Call the Bottom went live in the middle of a washout and started keeping score immediately. This page is the desk's research write-up — how the engine reads a bottom — and the full record of what happened next: the exact deployment moment, the prices then and now, every call graded live, and the complete Bottom Report trail. Everything below is computed from the board's own live state at the moment you load it. Paper game: no real funds, and the score is a transparent composite — a monitor, not advice.
Five instruments, one stance. The desk publishes a Bottom Report roughly every 30 minutes; a bottom is "forming" while the lows hold, and only "confirmed" when the engine flips — a Smart Play score through +10, or Jev's up-probability crossing its down-probability. A fresh low under a running bottom voids the read for that market.
For each market the desk tracks the lowest hourly-candle low on Coinbase since the season opened. That print is the running bottom; every price, call and score on the board is measured against it. If price prints a new low, the bottom moves, the market's V is broken, and the recovery score resets with it.
A 0–100 composite grading the shape of the recovery off the running bottom — not a prediction that the bottom holds. Live values for all three markets are in the scorecard below; the full formula is printed under it, weights and caps included.
The perps desk's per-asset score and bias (SHORT · LEAN SHORT · NEUTRAL · LEAN LONG · LONG), read live from the Clawd perps desk. Deeply negative scores into a holding low are the capitulation signature the desk waits for; the flip back through +10 is one of its two confirmation triggers.
For BTC, the Jev judgment model prices the next 24 hours as competing paths — up vs down, V-recovery, continued down, chop/base — plus a V-setup score. Jev's up-probability crossing its down-probability is the desk's second confirmation trigger. The board labels the feed "live Jev" only when the source itself reports live.
The board carries a prediction-market snapshot (Polymarket via the Jupiter Prediction API, cross-checked on Polymarket Gamma). The Jev desk posts a paper prediction call only when its probability disagrees with the crowd by 12 points or more — entry stamped server-side from the snapshot, never named by the caller — and every call is graded in public, win or lose.
One call per wallet per market. A call spends 10,000 paper USDC at the live Coinbase spot price at that moment, stamped by the server. Calls are ranked by closeness of entry to the running bottom. Paper Jev — Clawd's paper decision agent — plays by the same rules on the same board. No real funds move, ever.
Honest labels: the V Score is a composite monitor built from the four published components — it grades recovery shape, it does not guarantee the bottom holds. Smart Play and Jev are model outputs, shown with their live values so you can judge them against the tape yourself. Bottom timestamps are the hour of the lowest hourly candle since deployment.
Loading the record…
| Market | Price at deploy* | Running bottom | Price now | Above bottom | Deploy → bottom | Since deploy | V Score now |
|---|---|---|---|---|---|---|---|
| Loading… | |||||||
Every bottom call on the board, graded against the live price right now. "% above bottom at call" uses the running bottom as it stood in the hour the call was placed.
| Caller | Market | Called (ET) | Entry | Bottom at call | Above bottom at call | Price now | Vs entry | Paper P&L |
|---|---|---|---|---|---|---|---|---|
| Loading… | ||||||||
The Jev desk's paper calls on the prediction-market snapshot. None of these markets has resolved — every one closes in December 2026 — so each is graded mark-to-market against the crowd's current price, not as a win or a loss.
| Market | Side | Called (ET) | Entry (crowd) | Jev vs crowd at entry | Crowd YES now | Side price now | Mark | Status |
|---|---|---|---|---|---|---|---|---|
| Loading… | ||||||||
Loading…