Teardown 19 inventory systems measured Worst action cost recorded: eleven Two games abandoned the grid
The Interface Files mark The Interface Files
Teardown

The Inventory Grid Is A Storage Format Pretending To Be A Design Choice

Nineteen games, one screen, the same four measurements each time. The grid of identical squares is not a considered answer to the problem of carrying things; it is the cheapest possible answer, and it has been the default for long enough that most of the genre has stopped noticing it is an answer at all.

Teardown 19 games Updated monthly 17 min read
A floating three-dimensional grid of empty square slots rendered as glowing orange wireframe outlines with a few abstract geometric solids inside
Fifteen of nineteen games represent everything you own as identical squares of equal importance.

An inventory grid solves a specific engineering problem elegantly. It gives every object a uniform footprint, it makes capacity trivially easy to calculate and balance, it renders in one pass, and it can be extended by adding rows. These are real virtues, and they are all virtues for the people building the game rather than for the person playing it.

What the grid does not do is tell you anything. Fifteen of the nineteen games we measured present a lantern, a letter, a sword and four hundred grams of salt as four visually equivalent squares, and then require you to hover each one to learn which is which.

The comparison problem

Eleven of nineteen games offer no way to see two items at once. This is the single most consistent finding in the study, and it matters because comparison is the only reason to open an inventory at all. You are never browsing for pleasure; you are deciding what to keep.

The median action cost for a comparison across all nineteen was four. The best was two, in a game that abandoned the grid for a sortable list with columns. The worst was eleven, in a game that requires unequipping an item before its statistics become visible, which means you cannot compare without first destroying the state you were comparing against.

MeasureBestMedianWorst
Actions to compare two items2411
Actions to move a stack125
Sort options offered620
Slots visible without scrolling904018

Sorting, or the absence of it

Two sort options is the median. Three games offer none at all, which in a system with over a hundred slots amounts to asking the player to maintain the order manually, and several players do exactly that, which is a hobby the game did not advertise.

The game with six sort options is also the game with the lowest comparison cost, and that is not a coincidence. Once a designer starts thinking of an inventory as a set of records rather than a picture of a bag, the useful features follow more or less automatically.

A bag is a metaphor. A table is a tool. Most games commit to the metaphor and then spend years apologising for it with tooltips.

Encumbrance as a number

Six of nineteen games do not show carrying capacity as a figure. They show a bar, or a colour change, or nothing until you are over the limit and moving slowly. In four of those six, encumbrance affects movement speed on a continuous curve, which means the game is running a calculation on a quantity it will not display.

This is the same failure we recorded in Quartermaster with spoilage and road wear, and it is worth naming precisely: it is not difficulty, and it is not mystery. It is a system operating on information the player is not given.

The two that left the grid

Two games in the nineteen dropped the uniform grid entirely. One replaced it with a sortable table of rows with columns for weight, value and condition. The other replaced it with a small number of named containers, each with a strict item limit and a purpose, so that a decision about what to carry became a decision about which bag.

Both scored above the study average on every measure we took. Neither is a large or expensive production, which suggests the grid persists out of habit rather than necessity.

FormatGamesMedian comparison cost
Uniform square grid155
Weight-based list23
Sortable table12
Named containers12

What we are not arguing

Not that grids are worthless. A grid with a comparison view, real sort options and a visible encumbrance figure is a perfectly good interface, and three of the fifteen grid games here qualify. The argument is narrower: the grid is chosen by default rather than selected on merit, and the features that make it usable are the ones most often left out.

Nor are we arguing that inventories should be minimal. A dense inventory that rewards study is a legitimate design. Density is not the same as opacity, and the games in this study that got it right were frequently the busiest ones.

Where the data leaves it

Comparison is the task an inventory exists to serve, and eleven of nineteen games make it impossible on one screen. The grid is not the cause of that on its own, but it is the format under which the omission goes unnoticed, and the four games that left the grid behind were better on every measurement we took.