Chris Wilson Recalls Path of Exile’s Decoy Assets and the Cost of Tricking Dataminers
Wilson discusses removing assets, encryption and planted decoys, with the last approach creating extra work while making discoveries harder to interpret.
Path of Exile’s developers tried planting misleading assets for dataminers, according to co-creator Chris Wilson. His account describes a prank that created extra work for the team while making genuine discoveries harder to interpret.
Wilson discusses the experience in his video about what developers can do about datamining. GamesRadar’s coverage identifies three approaches in his discussion: removing unnecessary assets from public builds, using encryption before release, and leaving misleading material for people to find.
Three approaches to unreleased assets
| Method | What it is trying to change |
|---|---|
| Remove unused material | What the distributed build contains. |
| Encrypt material | When its contents become accessible, with limits once the game needs to use them. |
| Add decoys | How confidently people can interpret a discovery. |
Removing unused material reduces what a public build exposes. Decoys leave something available to find while changing what a discovery appears to mean. Wilson’s discussion of automated scrubbing presents a production-focused alternative to spending time constructing false clues.
Why a discovered asset can be misleading
An unused asset can be an abandoned real idea rather than a planted fake. Older files may show a direction the team explored and rejected. Neither explanation supplies a launch date, but both help explain why inspecting development material can produce a different picture from playing the finished update.
The reporting problem created by decoys
That is also the distinction behind our coverage of the Steam achievement exposure: accessible metadata can be real without resolving every question about its intended context.
The production trade-off favors spending less effort on elaborate false clues. A team creating decoys still needs to make the actual update, while players spend time sorting intentional misdirection from abandoned work. Wilson’s story exposes that mismatch between a successful joke and useful development work.
A prank still consumes production time
A misleading asset still needs to be created, named, included, and kept from interfering with the actual game. Even a successful joke uses some of the same production attention that could be spent on playable material. That is the practical tension in Wilson’s account of the experiment.
Removing unused material tackles a different part of the problem. It can reduce what a public build exposes without asking players to sort an intentional false clue from an abandoned real one. Encryption changes access, but material the game eventually needs to use presents another set of technical constraints. None of those methods makes a filename explain itself.
That approach leaves room for interesting discoveries without promising a future update. A cut idea can tell us something about a game’s development even when it never becomes playable. Its historical value and its value as a release prediction are different things, and a careful datamining story can explain both.
2 Comments