When a test is flagged for an article

Apply the findings log's p-value ceiling and independent-point floor to two current tests, one comparison at a time.

Written by Dean Shalem, published Sep 29, 2026, about 4 minutes to read.

A gate with two bars to clear: a p-value ceiling bar and an independent-points floor bar; two test tokens, one clears both bars and gets a green flag, one clears only one and stays unflagged. A small MinerTerminal mark sits in the top-left corner. — Illustration: AI-generated artwork created for MinerTerminal in ChatGPT from original article concepts.

How this article reads

Signal. Apply the findings log's p-value ceiling and independent-point floor to two current tests, one comparison at a time.
Mechanism and evidence. The "Two findings against the p-value ceiling" chart, drawn from two rows of the findings log, one built on filed figures and one on daily closing prices.
Implication. Only the CleanSpark leverage row clears both the p-value ceiling and the independent-point floor in this example.
Next catalyst. Where the figures update: the analysis page behind this example.
Sources. The sources list at the foot carries every document.

Corrected

Oct 3, 2026: The worked example printed the p-value gap for the hashrate and revenue test as -0.1021, beside the ceiling 0.01 and the p-value 0.112, so a reader's subtraction gave -0.102. The gap lines now print to three decimals: -0.102 for that test, and 0.010 for the CleanSpark leverage test, which read 0.0100.
Oct 3, 2026: The chart "Two findings against the p-value ceiling" printed the CleanSpark leverage test's p-value as 0 in its table and its hover labels, and the other two figures as 0.11 and 0.01. It now prints "under 0.0001", 0.112 and 0.010, the way the analysis tables print a p-value.
Oct 2, 2026: The Mechanism and evidence line under How this article reads said the chart "Two findings against the p-value ceiling" was drawn from the filed sample the sources list carries. It now says the chart is drawn from two rows of the findings log, one built on filed figures and one on daily closing prices.

The corrections log gives each change with the document that decided it.

Work through the figures.

Two findings against the p-value ceiling

Two current rows beside the p-value ceiling for article candidates. As of Oct 14, 2026.

Method and sources

Only the CleanSpark leverage row clears both the p-value ceiling and the independent-point floor in this example.

Covers: current findings log
Unit: p-value

Source: MinerTerminal analysis, computed from the cited source data

Two findings against the p-value ceiling
Finding	p-value (p-value)
Hashrate and revenue, n 11	0.063
CleanSpark leverage, n 365	under 0.0001
Candidate ceiling	0.010

A test earns a place in the Learn section when two conditions hold at the same time: its p-value sits below a set ceiling, and the count of independent points behind it reaches a set floor. Two comparisons, both must pass. That is the whole rule.

For each row, the p-value is the probability, under that row's null hypothesis and test assumptions, of a result at least as extreme as the one observed. It is not a verdict on whether the relation is real.

We walk through the rule on two rows from the findings log. The first is the level test of revenue against hashrate, filed with a p-value of 0.063 on 11 independent points. The second is the longest fixed-window leverage test for CleanSpark, filed with a p-value of under 0.0001 on 365 sessions.

The rule sets two bars. The p-value ceiling is 0.0100. The minimum point count is 8. Pooled quarter-to-quarter tests never enter this check.

The arithmetic below subtracts each value from its bar. A positive gap on the p-value line tells us the p-value is below the ceiling. A gap of zero or more on the count line tells us the count meets the floor. It is just subtraction, one line at a time.

P-value gap, hashrate-revenue test:

-0.053

Count gap, hashrate-revenue test:

P-value gap, CleanSpark leverage test:

0.010

Count gap, CleanSpark leverage test:

357

We ran two rows through two comparisons. The row where both gaps are non-negative is flagged for the Learn section; the other is not. That is the rule doing its work.

Sources

See the rule applied to every test: the analysis page behind this example.