The Rapture Signal 155.2 Easing 100 = 2024 · through 18 Aug 2026 Get tomorrow’s Signal →
WATCHING FOR THE BLESSED HOPE — TITUS 2:13
← BACK TO THE WIRE
LAST UPDATED · 19 AUG 2026

How The Rapture Signal™ Is Made

BUILT LIKE AN INTELLIGENCE PRODUCT. NOT AN OPINION.

THE RAPTURE SIGNAL™ is a daily reading of how loudly the world's news is running on the signs the Bible names, measured against a fixed 2024 anchor of 100. This page is the full derivation: what is collected, how long the record runs, the exact computation, and how a stranger can check every claim. Nothing on this page is a forecast, and no probability of any prophetic event is computed anywhere in the system.

The Pipeline

One diagram, the whole system: where the data comes from, where it is aggregated, and how the number is compiled. Every step runs by script; a person starts the scripts and ships the result. The only human hands in the system write the wire. They never touch the Signal.

The Rapture Signal™ pipeline, collection to verification 01 · COLLECTION THE GDELT PROJECT RAW PRESS FEED, PUBLISHED AS MINUTE FILES DOWNLOADED DAILY, ARCHIVED, COUNTED LOCALLY FIXED QUERY DEFINITIONS, HELD WITH THE ARCHIVE 02 · AGGREGATION NINE WATCH-AREA QUERIES · EACH ANCHORED TO SCRIPTURE ISRAEL & MIDEAST WARS & RUMORS BABYLON RISING MARK OF THE BEAST SIGNS IN THE EARTH FAITH & APOSTASY DAYS OF NOAH STRONG DELUSION RISE OF MACHINES DAILY COUNTS · NINE SERIES 03 · THE RECORD THE RECORD 130 WEEKS · CONTINUOUS FIXED 2024 ANCHOR EACH AREA INDEXED TO ITS OWN 2024 AVERAGE · 100 = ITS 2024 NORMAL 11-YEAR BACKDROP GDELT GKG · BACK TO 2015 SECOND SENSOR NEVER AVERAGED IN NINE SERIES SMOOTHED TO ONE MEAN CONTEXT ONLY 04 · COMPUTATION FIXED-ANCHOR COMPUTATION INDEXED TO 2024 = 100 · SMOOTHED 13 WEEKS COMPUTED TO ONE DECIMAL THE BIRTH PAINS MONITOR · SAME SERIES, THREE READS VELOCITY 1ST 4-WEEK DIFFERENCE ACCELERATION 2ND DIFFERENCE · ±2 SYNCHRONY 26-WEEK CORRELATION COLOR BY FORMULA · RED ONLY AT RECORD HIGHS AND ACCELERATING SEALED PIPELINE · NO EDITORIAL INPUT A HUMAN EDITOR CANNOT MOVE THE NUMBER 05 · PUBLICATION THE RAPTURE SIGNAL™ 2024 ANCHOR = 100 · ONE DECIMAL · NEVER WITHOUT ITS BAND ABOVE 100 = LOUDER THAN 2024 RAN. NOTHING MORE. PUBLISHED BESIDE THE WIRE 06 · VERIFICATION EVERY READING DOWNLOADABLE · CC-BY-4.0 tempo-history.json birthpains.json tempo_fixed.json backdrop_series.json ONE NUMBER A STRANGER CAN CHECK.

Collection

SENSOR
The GDELT Project publishes its raw press feed, minute by minute, as public files. We download those files every day, keep every one permanently in our own archive, and count matching English-language articles ourselves against fixed query definitions. Forty-eight minute-windows are archived per day, evenly spaced around the clock, and all of them are counted. Because we hold both the raw files and the queries, any day's number can be recomputed from the archive.
COVERAGE
Nine biblical watch areas, each anchored to a passage of Scripture: Israel & the Middle East, Wars & Rumors, Babylon Rising, Mark of the Beast Watch, Signs in the Earth, Faith & Apostasy, Days of Noah, Strong Delusion, Rise of the Machines.
INPUT
Measured press volume. One input only: no editorial scoring, no story weighting, no vibes. A human editor cannot move the number.

The Record

PRIMARY
More than 130 weeks of continuous readings across the nine areas, on a fixed 2024 anchor, with daily collection begun August 2026. Collection can stall; when it does, every reading carries the date its data actually reaches. Daily sampling adds resolution, not history.
BACKDROP
An 11-year historical context series from GDELT's Global Knowledge Graph, reaching back to 2015. Kept as a second instrument on a second sensor, never averaged with the primary.
BASELINE
Each area is indexed against its own 2024 average, so 100 is that area's own normal, not an arbitrary scale.

