of "solving chess". We are still practically guaranteed of never being able to even prove that black isn't winning in the starting position.

It remains to be seen whether this will signal the end of chess (switch to Go, perhaps?) or just the end of players strapping themselves to computers as Ed hopes. The view from OTB top-level chess is certainly very hope-inspiring, with the World Champion going down to the KID.

The trend is showing that it's getting harder and harder for white to actually force a chance at a winning advantage. In my example about the Petroff, a multi-core computer is downright guaranteed a draw with black so long as you use an opening book derived from the ICCF database. White CANNOT get anything more than a draw against it under those circumstances. That's why I was able to easily handle the Englishman and his 'beast' computer. He was used to it playing against uber-complicated Sicilian openings and crushing the other weaker computers, but when I trotted out the Petroff, his 'beast' never mustered more than +0.08 on the eval score against my rather outdated computer.

So solving chess doesn't have to be about calculating out every variation. It can merely be based on a simple fact that white cannot break through an opening book derived from hundreds of thousands of CC games starting from 2007 on. As I've shown you from the data, each passing year comes closer and closer to nothing but draws as computers/software gets stronger and stronger. Like it or not, that is a form of solving chess. At least rendering it a hopeless battle on that front.

Out of interest: what about 1. d4?



You can use the phrase "solving chess" that way, but it's not the way it's usually used. Even if no practical way of winning with white were known, that would hardly constitute proof. For example, when some time ago it was announced that 8x8 checkers had been solved, that referred to a proof that neither side can force a win. (Even that was a "weak" solution in the sense that it didn't provide the correct responses to imperfect play, so even someone who possessed the proof could lose a game against someone who refuses to follow the perfect lines).

But yeah, you're right, chess will probably be rendered a hopeless battle in the (limited) sense that computers which are told to go into drawish variations will draw every game. I don't much mind. It has zero effect on my experience of chess as a game, and apparently even the strongest humans aren't that much affected.

I recall something put up by Permanent Brain that even when top computers are pitted against one another that decisive games result. I doubt that computers are perfect calculators.

I'd like to amplify my earlier comment a bit.

First, what this does at most is to rule out playing to draw. If 1. P-K4 P-K4, 2. N-KB3 means a draw against the Petroff, then White won't play that sequence. I have my doubts that this is so. One example doesn't prove much.

Second, what these examples that FBX put up illustrate two likely outcomes. First, people will give up 'playing chess' as centaurs. After all, what is the point of doing that if it will always be a draw? I could see using computers in post-mortem, but not during play. The second outcome is that even 'centaurs' will be force to not play drawish lines because of the nature of ICCF events. When I started playing correspondence back in the mid 90's I discovered that OTB style play didn't work at the master level. Either one had to play like Tal or Petrosian. Today, with computers along FBX's line, a player, if he wants to win the event, must win with Black. If he plays to draw with Black and everyone else does the same, then the event winds up a draw for everyone. So instead of seeing more Petrosian-like play, we will see more Tal-like play. Instead of more draws, I think we will eventually see more decisive games. So correspondence chess will probably go in two directions, to non-centaur play and "Tal-taurs."

The third thing I think will happen is an evolution of human chess understanding. Computers do not play chess; they maxmize programs that mimic human understanding of chess. If chess 'plays out', as FBX says, then somewhere along the line new ideas will come along to alter styles of playing chess.

This is all predicated on FBX's sample size, which may be too small, or too biased. There is no breakout of games played between people who are not using computers, or games where only one players was not using computers. I know I've got about twenty games in there, and I've lost a lot of games to centaurs without knowing they were centaurs. I'm sure I'm not the only one.

Not quite solving chess, but you can say: "I can take Black and draw God".

If I did, and He was angry, would that make Him a sore loser? I mean, it's not like He lost.

Unless God plays d4, c4 or Nf3?