Test your Drift settings with a locked rule.
Practise a Drift Desk setting, write your pass rule, lock it, then take one final on coins and months practice never used.
Planned for members, included in Academy Premium when it opens; being built now, no date yet.
Our own Drift exam.
Our Drift research went through these steps and failed its rule; here it is as recorded.
1 · Choose
Design preview · tool being builtOur setting
Drift Desk 1.3.1 · 4h · long only · exit at the first opposite trend flip · REAL costs
Outside the native TradingView check
Engine-to-TradingView parity is established for Drift Free 1.3.1. Drift Premium has no Pine build yet, so none of its figures, including the TV-cost ones, has a TradingView Strategy Tester check.
Its rules, word for word (6)
- Signals: Drift Free 1.3.1 with its default inputs (fresh Drift flips; every optional gate off).
- Long entries only: a short flip is skipped, never queued.
- Initial stop: Drift Free's 10-bar structure stop plus 0.25 ATR(14), sized to 100 USDT of planned risk (1R), and kept for the life of the position.
- No profit target: the position closes at the close of the first opposite (bearish) Drift flip.
- Timeout raised from Drift Free's 120 bars to 5000 bars; no FINAL position reached it.
- Timeframe: 4h.
The Exam’s flip exit lifts Drift Free’s 120-bar timeout to 5,000 bars, as our rule 5 did.
In the Exam: you choose these four settings; long only and the flip exit stay engine extensions.
2 · Practice
Design preview · tool being builtOn the 10 practice coins, entries 2020-01-01 → 2024-01-01, we screened 36 distinct configurations; 4 passed our screen and 2 were frozen as candidates. Ours was frozen on 2026-09-23 23:31 UTC.
BTC · ETH · BNB · XRP · ADA · LINK · LTC · TRX · ATOM · XLM
Ours passed that screen only under a reading of one screening rule that we chose after the results were known (A6.1).
Our extra step, validation. Our 2 candidates ran on Set A, entries 2024-01-01 → 2025-04-01; 1 qualified. Results recorded 2026-09-23 23:51 UTC. No validation numbers are shown here.
The size of our search, and the reading we chose
Selection bias: 42 configuration × stage pairs (36 distinct configurations) were evaluated in this lab, and configs-ledger.csv lists every one. Before it, earlier ProEA Lab research tried 31 recipes on BTC and ETH and measured three 4h candidates; that work shaped this lab's hypotheses and grid.
The DEV neighbour rule drops a survivor whose immediate numeric neighbours all fail the screen. We read it so that a rule with no numeric parameter (long-only) has no neighbour, and the Free base is the control rather than a neighbour. This reading was applied after DEV results were known and ratified after VAL.
In the Exam: practice is unlimited on the same coins and years, and the validation months sit inside the final’s part 1, so you cannot see them before you lock. Your receipt records how many practice runs came first.
3 · Pass rule
Design preview · tool being built- In both Set A (holdout months) and Set B (untouched coins): Premium net R > 0.
- In both sets: Premium net R > Drift Free net R.
- In both sets: Premium profit factor >= 1.2.
- In both sets: Premium profitable on at least 6 symbols.
- Premium net R under STRESS costs (both sets) > 0.
- In Set B: Premium net R > Drift Free taking only its long signals.
First written 2026-09-24 00:24 UTC and committed 2026-09-24 00:28 UTC: after our validation results, before the final ran.
Amendments written after validation: A5, A6.
The lab’s account of the timing
The promotion gate was written and committed after VAL had run and after the VAL choice was committed, and before FINAL started. The founder chose its conditions without being shown the VAL results (A5). The Set B condition against Drift Free long-only was added after the founder saw the VAL robustness audit, in the stricter direction only (A6.2).
In the Exam: the default rule is this one, read onto the final’s two parts; you can change it, and it is locked before the final.
4 · Lock
Design preview · tool being built- Setting hashdaadfadfd47d…
- Frozen2026-09-23 23:31 UTC
- Rule file sha256287dc7b1ae95…
- Rule committed2026-09-24 00:28 UTC · 3544b22
- Protocol sha25658175c1c87f8…
In the Exam: locking records one fingerprint of your settings and rule, with the time, and closes them for that exam.
5 · Final
Design preview · tool being builtFailed our rule
Set A: Premium profit factor 1.16 is below 1.2.
Set B: Premium profit factor 1.12 is below 1.2.
Final ran 2026-09-24 01:31 UTC; our rule: the 6 conditions in step 3.
Set B: the 10 coins practice never used (DOGE, QTUM, ETC, XTZ, VET, ALGO, ZEC, DASH, HBAR, NEO), entries 2020-01-01 → 2026-09-23. The Exam’s part 2 uses the same coins and window.
| Set B | Our setting | Drift Free, long only | Drift Free (default settings) | Always long, same capital (all costs) |
|---|---|---|---|---|
| Net R | +197.60R | +170.73R | −152.32R | +299.09R |
| Profit factor | 1.12 | 1.12 | 0.91 | n/a |
| Positions | 4,685 | 2,430 | 2,607 | held at every 4h close |
| Coins with net R above zero | 6 of 10 | 10 of 10 | 1 of 10 | 10 of 10 |
| Largest fall, marked at every 4h close | 177.77R | 133.82R | 246.64R | 93.54R |
| Net R under STRESS costs | +85.38R | +119.92R | −205.83R | not in the export |
Always long keeps 6,371 USDT, our setting’s average capital, in these coins, rebalanced at every 4h close, and pays every fill: +299.09R (all costs). Charging only the first and last fill it would be +308.80R (rebalancing costs excluded).
