Berry Avenue in September: Which IDs Were Real

September produced the biggest Berry Avenue list of the year: one tracker published 31 active IDs, another published 14, a third published 45 including a batch of alphabetic strings that turned out to be unrelated. Sorting that mess is the whole job.

The overlap is smaller than the headlines

Three sites, three counts, and only a handful of numbers appear in all of them. The 31-item list and the 14-item list shared nine entries; the 45-item list used word labels like HEADLESS and PURPLE alongside numbers, which is a different thing dressed up as the same thing.

What survives re-testing in October is eight. Not because the rest died in the last six weeks, but because most of them never worked — or worked once, for the person who published them, on a server running a different catalog.

The alphabetic strings were the trap

Lists carrying entries like PURPLE, FLUFFY and HEADLESS looked authoritative because they had names. Those are not catalog IDs and nothing in the avatar editor accepts them. Import ID takes digits only, which makes the entire category untestable by definition.

That is the single most useful filter for this game: if a “code” contains a letter, it is not an import ID, whatever the page says.

Why the numbers shrink over time

Every entry here is a catalog asset owned by a Roblox creator. Creators delete assets, make them private, or get their accounts terminated, and the ID stops resolving. Nothing on your side changed and nothing can be fixed.

That is also why September’s list looked healthier than October’s. Not decay, just a smaller and better-tested sample once the noise was removed.

The practical habit

Treat any Berry Avenue list as a starting set to test, not a menu to trust. Import the ones that look interesting, keep what resolves, and ignore the count at the top of the page — it is measuring the author’s copy-paste diligence, not the game’s catalog.