|| The Background
One of the reasons I think we find games of abstract strategy across a variety of fantasy novels is the shortcut they offer to characterization. Want to evoke a clever, calculating protagonist? Have another character teach them an abstract strategy game. Want to preserve a sense of otherworldliness as you do? Have the game be specific to the creative universe (even if it’s only a slight adjustment as in Harry Potter’s wizard’s chess).
These portrayals neither require nor offer elaborate explanations of rules and strategies. Their plots don’t turn upon gameplay (even if, again like in Harry Potter, they turn upon the results of gameplay), but rather upon the shared meanings we all associate with games of the mind. Contrast them with the final hand of poker in the film Maverick, where the scene’s tension stems from the audience’s specific knowledge of rules, probabilities, and hierarchies.
Tak's narrative purpose aside, however, I think the game itself offers an experience that has little to do with archetypes and everything to do with its particulars: Tak might be a good metaphor for moral particularism, a view of ethics that emphasizes consideration of situations before (though not without) consideration of principles.
Considering this possibility lends itself to some discussion of the game’s complexity, recently the subject of a Reddit thread. I tentatively disagree with a claim made there that “main gameplay strategies” are likely to emerge as the community grows and individual players’ understanding and experience increases. While guiding principles will remain important, my intuition is that, for a few reasons, effective play is sensitive to the arrangement of pieces on the board in a way that far surpasses chess.
First is the question of convergence and divergence. Though I am no chess expert, I’m familiar enough to know that, at high-level play, there are well-known and distinct sequences of moves which nonetheless arrive at identical board positions (called transposition). This is true, not only of the early game, but also of late-game positions—where most pieces have been removed. Relatedly, chess exhibits what is likely a different variety of convergence stemming not from identical board positions with disparate board histories, but from similar outcomes of slightly different moves. There is a strategic attractor that pulls many games toward a checkmate with two rooks (or a rook and a queen) leapfrogging each other as they herd a king to the side of the board. While not impossible in tak, its branching makes these convergences vanishingly unlikely beyond the early game.
Second is sensitivity. While chess strategy, too, is dependent upon the specific arrangement of pieces on the board, I think there is room to speak in shades of grey about the degree of sensitivity, and my intuition is that Tak’s is greater. Consider a rook, poised to run the length of the board for a threat of checkmate to a castled king: if the column is clear, it may matter little (though it may matter a lot) with respect to the threat where exactly in the column the rook is located. Put differently, in chess there may be many different arrangements that are functionally equivalent. In Tak, the downstack composition may be interchangeable if all you need to do is throw its top piece across the board to complete a road, so this is, again, a matter of degree.
A third reason is the route by which a game arrives at its respective complexity. Another claim from the Reddit thread is that chess is inherently complex as a function (at least in part) of how many different kinds of pieces there are. In my experience, this aspect of chess often results in dynamics of play that are similar across many matches. Late-game, there might be significant convergence of strategy associated with the particular set of surviving pieces on each side. Tak’s complexity, more emergent and combinatorial in nature (perhaps because it is additive rather than subtractive) seems less prone to this similarity.
Lastly, it would be easy to imagine different board states with identical movesets in a Tak match (e.g. adding a flat at the bottom of stacks already taller than the dimensions of the board)—at least for that ply; it could still result in functional difference downstream if the top part of that stack is thrown. But the additional endgame conditions in Tak, essentially a countdown, add an additional dimension along which gameplay varies—often in decisive ways—that translates to important functional strategic differences despite those identical movesets.
As a starter for why we might care, I want to offer a passage from social philosopher Nel Noddings’ book Caring: A Relational Approach to Ethics and Moral Education:
“Indeed, I shall reject ethics of principle as ambiguous and unstable. Wherever there is a principle, there is implied its exception and, too often, principles function to separate us from each other… Along with the rejection of principles and rules as the major guide to ethical behavior, I shall also reject the notion of universalizability. Many of those writing and thinking about ethics insist that any ethical judgment—by virtue of its being an ethical judgment—must be universalizable; that is, it must be the case that, if under conditions X you are required to do A, then under sufficiently similar conditions, I too am required to do A. I shall reject this emphatically. First, my attention is not on judgment and not on the particular acts we perform, but on how we meet the other morally. Second, in recognition of the feminine approach to meeting the other morally—our insistence on caring for the other—I shall want to preserve the uniqueness of human encounters,” [Italics original, bold added].
It is this last piece of Noddings’ argument that I think corresponds so well to (at least my experience of) Tak gameplay. If in Tak, as Noddings argues of larger life, there isn’t sufficient similarity between distinct scenarios to formulate rules to govern action, then maybe the game can be an exercise in acknowledging the “uniqueness of human encounters.”
How many families are sufficiently similar to warrant rules for appropriate parenting? How many ecosystems are sufficiently similar to formulate a single set of stewardship principles? How many students are sufficiently similar to generate universally applicable performance feedback? More dangerously, how many people subscribe to an inflated sense of similarity, or a convenient one, that enables us to ignore the particulars in favor of easy “solutions”? Noddings isn’t out to dispense with ethics altogether. Instead, her whole work is a pursuit of a way to offer and receive more effective care in our relationships. She wants us to acknowledge that this care emanates from our familiarity with, and responsiveness to, particulars—situation and people. Maybe if we view Tak as an embodiment of this kind of thinking—or at least an example (even if just a computational one)—we can better apply it to other aspects of our lives.
Postscript: Possible Empirical Support
One way to probe the soundness of the claims that undergird this metaphor would be to analyze the database of completed games. How many games, especially ones above some duration threshold, transpose onto one another? When does the latest transposition occur (so far)?