# Vector Desk scope

Updated 21 September 2026. Public research and paper practice.

Our mission is to investigate whether defined vector events differ from comparable ordinary candles, and separately whether fixed trading rules profit after modeled costs, using automatic collection and visible results.

## This phase

- Preserve the frozen volume-filter strategy comparisons for 20 September–19 November 2026, ending at 00:00 UTC on 19 November. UBTC is primary; UETH and HYPE are replication markets. Both sweep arms keep vector targets, so this comparison does not isolate the value of those targets.
- Preserve the separate descriptive green-down/red-up recovery study beginning 21 September. Each event needs 24 hours of follow-up under its declared rules; pending observations are not failures.
- Show the saved July–September historical recovery comparison separately from forward evidence. Historical results are mixed; they do not establish a dependable recovery advantage. Recovery is not a trading return.
- Keep automatic shared research, the existing private owner recorder and browser-local paper notebooks working. Visitors do not have personal cloud recording or accounts. Notebook, recorder and replay feature expansion is paused while we verify reliability and comprehension.
- Make data age, missing observations, costs, provisional results and limitations visible. Correct presentation without retuning the frozen studies or rewriting old reports.

## What the activity map measures

The map shows portions of historical vector-candle price ranges that later prices have not revisited. “Vector” includes elevated and climax activity. Amounts estimate the whole original candle's traded value from volume × typical price. They do not measure value in the remaining fragment, outstanding orders, liquidation value or remaining inventory. Source-candle time is distinct from the report's observation time.

This candle dataset does not identify which addresses caused the activity, their motives or their off-venue exposure. UBTC and UETH are Hyperliquid spot representations, not a measurement of global BTC or ETH inventory.

## Deferred: participant Study B

Address/agent classification, wallet clustering, perpetual funding or liquidation analysis, new entry rules and live execution are outside this phase. No participant study is running behind owner sign-in.

Study B would investigate whether fill-derived participant behaviour adds useful information beyond the candle baseline. It is deferred until a scope review on **19 November 2026**, not promised for launch then. Before starting, it needs an agreed time/resource budget, measured data coverage, defensible labels based only on information available before the event, a sample and uncertainty plan, and a separately frozen future evaluation window.

The existing freeze does not prevent an independent new study. Deferral is a capacity and research-design decision, not a claim that participant research is impossible. Address labels alone would not establish identity, inventory or motive.

## Next decisions

1. Verify unattended collection, evidence recovery and version-attributed operational observation. A manual catch-up or passing local tests does not complete operational acceptance.
2. Run 3–5 independent beta sessions. Check that visitors can explain the existing findings, data age, simulated outcomes and notebook limits. Fix observed failures before expanding features.
3. Review the frozen research at or after 19 November with its required data and follow-up. The collection cutoff remains 21 November. Too little evidence is an allowed result; the 30-trade operational gate is not a statistical power calculation.
4. Decide whether to continue, revise or retire each question. Accounts, subscriptions and investment offers require separate customer evidence and a service decision. Live trading is a separate milestone.

[Back to the desk](./) · [Quick start](quick-start.md) · [Frozen trade protocol](protocol.json) · [Historical recovery data](research-update-v1.1.json)