Rarity, Explained

The collection was produced with the HashLips Art Engine, a generative tool that stacks image layers to build each NFT. Every layer (the planet, the character, the weapon, and so on) holds a set of trait options, and each option is assigned a weight that controls how often it should appear. The engine then rolls 10,000 unique combinations against those weights.

Two consequences of that process matter for rarity. First, rarity is something the designers set on purpose through the weights — it isn't an accident of the art. Second, because 10,000 draws is a finite sample, the actual mint counts land close to the planned weights but rarely exactly on them. Both facts show up below.

The collection's own rarity reference lives in the official AllianceDAO NFT Rarity document. This page expands on it with the marketplace comparison and the data behind each claim.

There are 40 Objects and 40 grades, one-to-one. The grade reflects the Object's planned weight: the lower the planned count, the rarer it was meant to be, and the higher its grade. Phoenix Rising sits at the top — Grade 40 — because it was planned as the single rarest Object (a planned count of 12).

Generation variance then nudged the realized counts. Phoenix overshot to 25 mints; the Sindarin Fire Saber, planned at 24, undershot all the way to 6. So by raw count the Saber ended up scarcer than the Phoenix — but the designed grade keeps Phoenix as the apex, because that's what the weights intended. This is the single most important thing to understand about aDAO rarity.

A grade on its own can't say “I'm #3 of 10,000” — hundreds of tokens share a grade (466 share Grade 1). To give every token a usable ordinal, we lay the grades end to end from the apex down: the 25 Phoenixes occupy ranks #1–25, the 6 Sabers take #26–31, the 14 Swords of Zando #32–45, and so on to the common Objects at the bottom.

Within a single grade every token is equal rarity by design — they're the same Object. To assign distinct numbers we use the token's mint ID (lower = minted earlier), so the apex tokens get the lowest ordinals in mint order. Think of it as a tiebreaker, not a ranking: a Phoenix is “top 0.25%, mint #X” — the grade is the rarity, the within-grade number just records who showed up first. The meaningful signal is the grade and its band; the exact spot inside it doesn't carry further design weight, and the value of one Phoenix over another comes from its scene (planet, inhabitant, light, weather), not its rank within the grade.

We call this the Intended Rank. It's derived entirely from the collection's own Rarity grade, so anyone can reproduce and verify it from the public metadata.

Marketplaces like BackBone Labs (BBL) apply a generic, collection-agnostic rarity score: every trait counts equally, and a trait's rarity is just how few tokens share it. It's a reasonable default that works across thousands of collections, and for collections without a deliberate grade ladder it usually produces sensible results. For aDAO it just doesn't quite match the design intent, in three specific ways.

1. It inverts the apex. Because the score ranks by realized count, the 6–mint Sindarin Fire Saber lands at BBL #1, while Phoenix Rising — the designed apex — drifts down to roughly #66.

2. It scores the weather. Atmosphere counts the same as the weapon, so an ultra-rare sky can rocket a common weapon up the board. The clearest example is the six tokens BBL ranks #7–12: every one carries a common Object but the rare Weather = Lightning strike (6 mints), which BBL treats as exactly as rare as the rarest weapon.

3. Broken-NFT handling differs. BBL leaves most broken NFTs unranked — though a small number still slip through with a rank, so the behavior isn't fully consistent. The Intended ranking, by contrast, ranks every token the same way: the grade is set by the Object, so brokenness simply doesn't change it.

Honestly, both are valid — they answer different questions. The Intended Rank says “how rare was this meant to be”; BBL's rank says “how scarce is each trait in the final mint, atmosphere included.”

There's a practical point that matters here: BBL launched this collection and has been the primary marketplace, carrying nearly all of aDAO's trading volume. Its ranking is the one buyers and sellers have actually used to price NFTs since day one. Even though it diverges from the original design, it has effectively become the market's reference — so it can't simply be dismissed as “wrong.”

Rather than pick a winner, the NFT Explorer lets you choose. A BBL Rank / Intended Rank toggle switches every rank shown on the page, so you can browse by the market's standard or by the collection's design and compare them directly.

To keep the design standard consistent everywhere, the full Intended Rank is published as an open data file (below) that any marketplace — Atrium included — can read directly or host themselves. The goal is a single, verifiable source of truth for “as intended” rarity.

Only the Object is a graded rarity layer. The other four describe where the character is, who they are, and the scene around them — flat by design. But "flat by design" doesn't mean uniform in practice: generation variance and a few intentionally narrow trait weights produced some genuinely scarce values that the Intended grade doesn't capture. They aren't rarity in the design sense, but a collector can absolutely value them. The tables below are the public scoreboard so you can decide for yourself.

There's one more lens that doesn't appear in any single trait: how often the layers line up. Each Planet has a paired species (Sindari ↔ Sindarin, Cristall ↔ Cristallian, and so on), so an NFT can show an inhabitant on their home planet, sometimes holding their species's signature weapon. These aren't rarity by the design spec, but they're cohesive in a way the metadata makes legible only if you go looking.