NFTs don't have a "price" the way XRP does. There's no order book matching identical units. Each token is, by definition, non-fungible — its price is whatever the next willing buyer pays.
This makes pricing one of the most-manipulated parts of the NFT market. Understanding the manipulations is half the defense against them.
Three numbers to read on every collection
1. Floor price
The cheapest currently-listed NFT in a collection. People treat this as "the price of the collection." It isn't — it's the most desperate seller's price.
What it actually tells you: how much it costs to acquire the worst-currently-available member of the collection. Useful as a lower bound for any individual NFT's market value. Useless as a single-number summary.
Common manipulation: the project team buys their own NFTs at successively higher prices to lift the floor. New buyers see "floor going up!" and FOMO in. Trade volume is mostly the project trading with itself.
The Gopnik marketplace labels collections where the project's known wallet has > 30% of recent volume. Treat the floor accordingly.
2. Volume (24h, 7d)
How much trading actually happened. Real volume from many distinct wallets is a sign of organic interest. Concentrated volume (3 wallets doing all the trades) is a sign of wash trading.
What to look for:
- 24h volume in line with collection size (e.g., a 10,000-NFT collection should see at least dozens of trades per week if it's "active").
- The distribution of buyers and sellers — many unique addresses, not the same ten.
- Volume that survives the project team's not trading. If the team pauses and volume craters, you've found the wash.
3. Holder distribution
How many unique wallets hold the collection, and what's the concentration. A healthy 10,000-NFT collection has maybe 4,000–6,000 unique holders. A "10,000-NFT" collection where 200 wallets hold 8,000 of the NFTs is either an art project (legitimate, but tiny) or a manipulation (not legitimate).
The Gopnik marketplace shows holder distribution as a Gini coefficient. Above 0.6 is concentrated. Above 0.8 is suspicious.
The rarity trap
Most collections assign "rarity scores" to their NFTs based on trait combinations. NFT #4271 has "Glowing Eyes" (3% of the collection) and "Birch Background" (1.5%) and "Golden Crown" (0.3%) — its rarity rank is 14 of 10,000. People pay premiums for low-numbered ranks.
The trap is in which traits get scored.
Rarity is whatever the calculator decides it is. Different rarity sites use different formulas: weighted product of trait probabilities, statistical scarcity, hand-curated tier lists. The same NFT can be #14 on one site and #2,830 on another.
Some traits are economically meaningful; most aren't. "Has a hat" might be in 3% of the collection. So might "background is slightly more blue than average." The first might command a 5x premium because the community decided hats are cool. The second commands zero because nobody cares. A naive rarity score weights them equally.
The "right" rarity is what the market currently prizes. Look at recent sales of high-rank NFTs in the collection. If the top-100 ranks consistently sell for 5x the floor, the market is endorsing the scoring. If they sell at floor + 10% randomly, the market doesn't care about the scoring.
How to value an NFT before buying
A pragmatic four-question checklist:
What's the floor price, and is it manipulated? Check holder distribution and trade-concentration.
What's the recent (last 30 days) median sale of similar-rarity NFTs? Median, not average — averages get blown by single outlier sales.
Is the off-chain metadata going to outlive me? Check the IPFS CID's pin status. If the project pinned for 30 days and disappeared, the image dies when current pins lapse.
What's the issuer's track record? Has this account minted before? Did those collections survive? Do they have any soulbound credentials proving identity?
If you can't answer three of those four, you're guessing. Guessing is fine for €20. Guessing is expensive at €2,000.
The Gopnik marketplace shows you, in one row
Every NFT detail page in our marketplace shows:
- Floor, with manipulation flag if applicable.
- 30-day median sale for similar-rarity peers.
- Off-chain metadata pin status (3-of-3 / 2-of-3 / 1-of-3 / unpinned).
- Issuer's account age and any previously-minted collections.
- Rarity rank (from a published, open-source scorer — the same formula for every collection).
It does not show:
- A "fair value" estimate. We don't believe in that for non-fungible assets.
- "Investment advice" of any kind. NFTs are not financial products in most jurisdictions and we don't model them as such.
What the course unlocks
Passing NFT 101 unlocks minting 5/day, buying up to €500/tx, and is the prereq for listing on the marketplace. The exam is 25 questions, 70% to pass.
NFT 201 covers auctions, royalty configuration, large-collection minting workflows, and the brokered-vs-unbrokered sale economics (where most royalty-avoidance happens).
Next: rugs, fakes, and what royalties actually do.