96 rule sets tested. One made the cut.
Hypothetical backtest after estimated costs, not a forecast. Never places orders.
≈ +$892 at $100 risk · after fees and funding
Simulated four-hour market · the study's own engine.
Window 1 is the first simulated trade that ends on its 10-candle exit, so every step shows. The next windows play every later trade in order, stops included.
| Candle close | Event | Prices | Alert |
|---|---|---|---|
| 25 Jan 20:00 | Close above the 20-candle high: long | entry 85.450 · stop 82.054 (2 × ATR) | BA Trend · SIM · ENTRY LONG 85.450 · stop 82.054 |
| 9 Feb 00:00 | Close below the 10-candle low: exit | 116.2 | BA Trend · SIM · EXIT +8.9R after estimated costs |
Picked on 2021–23. Confirmed on 2025. Run once on 2026.
Hypothetical backtest of 137 trades. R is the amount risked per trade. The result fell 60.7R before August recovered it.
| Period | Trades | Won | Net | Profit factor | Max drop | Markets + |
|---|---|---|---|---|---|---|
| 2021–23 · chosen | 554 | 30% | +314.2R | 1.96 | 59.8R | 6 / 6 |
| 2024 · left out | 197 | 34% | +118.7R | 2.13 | 21.6R | 6 / 6 |
| 2025 · confirmed | 203 | 29% | +21.6R | 1.18 | 28.2R | 4 / 6 |
| 2026 · final test | 137 | 27% | +36.1R | 1.41 | 60.7R | 5 / 6 |
The protocol and its programs were fingerprinted before the first run and are unchanged. Filling at the next candle's open instead of the close gives the same final test. The TradingView script's trade log matched the research engine on 38 of 38 BTC trades.
All 137 trades of the final test, worst to best. Most give back a little; a few long trends pay for all of them.
Every trade, sorted worst to best
One table. One message per 4-hour close.
Waiting for a 4-hour close that enters or exits
Relayed to Discord or Telegram. Never places orders.
Rerun the study. Then build on it with your AI.
$ node lab/bin/lab.mjs study dev 554 trades · net +314.154R · PF 1.962 · max drop 59.81R · 6/6 markets matches the evidence note val 203 trades · net +21.59R · PF 1.182 · max drop 28.168R · 4/6 markets matches the evidence note holdout 137 trades · net +36.055R · PF 1.412 · max drop 60.669R · 5/6 markets matches the evidence note REPRODUCED: every published figure matches.
The free desk on six intraday markets. A monitor: these intraday rules lost money in our validation.
Native TradingView capture · 0 mismatches Radar vs chart engine.
Ten things we make, one payment, and nothing that renews.
$9.70 a product. $411.98 stays with you.
Secure checkout by Stripe
Renews never/ Full source/ Yours forever/ 7-day refund before download
Each price on its own is that product's one-time price today: the Pine Complete Bundle at its current $46.00, the AI packs as the Starter Stack plus Chart Reader. The tools draw levels and conditions; you place the trades. Trading involves risk.
No. It is a hypothetical backtest of the rules on January–August 2026 data, after a 0.05% fee per side, one tick of slippage on stops and funding for every started 8 hours. R is the amount you risk per trade, so the result depends on how much you risk. It is not a live record and not a forecast.
Crypto fell and chopped for most of 2026, and a trend rule pays for failed breakouts with small losses. The August rally paid for them. Expect the same shape again; size small enough to stay in.
That is what passed. Long-and-short versions made the shortlist but lost money on 2025, so they never reached the final test. Other markets were not tested.
Yes, they are inputs, but only these values were tested. The inputs say so.
No. It shows the table, draws the trades and sends alerts. You place any order yourself.
TradingView serves CME data about ten minutes late without a CME subscription. The Radar reads what your account receives, and cards and alerts say how late it is.
Yes. The Research Lab downloads Binance's public archives and reruns the study; it prints REPRODUCED when every published figure matches. You need Node.js 18 or newer and a few minutes.
Yes. Open the folder in Claude Code, Codex, Cursor or Windsurf. AGENTS.md and the skill tell the agent what was tested, what must not change and how to check a change. You paste the result into TradingView, which is the only Pine compiler.
No. Pay once and keep the source.
It's included; download it from your account.
Same policy as the rest of the store (7 days, before download).
No. It is an independent build of a publicly explained concept, not the private Breakaway Bot or its signals.
A native TradingView capture, 24 Sep 2026, BINANCE:BTCUSDT.P 4h, default inputs. The September trades on the right came after the final test and are in no statistic.
Every result on this page is a hypothetical backtest after estimated costs, not a live record and not a forecast. The Trend Radar's rules passed a pre-registered test on January–August 2026 data; the Breakaway intraday rules lost money in the author's validation. Not the private Breakaway Bot or its signals. Not financial advice. Nothing in Premium places orders. Breakaway Desk Premium is charting software by ProEA Lab.