Reviews describe one game at a time. These surveys draw on all twenty-six, because the interesting patterns in modding are only visible across a file: which problems communities keep having to solve, and which architectural decisions make that work survivable.
Four hundred and eleven mods sorted by the problem they address rather than by download count. Interface repair is the largest single category and almost none of it is cosmetic: it is people rebuilding inventory screens and quest logs that shipped unusable.
Ninety-four updates across fourteen games with active modding scenes. Breakage rates range from three per cent to eighty-one, and the spread is explained almost entirely by one decision: whether the game loads content through a versioned interface or reads it directly.
Begun as a separate count of mods that add rather than repair, and closed after eleven games because the distinction proved impossible to draw cleanly. A texture pack that replaces illegible icons is both, and most of the sample was both. The usable part now sits inside the first survey.
Each of these pages is a count that changes. When a new game enters the ledger its mods are catalogued into the first survey and its update history into the second, and the conclusions on those pages are revised if the numbers move. Nothing is deleted when that happens — a superseded figure is struck through and the replacement noted beside it.
This is slower than publishing a fixed piece and arguing for it, and it is the only honest way to present a running tally. Two conclusions in the first survey have already reversed since it opened, both because a large game entered the file and shifted a category that had looked settled at fifteen games.
