Evidence
How Nansen drives every trial.
Nansen supplies the candidate universe and the point-in-time market data used to construct every blind trial. Trench Trials deterministically selects eligible assets from that data, derives the blind clues, and uses Nansen price candles to calculate the outcome. Without Nansen, no new Replay, Daily or Live round can be produced.
Historical inputs are frozen and hash-sealed before a player enters. This prevents the evidence from changing after a decision has been made and makes every verdict reproducible. Existing rounds can survive a temporary Nansen outage, but the product cannot create new content without Nansen. There is no substitute data source.
The pipeline
Nansen API → eligible candidates → blind clues → locked decisions → Nansen outcome candles → verdict
01
Two Historical Token Screener requests — a 7-day and a 1-day window — read as of the day before the cutoff, so nothing in them comes from the outcome window.
02
Only freely traded tokens that Nansen positively classifies qualify (memecoins, AI, DEX, gaming, infrastructure, derivatives). Stablecoins, wrapped or bridged majors, staking, LP, yield, tokenized-stock and RWA tokens are excluded, and anything unclassified fails closed. Floors: $2M liquidity, $100K 24h volume, 30 days of age. The three most liquid qualifying tokens are chosen, and no two playable rounds share two tokens.
03
Four signals per token, computed from the screener rows below and bucketed. Names, addresses, dates and prices stay hidden.
04
Before anyone plays, the candidates, signals, receipts and outcome rules are written into a manifest and sealed with a SHA-256 commitment. The inputs are frozen from then on.
05
Your blind pick and your final keep-or-switch are locked on the server’s clock, in that order, before any outcome is released.
06
One Historical Token OHLCV request per token supplies the hourly candles at the round’s fixed entry and exit boundaries. The whole outcome window must be present; a missing or broken candle rejects the round.
07
Each return is exit close ÷ entry close − 1. The highest return wins; returns within 0.01 percentage points share the win. The manifest is revealed so the commitment can be recomputed.
08
Your blind pick’s return minus your final pick’s return: what the reveal cost or gained you. Keeping your pick costs zero.
Endpoints and what they drive
| Nansen endpoint | Drives | Credits / call |
|---|
Historical Token Screener/api/v1beta1/token-screener/historical | Which tokens qualify, and the four blind signals | 5 |
Historical Token OHLCV (1h)/api/v1beta1/tgm/historical-token-ohlcv | Exact entry and exit prices, returns and the winner | 5 |
Token Screener (current)/api/v1/token-screener | Live candidate capture | 1 |
Historical Token OHLCV (5m)/api/v1beta1/tgm/historical-token-ohlcv | Live resolution | 5 |
| Clue | Formula | Nansen fields |
|---|
| Buy / sell balance | (buys − sells) ÷ (buys + sells), 1-day window | buy_volume, sell_volume |
| Trading pace | 1-day volume ÷ (7-day volume ÷ 7) | volume (1-day and 7-day windows) |
| Netflow vs liquidity | 1-day netflow ÷ liquidity | netflow, liquidity |
| 7-day momentum | 7-day price change | price_change (7-day window) |
What verification means
- Every input traces to a logged Nansen response, by request ID and SHA-256, sealed in the round’s commitment.
- Each return is exit close ÷ entry close − 1 on the committed candles, and anyone can recompute it.
- The round was reviewed and approved under the current eligibility policy.
- Only approved rounds are offered for play. A withdrawn round leaves play and scoring; nothing is deleted.
No fallback, ever
Missing or malformed Nansen data invalidates or prevents a round. No alternate market-data provider, estimate or mock input is ever used in production.