Great Leaps Forward

I'm interested in the history of popular sports and games--how they evolved both in terms of the rules and the strategies employed. Modern spectator sports are for the most part young enough that we know exactly when certain innovations occurred, and why. In American football, for instance, the forward pass was introduced in the first few years of the 20th century because the massive scrums and collisions that were occurring at the line of scrimmage resulted in a number of serious injuries and even a few deaths, so the forward pass was meant to spread the players out over a wider area.

That was a change in the rules. Besides rule changes I'm interested in strategic insights. Something that once it was discovered, everyone realized that the previous way of playing the game was suboptimal, and that there was a different way of playing that got better results. In baseball, for instance, it's now recognized that a batter drawing a walk was not only a skill by the batter (for decades it had been assumed that a walk was something the pitcher did and the batter had nothing to do with it), but that a batter's ability to get on base in any way, including a walk, was underappreciated and undervalued. Once this was accepted, teams started to value players who could draw a walk, much more than they had previously. It changed the way the game was played.

In chess it's a lot harder to figure out what the big advances have been in chess thinking since the rules stabilized in more or less their current form since most of the advances occurred hundreds of years ago, literature about chess was rare and much of what did exist did not survive. Nevertheless, just because we don't know when some of these things happened doesn't mean they weren't revolutionary. Just as a thought experiment I figured I'd list some of the things we now know about chess that must have been Great Leaps Forward at the time.

The Relative Value of the Pieces

One of the first things we're taught when we learn how to play chess is the relative value of the various chess pieces. We learn the "Reinfeld values" that if a pawn is the basic unit of value, knights and bishops are worth about 3 pawns, rooks 5 and the queen 9. Even though this is inexact, it's essential information that at one time presumably was not generally known. How many games had to be played until it became common knowledge that it's usually better to have two minor pieces instead of a rook? What an advantage it would have been for a player to know, when others did not, that he could give up his queen if he got three minor pieces in return. Once the relative value of the pieces became common knowledge, the quality of the average chess game must have increased considerably.

A less revolutionary but more contemporary bit of understanding along the same line is the value of having two bishops when the opponent does not.

The Center

Control the center, occupy the center, dominate the center! After you learn how the pieces move this is the first thing you read or are told about how to play a good game of chess. Once you have some experience this seems obvious, but if you watch many beginners' games you will see that it isn't, at least at first. They'll move wing pawns and try to get their rooks out by going forward. They'll keep their king in the center of the board and not move the center pawns, figuring the king will be safer that way.

The earliest manuscripts we know of that discuss specific chess moves already understand the importance of the center, whether they mention it or not. The fact that the importance of the center often isn't mentioned is probably an even better indicator that it was already an accepted part of the game, so well known that it didn't need to be emphasized. Just as it's one of the first things a beginning player learns, it must have been one of the first generally accepted concepts in understanding the game.

It wasn't until the rise of the Hypermodern school headed by Nimzovich and Reti in the 1920s that it was understood the center could be influenced and controlled by pieces at a distance, and did not necessarily need to be occupied by pawns. The Hypermoderns didn't prove that the old way of playing was incorrect, just that there was an alternative to the conventional openings that had dominated chess for hundreds of years.

Positional sacrifices

The sacrifice in and of itself probably can't be regarded as an innovation; anyone can understand the idea of letting a piece be captured if you force checkmate as a result, or you "lose" a piece to get it back with interest a few moves later. The long-term positional sacrifice, on the other hand, must have made quite a stir. Why would you purposely give up material for some sort of ill-defined "compensation?" The practitioners of such sacrifices must have had to win many games before everybody was convinced that sometimes it's actually good to give away material on purpose.

These days we're familiar with the Romantic era of chess in the 19th century when gambits and similar tactics were employed regularly, so it seems like an old-fashioned way of playing. But I'll bet at one time it was the other way around and the gambit was looked upon first as pure craziness, and then accepted as a legitimate way to play, possibly even the most effective. Apparently it didn't take very long; the King's Gambit, for instance, is one of the oldest known openings in chess, going back more than 500 years.

Pawn Structure

"Pawns are the soul of chess," quoth Philidor. In the mid- to late 18th century Philidor recognized that the way you had to play depended upon the pawn structure that was created in the opening and early middlegame. The way the pawns were set up was a guide that told you what to do. This is still an important aspect of proficiency at chess, one that I personally and a lot of weaker players struggle with.

