Have Engines Killed Chess?
Exploring the claim that engines have killed chess.

Have Engines Killed Chess?


Part of a comment from a now closed account which expresses the thoughts and feelings that many Chess players seem to have about engines. 

I don't use computers before a game, only after. And so many times I analyze a game where I thought I played brilliantly only to find out I blundered twice and made four mistakes. And even at the top level it's pretty much the same, only slightly less so. You can put a Tal or Fischer game on an engine and find mistakes and inaccuracies every time with the occasional blunder. It's like, you can't even be proud of your performance anymore because the computer always shows you up. It's gotten to the point where I cringe when putting a game I'm proud of on an engine.

I personally think computers have taken the soul out of chess. I think they should be banned. You can't even have adjournments anymore because of them. They've completely ruined chess in my opinion.


What do engines bring to chess? Essentially, deeper knowledge. Can deep knowledge kill a game? Of course. Consider tic-tac-toe. It's quite dead, because humans are deeper than it. If humans could not see a few steps ahead, the game would remain deep for them. The deepness of a game is relative to the player. 

Can a game be killed (made uninteresting or boring for humans) if it is mastered or even solved by an engine? Checkers has been solved, and yet people still play it. Even if an engine solves a game, that solution was already there before anyone discovered it. The fact that some other player knows the solution to the game you are playing tells you nothing during the game. Tal himself was of course aware that many of his sacrifices posed practical difficulties for his opponents without necessarily being theoretically best--what player thinks their play is perfect? All engines do is give us a deeper understanding of what works, and what doesn't work. This will necessarily lead to moments where human beings do not feel flattered. But this is to be expected--it's part of learning and it's a very human practice. Players have always critiqued themselves and one another. We should not expect the engines to heap humans with praise--they don't have the social skills for that--just like calculators don't praise the best human mathematicians. Calculators are like all machines very stingy with their praise. Of course we all need a little gratuitous praise once in a while. This might help. 

Games In Principle

Abstract strategy games are fun because they are deep--deeper than humans. Chess, being one of these, is fun because it's so deep and difficult. Not knowing the best moves is the whole point--it's what makes the game possible. Chess is predicated on ignorance. But deepening knowledge is also part of the fun--otherwise why play more than one game? 

When players lament the absence of the King's Gambit at the highest levels, for instance, what is that they are really nostalgic for? To be charitable, they might be nostalgic for "soul" as the comment above says. Less charitably, they are nostalgic for a time when things were less understood--unfamiliarity. To be nostalgic for unfamiliarity is a mistake--a desire to go backwards. 

When people say that in Chess there is too much theory, they might as well be saying that there is too much knowledge. If there is too much knowledge of chess, then with Chess there is nothing left to learn. In principle, this could describe a real state of affairs. The proper response is not to pine for the days when we didn't understand Chess as deeply as we do now. When children reach a certain age, they don't yearn for days of competitive tic-tac-toe. 

In Crazyhouse Chess, where the pieces which you capture become yours to drop on the board, there is a once-popular line now considered refuted by computer analysis. 

Many crazyhouse players yearn for this line to be viable. Confused nostalgia for ignorance! "Why can't we go back to the days when we could have fun with unsound moves?" And calling it romanticism doesn't help--it just makes the word sound bad. 

Cultural Shock  

The thing with computers is that they are not persons. We anthropomorphize them. We think they are "cold" and "heartless". Because "human vs machine" sells so well. Machines are not agents. Calling them "silicon monsters" or something implies an agency they don't have. As far as I can see, there is nothing in the quoted comment above which offers any reason to think that engines have killed Chess. When you have no way to convey meaningful information, you can talk about "soul", or magic, or authenticity. But "soul" can mean anything, because it refers to basically nothing. If you say something is "good" you can usually offer some criteria according to which that claim might be considered. Saying something has "soul" is like saying something is good in a way which can't be explained. 

For instance, the comment about adjourned games--adjourning was arguably a sub-optimal practice. When two players adjourn a game, they may spend their time analyzing with their seconds, who may even do some of the analyzing for them while they sleep, sometimes in teams. So the quality of the moves will go up, since the decisions are made collectively and players may check one another's work. There is more time to spend on the analysis which increases depth. And the work is not necessarily the player's own. So losing adjournment is not necessarily a bad thing. 

My diagnosis here is that humans are collectively shocked--we have been recovering from it since 1997 when Garry Kasparov lost to Deep Blue. It has affected our self-image as humans. But it needn't have. People will let practically anything affect their self-image, it seems. Imagine being upset that the sun doesn't orbit the earth, for instance. Most people don't worry too much about that now. 

Alive or Dead?

Since I am very interested in variants, I am a little tempted to repeat the immortal Mike Myers line of "honey, it was ruined when she bought it".

Anyway, perhaps Chess wasn't killed by engines--because it was dead long before their invention. Capablanca wanted in the 1920s to make chess larger and add new pieces--which I would also like to see. The distinctive idea behind Chess is that there are pieces which have different movements and values. This essential idea could be realized in many ways, and some of these ways must be better than modern standard Chess because there are infinitely many possibilities. If engines ever really threatened Chess, humans could always migrate to something too deep for the engines. But I think humans should do that regardless of what engines are up to. gbtami, the founder of pychess.org, once said to me "I have always loved chess in all its forms" which I have come to agree with. I still work at standard Chess. For me it is like the roots of a tree, where the branches are the many variations. But the basic idea here is that you can have the fun of new problems without nurturing nostalgia for unfamiliarity or being scandalized by, or resistant to, new knowledge. 

But engines don't make humans stronger at games unless the humans already know how to learn from the engine. You can only remember pages and pages of opening analyses if you already have the background knowledge to digest it. Even if Chess had been solved, that would only matter if there was some way for engines to convey knowledge of the solution to humans. But the question should not be "have engines killed Chess?" but "have humans outgrown Chess?" Even if engines reveal certain variations which would show that Chess is like tic-tac-toe, drawn with perfect play, knowing that there is a solution is not the same as knowing what the solution is. 

One surprising thing I have heard many times is that Chess engines ruin high level chess because the spectators no longer appreciate the moves. Their evaluation of the game is reduced to a number such as +1 or +0.4. It barely seems worth mentioning, but this isn't a problem with the engines, is it?

Another argument I've heard is that over the board games occur in which all of the moves have been analysed by engines. But there have been games predating engines where the players had already analysed all of the moves beforehand. 

Although I think almost everything I have said here is pretty obvious, I haven't seen anyone else saying it. But perhaps you think I'm completely wrong. Let me know. Thanks for reading.