Practice · Indicator-off curriculum

The naked-chart phase — 30 days before the indicators come on

The audit found that the FibLab indicator stack (Hold Levels, Break/Origin Detector, Origin Bot, Bias Scanner, Fib-Lab Dashboard, Leviathan) hides the discretion that IS the skill — and creates a single point of failure if any of those tools breaks. This 30-day phase builds the manual reflex first. Indicators come on at day 31 as confirmation tools, not crutches.

The premise: if the indicators were doing the pattern recognition for you, you've outsourced your edge to TradingView's uptime. If you can mark holds, breaks, and origins on a blank chart with ≥85% agreement to the indicator, the indicators are confirming what you already see. If you can't, the indicators are deciding for you — and you'll have no idea why a trade failed when one of them breaks.

The four skills you're building

Skill	What it looks like	Why it matters
Mark holds	On a static chart screenshot, draw the hold line at the body of the engulfed candle. Body engulfment only — no wick engulfments yet.	The Hold Levels indicator does this automatically. If you can't reproduce it manually, you can't tell when the indicator is showing you a non-standard hold (wick mode left on, etc.).
Mark breaks	Identify the range, then mark the wick extreme. Distribution top = upper break; accumulation bottom = lower break.	Break placement determines stop placement. If you delegate it to the indicator, your stops drift when the indicator misclassifies the range.
Detect non-pure ranges	Count candles. 2 = pure. 3+ = non-pure → "investigate higher TF." Make the call before opening the dashboard.	This is the single most valuable manual skill — the dashboard tells you the true TF, but you need to know when to ask it. That decision starts with the candle count.
Spot origins	Identify a break attempt that didn't body-close past, then mark it as "pending origin." After the conversion candle, mark it as "active origin."	The Origin Bot fires alerts. Spotting them manually forces you to reason about the conversion event — which is what the RT-vs-hindsight box on layer-1 chapter 1.1 warns about.

Week 1 · Holds & Breaks

Foundation

Daily exercise: open BTC 4H. Hide all FibLab indicators. Screenshot. On the screenshot, manually mark every hold and break you can find in the last 30 days. Save the marked screenshot.

D1BTC 4H · holds only · last 30 days

D2BTC 4H · breaks only

D3SOL 4H · holds + breaks

D4ETH 4H · holds + breaks

D5BTC 1H · holds + breaks (faster, more)

D6BTC Daily · holds + breaks (slower, fewer)

D7Calibration: turn indicators on. Compare. Note disagreements.

Pass criterion: by D7, ≥80% of the holds you marked match what the Hold Levels indicator drew. If not, repeat week 1 before continuing.

Week 2 · Pure vs Non-Pure

TF discipline

For every range you marked in week 1, count candles. If 3+, write down the next-higher TF you suspect the level actually lives on. Then verify with the dashboard.

D8BTC 4H · classify each range pure/non-pure

D9BTC 1H · same; expect more non-pure

D10BTC 15m · same; expect mostly non-pure

D11SOL 4H · classify + predict true TF

D12ETH 4H · same

D13BTC Daily · classify (most should be pure)

D14Calibration: dashboard's true-TF vs your guess. Score yourself.

Pass criterion: ≥75% of your true-TF guesses for non-pure levels match the dashboard within ±1 TF step. The point isn't perfect — it's developing the reflex of asking "is this on my chart's TF, or borrowed from above?"

Week 3 · Origins (the hardest)

Conversion event recognition

For every break you marked in week 1, classify: did it convert to an origin? When? Which candle was the conversion event? Then check against the Origin Detector.

D15BTC 4H · find every origin in last 30 days

D16SOL 4H · same

D17ETH 4H · same

D18BTC 1H · classify each origin's conversion candle

D19BTC Daily · find singularities (1 TF only)

D20BTC 4H · count touches: 1st, 2nd, 3rd. Measure each touch's outcome (held / failed)

D21Calibration: Origin Detector vs your marks + first-touch hit-rate vs published 85-90%

Pass criterion: ≥70% of origins you marked also appear in the Origin Detector. AND your measured 1st-touch hit-rate from D20 is within ±15 points of the published 85-90% — if it's lower, you're either marking weak origins, or the published rate is sample-biased on the videos. Either matters.

Week 4 · Live mark + paper trade

Real-time discipline

Indicators stay off. For real candles closing today, mark holds / breaks / origins as they form. Paper-trade the setups. Log to the journal as you would for live.

D22Live mark BTC 4H · paper trade what you'd take

D23Same · note what you skipped and why

D24Add SOL · two assets at once (the realistic load)

D25Live · 4H + 1H stack. Refine 4H signals on 1H

D26Live · attempt a greedy refinement (15m or below)

D27Rest. Re-read the four loss walkthroughs

D28Final calibration: turn indicators on. Compare 6 days of live marks. Score.

Pass criterion: ≥85% live-mark agreement to indicators across the 6 live days, AND your paper-trade journal shows you skipped at least 50% of "looks valid" setups. (The skip discipline matters more than the recognition.)

Days 29–30 · Decision

Graduation

Two outcomes possible:

You passed all 4 weeks (≥85% agreement, skip discipline visible). Indicators come on at day 31 as confirmation tools. You now have a manual fallback for any tooling outage. You also have 4 weeks of paper trades in the journal — start tracking them in the trade-log gating page to figure out your starting tier.
You did not pass at least 2 of the 4 weekly calibrations. Indicators stay off. Go back to whichever week you failed and repeat. Cap risk at the Apprentice tier (0.25%) until you've completed 30 days. There is no shortcut — the indicators don't help if the manual reflex isn't there.

How to actually do the weekly calibration

Step 1. Take the screenshots from the week (you saved them, right?). Side by side with the indicator-on version of the same chart at the same zoom.

Step 2. Count: how many holds (or breaks, or origins) did the indicator draw that you also marked? How many did it draw that you missed? How many did you mark that it didn't?

Step 3. Agreement % = (you ∩ indicator) / (you ∪ indicator). Yours-not-theirs counts as a miss; theirs-not-yours counts as a miss.

Step 4. For every disagreement: write 1 sentence in your journal explaining why. "Indicator marked a wick engulfment as a hold — I had body-only on" is a useful note. "I marked X but indicator didn't" might mean you found one it missed (good), or you marked something that's not actually a hold (calibration target).

What this prevents: the median trader losing 40-70% in year one (per the audit) is sized for Syndotc's hit rate, not theirs. Syndotc has thousands of repetitions of manual pattern matching across these same charts. The naked-chart phase is one month of the same work, distilled — enough to know whether your eye is calibrated or whether you've been confusing "I see this on the indicator" with "I can identify this."

Companion reads: Layer 4 · Indicators (the tools you'll turn back on at day 31) · Antichart pairs · Trade-log gating.