Nine on-chain indicators, weighted into a single 0–100 Score, recomputed every morning from public data. Every weight is published below, so you can rebuild the number yourself and check that I am not making it up.
The six bands, in full — these are the only ones, and they are the same on every page, in the Sunday email and in the Discord bot:
Every one of those is a description of what the band has done before. None of them is a prediction — see the record.
I found a bug in my own arithmetic on 20 August 2026 and shipped the fix the same day. This is exactly what changed, and why the number it replaced was wrong. When the data source publishes no reading for one of the nine on a given day, my code was treating that gap as a reading of zero and averaging it in, instead of leaving it out. On 19 August that understated the Score by 1.8 points — it read 34.5 when it should have read 36.3. (The 34.4 quoted further down is the same bug one day earlier, on 18 August, which is the date that recompute was run.) Anyone who pulled the data and recomputed it properly would have got a different number from mine and would have been right. A missing reading is now excluded and the divisor drops to match, which is what the table above shows. The 64-signal record at /receipts is unaffected — it was generated by a separate script that always handled gaps correctly.
These are the exact numbers the live Score uses. Nothing is held back, and there is no tenth indicator behind a paywall.
| RHODL Ratio | 20% |
| Puell Multiple | 20% |
| Trolololo trend line | 15% |
| MVRV Z-Score | 15% |
| Pi Cycle Top | 10% |
| 2-Year MA Multiplier | 5% |
| Reserve Risk | 5% |
| Woobull Top Cap | 5% |
| RUPL | 5% |
Two things worth saying plainly: CBBI is the data source these readings are pulled from, not a component, and the Score contains no Fear & Greed and no Google Trends. I wrote about both for a long time; neither carries any weight in the number.
Want to check my working rather than your own? /api/cycle-score returns every component, its weight and its reading as JSON. Open it. An outside recompute on 18 August 2026 got 34.35 against a published 34.4 — and both of us were wrong by about 1.8 points, because we had each treated a missing reading as a reading of zero. Two people agreeing is not two people being right. That is why the offer below matters more than the match did.
If you run it and get a materially different number, tell me and I will fix it. That offer is the entire reason the weights are public.
Knowing where the cycle sits is not the same as knowing what you’ll do about it. The people who round-trip a cycle are rarely the ones who lacked information — they are the ones who had no rule written down before the moment arrived, and then made the biggest financial decision of the cycle on the worst possible day to make one.
The $29 plan is the nine levels I have written down for my own position, with the reasoning behind each one and a worksheet that turns your own stack size into your own ladder. One document, one payment, nothing renews. Read it for 30 days and if it isn’t worth $29, I refund it — no form, no reason needed.
See what’s in the plan — $29, once →Every zone crossing this model has ever produced — all 64, from September 2011 to November 2025 — with the date, the price, and what Bitcoin actually did over the next 30, 90 and 180 days. Including the 2017 exit signal after which Bitcoin rose 149% in a month.
Scored directionally, the model was right on 21 of 46 directional crossings at 30 days and 25 of 46 at 180. Neither figure is distinguishable from chance — 21 and 25 of 46 carry the same p-value, and it would take 31 of 46 to clear significance. It is published rather than buried, and it is not a win rate.
See all 64 signals →One email a week: where the Score sits, the largest weight change behind it, what that zone has historically meant, and what I am doing with my own stack — which is a disclosure, not an instruction. No trade calls, no tickers, no “buy now.”
Unsubscribe in one click. 18+ only. Educational content, not financial advice, and nothing in it is personalised to you.
Want the levels I am personally buying and selling at, written down in advance? That is the $29 plan. The Score above stays free either way.