Methodology · every assumption, published

The full method, open for inspection.

Most people show you a chart. This is the page where I show you how it was built, what it assumes, and where it's weakest — so you can decide whether to trust it instead of taking my word.

Read this before you read the numbers

I built the V7 formula, then ran it against Bitcoin's price history. It's a backtest, and it's retroactive — the formula didn't exist for most of the period it's measured against, so nobody — me included — traded these signals as they printed, and none of it is a prediction or a promise about what you'd make. One thing to keep separate: on /about I describe selling into the 2025 top myself. That is a personal account, unquantified and unauditable, and it is neither part of this backtest nor evidence for it. Judge the method on the log.

There are two separate tests on this site and I had been describing them as one. Fixing that here:

The signal log — the Score run over the full available indicator history, 23 Sep 2011 to 30 Nov 2025. It produced 64 zone crossings, all published at /receipts, winners and losers alike. That is the record to judge the method by.
The DCA simulation on this page — a narrower thing: $50 a week from 1 Jan 2017 to 12 May 2026, 489 weeks, with the Score's exit and re-entry rules applied. It produces 14 signal-triggered trades. Those are not rows from the 64-signal log and are not a subset of it: the log records zone crossings, this records what a $50-a-week ladder did in response, on weekly data. Two different objects, described here as one until now. It starts in 2017 because that is where clean weekly price and on-chain series were available to me, which means it does not cover the 2013 turn. The signal log does.

That's the caveat, stated once. Underneath it is a fully specified method: every input, every weight, every signal it produced — and every one of them is recomputable from CBBI's free public data.

DCA Sim On This Page

489 wks

Jan 2017 → May 2026

Indicators Weighted

every weight published

Signal Log, Separately

Sep 2011 → Nov 2025

Max Drawdown

-60%

vs -79% for pure HODL

Where the model fired, against price

BTC price across the tested period, with every signal the method produced marked on it — red for exits, green for buys. This is the chart worth checking: it shows when the model spoke and what the market was doing at the time. Hover any point.

What the exit discipline actually changed

Pure DCA, no cycle exits

-79%

deepest drawdown you'd have had to sit through

Same money, with the Score's exits

-60%

shallower ride, same contributions

That gap is the whole argument for having a written exit, and drawdown is the measure that shows it most plainly. Backtested over Jan 2017 – May 2026; past performance does not predict future results.

Every transaction in the DCA simulation, 2017 onward

$50 DCA buy (every Monday) Cycle-top SELL signal Cycle-bottom BUY signal

Every signal-triggered trade in the simulation (14 of the 64)

Why these dates don't match /receipts. This simulation fires on a tiered ladder — it sells in four tranches once the Score enters the exit band and buys back in tranches on the way down — so its trade dates are the tranche dates, not the zone-crossing dates. The zone log at /receipts records the crossing itself. Example: the ladder's first 2024 sell tranche lands 15 Dec 2024 at $104,551, while the Score's exit-zone crossing was 16 Nov 2024 at $90,568. Same model, two different questions. If a figure from this page ever appears anywhere else on the site, it must be labelled as the ladder simulation, not as a signal.

The full signal log — all 64 zone crossings, Sep 2011 to Nov 2025 →

The system that produced this isn't a secret.

The exact indicators, thresholds, and execution rules used in this backtest are taught in the LiftOffr course. Module 4 covers the indicators. Module 5 covers the execution.

Get the plan — $29 once →

One-time payment, no subscription. The scoreboard is free →

The DCA simulation on this page is based on historical BTC prices and on-chain indicator data (CBBI components from colintalkscrypto.com) from January 2017 through 12 May 2026. The separate 64-crossing signal log at /receipts runs from September 2011 to November 2025. Past performance does not guarantee future results. This is not financial advice. Trading carries risk — only invest what you can afford to lose. Backtest does not include taxes, exchange fees, or slippage. Real-world execution may differ.