Positional Heuristics in Games

In their landmark Characteristics of Games, George Skaff Elias, Richard Garfield, K. Robert Gutschera & Peter Whitley define positional heuristics as "heuristics that evaluate the state of the game." It's a fairly trivial concept, but one that is often overlooked in discussions about the drama and skill of games.

Heuristics aren't equally clear from game to game. The fuzziness of the players' position can vary quite drastically. For example, let's look at two games with very similar mechanics: Candy Land and Chutes & Ladders. The positional heuristics in Candy Land are fairly clear: being substantially further ahead means you're substantially more likely to win. There is a slight twist—the picture cards can swing a player quickly up... or down.

Chutes & Ladders plays very similarly, but the positional heuristics are much fuzzier. The eponymous chutes and ladders can cause massive swings in both directions, making it hard to judge who is winning at any point.

This gives drama to games, even without agency. Functionally, there is no difference between flipping a coin and playing a head-to-head match of either of the above games. Either way, the winner is random and nothing a player can do within the rules changes the outcome. However, ask most young children which they prefer: the drama provided by the positional heuristics gives a magic to Candy Land that the coin flip doesn't have. (Of course, the colorful candy theme doesn't hurt). Modern gamers, including myself, aren't immune to this drama. Magical Athlete is a game of fairly little agency, but the laughter and drama that comes from the ever-shifting chaotic positional heuristics has led it to be a hit of 2025.

There are a few design choices that can be made to finetune the positional heuristics fuzziness. Some designers will do this by placing resources behind a player screen (or another device) in order to obfuscate: this is called "hidden trackable information." It can be a divisive technique because it adds a memory skill to the game—but the memory aspect isn't why designers use it. The information is hidden behind a screen because the fuzzy positional heuristics make for a better gameplay experience.

Consider the classic Reiner Knizia game Tigris & Euphrates. In the rulebook (at least of my Mayfair edition), there is an explicit callout of which information is hidden: victory points & treasure tokens gained, tiles in the bag, and tiles removed are intentionally hidden from players. These each add some fuzziness to information around who is currently winning, what tiles are mostly likely to come out of the bag, and how much time is left in the game. None of these are completely hidden: you can track the general direction based on play. This helps to smooth some of the decisions in the game which would otherwise be prone to analysis paralysis.

Games often skill-test positional heuristics, sometimes exclusively. Betting games are a good example of this: Texas Hold'em is a series of bets over positional heuristics that are gradually becoming more clear throughout each hand. My personal favorite mechanic that skill-tests positional heuristics is the doubling cube in backgammon.

Before any backgammon roll, a player can offer to double the stakes. The opponent must either accept and play at double the stakes or resign immediately. The cube then belongs to the accepting player, who is the only one who can offer the next doubling. This allows players to profit from good positional heuristics. It also has the bonus of short-circuiting a boring round when both players know who has already won.

There are many other interesting parts of positional heuristics that I haven't even touched on—how they affect negotiation, kingmaking (or its inverse kingbreaking), and tempo in games. It's an easy part of games to overlook, but one well worth considering.