Luck test
We reshuffled Set B’s 352 weeks 20,000 times, keeping whole weeks together, to see how much of each result could be luck.
Our setting alone: The range includes zero: not distinguishable from luck at 95%.
total +197.60R · middle 95% −107.39R to +515.70R · at or below zero in 10.8% of reshuffles
Our setting minus Drift Free, long only: The range includes zero: not distinguishable from luck at 95%.
total +26.86R · middle 95% −227.67R to +301.05R · at or below zero in 43.9% of reshuffles
Our setting minus Drift Free: The range includes zero: not distinguishable from luck at 95%.
total +349.92R · middle 95% −21.20R to +722.52R · at or below zero in 3.2% of reshuffles
3.2% of reshuffles were at or below zero; zero falls outside the middle 95% only when that share is below 2.5%.
Our setting minus always long (all costs): The range includes zero: not distinguishable from luck at 95%.
total −101.49R · middle 95% −367.81R to +179.39R · at or below zero in 77.1% of reshuffles
How the luck test works
Premium beside its comparators on each set (REAL costs): Drift Free long-only first, then Drift Free, constant long exposure of Premium's average capital, and buy D/P0 units and hold. Drawdowns are mark-to-market at 4h closes. Margins carry the week-block bootstrap of the weekly differences; resolved95 means the 2.5%–97.5% interval excludes zero.
How always long is built
Capital-matched constant long exposure (added after FINAL, postFinalCorrections PFC-2). Per symbol and edition, D = the edition's average capital deployed on the symbol over the window (Σ quantity × entry price × time held / window length). Hold D of the coin at every 4h close: buy D at the first window close, trade back to D after each close, sell at the last. P&L = Σ D × (c_k / c_(k−1) − 1). REAL costs, 0.07% of each fill's notional (0.05% commission + 0.02% slippage), on every fill that P&L needs: the entry, the rebalancing after each close but the last, and the exit (costR = fillCostR + rebalanceCostR). entryExitCostsOnlyNetR charges only the entry and exit fills; it books the rebalanced P&L without the trades that make it, so it overstates the benchmark. maxDrawdownRMarkToMarket: the largest fall of the mark-to-market curve (every cost booked when paid), summed over the set's symbols, from 0.
Our Drift test publishes these coins’ results for 9 of the Exam’s 48 settings (three Drift editions, three cost models); for those, the part 2 result is already public here, so a final on one of them is not blind.
In the Exam: the final runs once per setting, whatever the rule, on coins and months practice never used, and the verdict is against your own rule.
6 · Receipt
Design preview · tool being builtOur receipt is the export itself: MANIFEST.json lists the sha256 and size of each of the export’s 10 other files. The key facts, as JSON:
{
"evidenceVersion": "2026-09-24-final-r6",
"setting": {
"label": "Drift Premium (research edition)",
"candidateId": "C1",
"configHash": "daadfadfd47d9685d66bdd6e745ed85ac41b98c46f5b575efa82e5f9b294604f",
"frozenAt": "2026-09-23T23:31:35.249Z"
},
"rule": {
"sha256": "287dc7b1ae95f5c0e99783e0d2f251e735388e33047bde1568025e5984003fa7",
"committedAt": "2026-09-24T00:28:48Z"
},
"finalRunAt": "2026-09-24T01:31:17.448Z",
"verdict": {
"pass": false,
"reasons": [
"Set A: Premium profit factor 1.16 is below 1.2.",
"Set B: Premium profit factor 1.12 is below 1.2."
]
},
"manifestSha256": "f0df7913c6337f95a4ae21339cfce9c3baa3894a87b147fb1d9350e4d664fe32"
}One final run, 2026-09-24 01:31 UTC; this export revises its analysis files, and its revision note says wording only: earlier versions described as never published and listed by their MANIFEST.json sha256 only, and "set" for when the gate and amendments A5 and A6 were written; no number, verdict or gate reason changed
.
In the Exam: your receipt is a JSON file you download and keep: your settings, rule, lock time and results.
What our order got wrong.
We wrote our pass rule after we had seen validation results. The Exam asks you to set yours and lock it before the final runs.
Planned for members.
Nothing is for sale until it opens; here is the plan.
- The Strategy Exam in your browser, for Drift Desk 1.3.1 settings.
- Unlimited practice; one final per setting, whatever the pass rule.
- Our Drift test’s data: 20 Binance spot coins, 2020–2026.
- Every result beside three baselines (Drift Free; Drift Free, long only; always long with the same average capital) and a luck test.
- A receipt for each final: settings, rule, lock time and results.
Included in Academy Premium ($49/month) when it opens; being built now, no date yet.
If you cancel, the Exam closes for you; the receipts you downloaded stay yours.
Straight answers.
Can I test my own Pine script?
No. The Exam tests Drift Desk settings only; bringing your own Pine script is not planned.
Does a pass mean it will make money?
No. A pass means your locked rule held on a hypothetical backtest; it is not a prediction or a live record.
Is the engine the same as TradingView?
For Drift Desk 1.3.1’s default settings under TradingView’s costs, the engine reproduced TradingView’s Strategy Tester exactly on 914 of 914 native trades. Long only, the flip exit and other target sizes are engine extensions: TradingView exactness is not claimed for them. 2 of the Exam’s 48 settings are inside that check.
Can I game it?
The exam is a discipline aid, not a security boundary: it enforces the order of the steps in this browser and records it; it cannot stop anyone from gaming their own test. Our own results are public too: our Drift test publishes the part 2 coins and months for 9 of the Exam’s 48 settings, so a final on one of those is not blind for anyone who has read it.
Hypothetical backtests, not a live record. We, ProEA Lab, are building the Strategy Exam with AI assistance. Trading involves risk.