Endgames

There are a huge number of specific pieces of knowledge about the endgame that are now generally known by all experienced players: the opposition; zugzwang; which pieces can force checkmate without any pawns, and how to do it. There are specific endgame positions such as the Lucena and Philidor. There are a lot more endgame concepts that are understood in theory, but difficult for weaker players to execute in practice such as mating with knight and bishop or winning with queen against rook. There are too many to list, many that I don't even know, but that are known to all master-level players. It's interesting to think that at some point in the past these things were not known and you couldn't just learn them from looking them up in a book. Someone discovered them by sitting down and making detailed notes of various positions and trying out every line of play they could think of until they knew the truth of the situation (Philidor was notorious for discovering things like this; he literally knew things about chess that no one else knew).

Clocks and shorter time controls

Prior to the introduction of clocks in the late 19th century a player could take as long as he wanted to make his move, resulting in marathon games that sometimes lasted all day and still had to be adjourned and resumed later. Over the decades time controls have become shorter and shorter, making time management an important part of competitive chess. Even more importantly, shorter time controls have basically eliminated the existence of adjournments which once were an important part of chess, particularly at the highest levels where top players would spend hours going over an endgame position where a game in progress had been stopped. At lower levels there were also adjudications, where if a game could not be finished in a reasonable amount of time but there was no provision for an adjournment and resumption of play later on, a strong player would evaluate the last position and determine whether one of the players had a winning position or the game was a draw, and the result that was determined by the adjudicator became the official result of the game. This is unknown now.

You could make the argument that one of the reasons the world's best players have to memorize so much opening theory is because they can't afford to waste clock time figuring out over the board all the implications of various opening lines. If you use too much time in the opening you won't have it available when those critical moments occur in the middlegame and endgame where one move means the difference between winning and losing so you have to get it right.

Engines and Artificial Intelligence

For decades chessplayers were convinced that a computer could never beat a top human player, that there was a subtle art to understanding chess that brute-force calculation machines would never be able to crack. This was still a widely held belief when I started playing chess in the 1990s and it stunned the chess world when Deep Blue beat Garry Kasparov in a six-game match. Even before it happened it seemed pretty obvious to me that it would happen eventually; the calculation powers of computers were so much greater than humans, and the gap only increased with time, that it was inevitable that brute-force calculation would defeat the more subjective human understanding of how to win a chess game. A lot of people were very upset when this threshold was crossed, and computers became better chessplayers than human beings. Personally I never understood this; a car is a lot faster than a human being but we still have foot races and still care who wins even though cars are much faster than any of them, so the fact that computers play chess better than people doesn't lessen my enjoyment of chess at all.

The influence of computers on chess is a lot deeper than that, however. Computer engines stronger than any human are available for free online to any level of player, and the strongest players in the world use computers extensively in study and preparation. The result has been that there is no longer such a thing as an "unclear" move; the computer will tell you if the move was actually good or not. Using engines so much in their training has resulted in the top players playing increasingly like the engines, and engines instantly know the proper way to refute creative sacrifices, even if it takes a series of "only moves." It's unlikely that we will see a magician like Mikhail Tal ever become world champion again--after all, many of his brilliant sacrifices were later discovered by the engines to be unsound.

Neural-network engines like AlphaZero and Leela have stepped up computer analysis of chess even further. The machines have already demonstrated to grandmasters the value of playing h4 (or h5) in the middlegame and it is becoming commonplace as a result. As engines continue to get stronger they will almost certainly come up with other new ideas that humans have not.

Conclusion

There's a temptation to think that all the big innovations have already been made, and nothing that is discovered about chess now will be truly revolutionary. I suspect that's true at the level of the common player, but it's been true for hundreds of years already. At the very highest levels of chess (human and machine) the evolution continues, and probably always will. Even if the engines eventually crunch out the impossibly large but finite number of possible chess games and conclude that indeed every game should be a draw (or, less likely, that White should win every game, or even less likely, that the starting position is zugzwang and Black should win every game), the humans can't remember it all and will need to use the engines to give them shortcuts and useful strategies for beating other humans. If the innovation ever completely stops under the current rules, we could always change the rules and let something like Chess960 become the standard version of chess. It could happen; the very best players in the world have complained for decades that the game as it is requires too much memorization of opening theory. But that day, if it ever comes, is still a long way off.