The document did not get amended. It got deleted.
On-chain, Genius Terminal ran a points program for Season 2. The official documentation listed a window — April 10 to August 10 — and a rate: 1,500,000 points distributed daily. That is the number users traded against. That is the number that priced their hours.
Then the page vanished. The daily rate collapsed to 50,000. A 96.7% reduction, executed without a governance vote, without a timelock, without a single line of on-chain coordination.
And the reward itself? Community records cite a promise of 70 million GENIUS unlockable by October 25, 2025. CoinMarketCap now shows the date as February 2028. Two years and four months of added latency, on a promise the project no longer describes as a promise.
Three operations. One weekend. Zero votes.
Co-founder Ryan Myher responded the following day. Calm. Composed. Prepared.
I have seen this movie. I have audited the contracts behind it.
Context
Genius Terminal positions itself as an on-chain trading terminal — a DEX aggregation interface with YZi Labs backing. The category is crowded: Axiom, Photon, BullX, and a rotating cast of forks. The technical moat is thin. These products compete on routing, speed, and fee structure, not cryptography.
Which means the real product is not the terminal. The real product is the incentive.
Points programs — PointFi — replaced yield farming as the standard user-acquisition vector after 2023. Users trade, they accumulate points, points convert to tokens, tokens convert to exit liquidity. The loop works only if the conversion is credible. Remove the credibility and the loop eats itself.
Genius built on that loop. Season 2 was the load-bearing wall.
There is a hole in the disclosure. No total supply. No team allocation. No investor vesting schedule. When those numbers are absent, the 70 million GENIUS figure floats without a denominator. Seventy million out of what? Two billion? Twenty billion? The dilution math is unknowable, and unknowable math always favors the party holding the spreadsheet.
From an audit standpoint, a terminal that can alter its own incentive docs is not a protocol. It is a database with a marketing department. The contract is not the risk here. The governance key is.
Core
I spent six weeks in 2017 tracing replay attacks across the Ethereum Classic fork boundary. Fifteen million transactions. Local node farm in Nairobi. The lesson I carried forward was simple: a system's true rules live in what it can enforce, not in what it publishes. Documentation is a hope. On-chain constraints are a fact.
Apply that lens to Genius. Three questions decide everything.
First: is there a timelock on the reward distribution? If the team can push an unlock from 2025 to 2028 by editing a frontend, the answer is no. There is no contract forcing the date. There is no Merkle root committing the 70 million tokens. There is a spreadsheet, and a person with access to it.
Second: is the points rate governed on-chain? The collapse from 1.5 million to 50,000 daily is a 30:1 cut. That is not a tuning. That is a repricing of every hour a user spent on the platform. An on-chain rate would require a transaction, a visible signer, a timestamp. A documentation change requires none of that. The absence of a visible transaction is itself the evidence.
Third: what does "points" legally mean? Myher's framing is precise: points are "additional perks," not promised returns. Users "agreed to the terms." The team's personal tokens "will not unlock before user tokens."
Read that last line slowly. If user unlocks move to 2028, team unlocks move to 2028. The user gains nothing. The team concedes nothing. It is alignment redefined as mutual delay — a rhetorical gift with no economic transfer.
This is the same playbook I mapped in Compound's v1 governance in 2020, where a 24-hour delay looked like safety and functioned as an attack surface. Delay is not protection. Delay is a timer someone else controls.
A deleted page is a gas leak with better formatting. Every gas leak is a story of human greed.
There is a harder edge to this. Some participants are demanding refunds for Season 2, which implies the season carried a cost — fees, slippage, or a paywall. If real money crossed the table in exchange for a points balance later devalued by 96.7% and locked until 2028, the dispute stops being about airdrop disappointment. It becomes a question of consideration paid and consideration withdrawn.
The 96.7% cut and the document deletion are the actual disclosure. Everything after is public relations.
I will not call Genius a fraud. Fraud implies intent I cannot prove from a documentation diff. But I will call it what the evidence supports: a centralized operator that discovered its incentive promises carried no enforcement cost, and acted accordingly.
I do not fix bugs. I reveal the truth you hid.
Contrarian
Give the bulls their due, because the cold read cuts both ways.
PointFi is fatigued industry-wide. Projects across the terminal and DEX landscape have trimmed, delayed, or quietly shelved airdrop seasons as token valuations compressed in the bear market. Genius is not an outlier. It is a symptom. If you are angry at one team, you are misreading the base rate — dozens of teams made the same calculation and most simply did it more quietly.
And the founder's alignment claim, hollow as it is, is not nothing. A team that could have unlocked its own allocation and chose to say otherwise has at least acknowledged the optics. That is a data point. It is a weak one, unverifiable without watching the treasury wallet, but it exists.
There is also a defensible reading of the cut. If points were being farmed by wash traders — and a volume-linked points rate guarantees wash trading — then 1.5 million daily was never real demand. It was noise paid for by future dilution. Collapsing the rate may be less a betrayal than a correction of a number that was always fiction.
The bulls are right that this happens everywhere. They are wrong that it makes it acceptable.
Takeaway
Watch the treasury address. Not the tweets.
If GENIUS ever flows from a team wallet to an exchange before February 2028, the "alignment" line dies on-chain, permanently and verifiably. That is the only promise worth auditing, because it is the only one that leaves a trace.
Everything else — the perks language, the clickwrap defense, the deleted season — is a story. Hype burns hot; logic survives the cold burn.
