Chess for Game Designers, Part 1
Chess is a fascinating game, both a major part of the Western game design tradition and an outlier in almost every way among other games. I think there's a lot we can learn about game design by analyzing it.
For one thing, chess is an open deterministic game – something that's very unusual among modern designs. The only prominent example I can think of is John Yianni's Hive.
But chess is also a tremendously influential game in the western board game tradition, obviously. Almost every competitive game scene we have today is touched by the history of chess. Swiss tournaments were invented for chess. The Elo rating system was invented for chess.
Most interesting of all, Chess is a game that has existed basically unchanged for 500 years, and it it has a metagame that has continued to evolve over time. Different ideas about the game and styles of play have emerged over the centuries, to the point that you can very nearly tell the era a high-level game was played in by the moves themselves.
This series of articles is not about how to play chess; I'm not a strong chess player, and I'm not an expert in chess theory. I'll probably discuss the broad strokes of strategy, just to contextualize things. Sporadically in this article series, I'll make use of embedded analysis boards, so you can read along with moves. But mainly I'm going to be talking about chess concepts from the perspective of design analysis – almost every explanation of these topics, of course, comes from the standpoint of trying to teach one how to play chess.
What's a chess opening, exactly?
Chess begins with the pieces mostly tucked behind the pawns and unable to move; you can't make direct progress towards attacking the enemy king. This means that the first few turns of the game must be spent in preparation; moving pieces out to better squares and pushing pawns to control more of the board.
The opening determines the 'flavor' of the middlegame; chess players tend to characterize these positions across two axes:
- Balanced vs imbalanced: Are there asymmetries in the position? For example, does one player have more pawns on one side of the board? Did one player keep their bishop while the other traded it off?
- Sharp vs positional: You could think of this as 'tactical' versus 'strategic'. Positional play is about controlling key squares and gaining small incremental advantages simply by making sure your pieces are on good squares. Sharp play is about noticing tactics – patterns of attacks, captures, or checks that lead to winning material or gaining an advantage.
At higher levels of play, players formulate goals around wanting certain types of positions and pursue those goals by making opening choices. For example, players will pursue more imbalanced positions when drawing the game is unnacceptable.
While the space of possible middlegames is vast and impossible to really apprehend, the space of practical opening play is much smaller. So much so that even intermediate players develop a pretty good mental map of the landscape of openings.
White has 20 possible opening moves at the start of a chess game, but most are simply bad. In practice, the vast majority of games at the intermediate level and above start with one of four moves.
Each of those four moves has a handful of common responses by Black, which in turn have a handful of common responses by White. In essence, neither player is really choosing which opening to play; they are both deciding together, taking turns curtailing the possible options. It's almost analogous to the pick/ban phase before a League of Legends match.
To extend the 'picks and bans in a moba' comparison: Imagine if picking your hero in a moba restricted your opponent to a subset of heros they could pick. One of the things that players rank when considering different openings is their relative agency – some openings give the opponent little choice about what the game will be like.
These choices also get filtered through the lens of tournament and ladder dynamics. Draws are common in chess, and it's generally desirable – for one's Elo rating or tournaement result – to draw games when you have the black pieces or are up against a higher rated opponent. This creates an incentive for players to steer the game into 'drawish' positions.
One opening, the Berlin Defense, is well-known for exactly this; it leads to a big exchange in the middle of the board where most of the pieces get traded off, after which point there's not enough material left for either player to really make progress. This has historically led strong-players to play openings that disallow entering the Berlin – like the Réti or English openings mentioned before. But it's also led to a lot of development in the theory of a variation known as the 'anti-Berlin'.
To answer the question of why the chess metagame has evolved and shifted over time so much in spite of the game being unchanging and deterministic: Partly, it's a matter of which openings are well-understood and which are not. Each generation has added more and more to chess opening theory, studying different variations to gain an edge; one of the best advantages you can have in chess is to be "in your prep" – playing a line that you understand and have studied – while your opponent is not, and is having to actually think about their moves over the board.
This, of course, incentivizes players to study and memorize and look for more and more obscure lines of play. Even today, the opening is still the most mysterious part of the game; computers are not as good at evaluating those positions as they are at understanding middlegames and endgames.
But another piece of the puzzle here is simply that openings tend to bring the players' context and expression into the game. Players have personal preferences and styles of play; players have differing goals going into a game – around whether they really need to win or are okay with a draw, for example. This expression is possible because openings are never really that much better or worse than other openings – unless someone makes a significant mistake, you're not really expecting to enter the middlegame already winning or losing.
This relative balance lets this broader decision-making bubble to the forefront, which is why we still see a relatively diverse set of openings being played in top-level play, in spite of how much computer analysis has changed the game.