Most wiki-style sites answer every question the moment someone asks it, whether or not the answer is actually known yet. "Best gear for beginners." "Top 5 setups." Confidence is the default output, and nobody checks whether it was earned.
I ran into this building a small fan wiki for a Roblox game called Dig & Clean, where players buy shovels, detectors, and pressure washers with in-game Gold. The obvious content to write is a buying guide: what should you purchase first, second, third. I don't actually know that yet, because I haven't played through enough runs to say which upgrade path is faster, and every other wiki for this kind of game just writes the guide anyway, presumably by guessing or copying whoever guessed first.
What I do know, reliably, is the price list. I pull it from the game's own API and can say, for a given Gold total, exactly which items are affordable. So the beginner page states only that: at 4,000 Gold, at 100,000, at 1,000,000, at 9,000,000, here's the priciest item you can afford in each category, sourced from a timestamped API snapshot. Then it says outright, "these are prices, not recommendations," and that the opinionated buying route is coming after actual playtesting, not before.
The part I think is worth more than the page itself is a small counter sitting next to the gear count: "48 gear listed, 0 verified by us." Everything scraped from the API counts as a fact. Nothing gets counted as verified until someone on our side has confirmed it against actual gameplay, not just against the API response. Right now that number is zero, and the site says so, in the same place it brags about having 48 items.
It would be easy to skip that counter. Nobody's forcing a small wiki to admit how little of its own content has been checked. But a "sourced" badge that's always full and never wrong isn't measuring anything, it's decoration. The counter only means something because it's allowed to sit at zero.
If you build anything that scrapes a live source and then layers claims on top of it, from game wikis to changelogs to internal dashboards, the same split is probably hiding in your data: numbers you pulled versus numbers you actually checked. Most tooling doesn't draw that line out loud. This one's at digandclean.org/guides/beginner if you want to see what a wiki looks like when it says "we don't know that yet" instead of making something up.
Top comments (0)