The Method

The Signal. Each of the nine areas is measured as its share of the day's press corpus, indexed so that 100 equals its own average across 2024, then averaged across the nine and smoothed on a trailing 13-week window. THE RAPTURE SIGNAL™ is that smoothed average, computed to one decimal and sampled daily: each morning's reading takes the same instrument over the trailing 91 days of usable data, a construction arithmetically identical to the weekly value at every week boundary. That identity is tested before every publish, and the number does not ship on a day the test fails. 100 is the 2024 anchor, not a ceiling and not a score out of 100. A reading of 150 means press volume on these nine signs is running 50 percent above where it ran in 2024. The anchor year is arbitrary; it is where the record begins, and a different anchor would give different numbers.

Why the rank is not the headline. Until 11 August 2026 the headline was the reading's percentile rank within the record. That number had gone dead: the record has sat at its own highs for months, one rank step is 1/130 of a point, and across 120 days the rank took six distinct values and read exactly 100.0 on 78 of them while the underlying reading swung from 137.4 to 158.9. A number that cannot move cannot inform, and a daily number that does not move is worse than a weekly one. The rank is still published, as a gauge beneath the headline, where a saturated statistic is honest context rather than a dead lead. It also still governs the colour rule, which was always rank language: red only at record highs and strengthening.

Three limits publish with the number. Daily sampling does not lengthen the record, which remains about two and a half years. Consecutive readings share 90 of their 91 days, so they are never counted as independent observations. And the number always carries the date its data actually reaches: a stale collection lane marks the reading PROVISIONAL rather than letting an old number wear today's date. The Signal never appears without its band word or a plain comparative sentence.

Beneath the number. The Birth Pains Monitor reads the same series three ways, all published: velocity and acceleration (first and second 4-week differences: ACCELERATING above +2, EASING below −2, STEADY between), and synchrony (mean pairwise correlation of week-over-week changes across the trailing 26 weeks: below 0.15 the signs move independently, above 0.40 they are locking).

Color is formulaic. Red requires the record at highs AND accelerating. Amber requires highs with acceleration merely positive. Everything else, including record highs that are easing, renders in plain ink. An editor cannot set the color.

The Tempo. The same calculation across all nine areas at once, indexed so 100 equals the previous 13 weeks of aggregate volume. It counts column inches. It does not count meaning.

Verification

Nobody has to trust us. Every reading is downloadable, license CC-BY-4.0, attribution "Rapture Radar™ (raptureradar.com)":

The corrections run on the record at /corrections, with a standing policy. Sourcing rules are published at /sources. The full instrument methodology, with every disclosure, lives on THE RADAR.

The Limits, Stated

THE BASELINE. The 100-line is the 2024 average of GDELT-measured press volume, not of this wire's own story counts. It swaps to the wire's own record once that log spans a full year; the log began in August 2026. Until the swap, this sentence is the disclosure. A stale collection lane still marks the reading PROVISIONAL rather than letting an old number wear today's date.
SENSOR CONTINUITY. Before August 2026 the counts came from GDELT's search service. When that road closed we moved to the raw files. The two methods were run side by side across their overlap, and the new counts were scaled to the old series lane by lane before a single day joined the record, so the seam has no artificial step. One lane whose new counts failed to track its old series was held on its old data until its query was corrected and revalidated. The validation log is kept with the project records.
VOLUME ONLY. A quiet week of grave news reads low here. The instrument does not judge significance. That is a real limitation, stated rather than hidden, and it is why the Signal is published beside the wire and never instead of it.
ENGLISH ONLY. The queries count English-language articles. Events with dense English coverage read larger than equivalent events without it. The wire is swept far more widely than the instrument.

The Covenant

We don't measure when Christ will return. We measure what He told us to watch. The Signal records news intensity in the areas believers are told to watch. It does not measure, estimate, or predict proximity to the fulfillment of any prophecy. No one knows the day or the hour. We watch. We do not predict.

SEND ME TOMORROW'S SIGNAL →
100% FREE · FIVE MINUTES · UNSUBSCRIBE ANYTIME
ONE NUMBER A STRANGER CAN CHECK.
WE REPORT. GOD DECIDES.
THE WIRE · THE RADAR · QUAKES · FULL INDEX · ANSWERS · ABOUT · METHODOLOGY · SOURCES · CORRECTIONS · CONTACT
ARE YOU READY? · PRIVACY · TERMS · DMCA · RSS · FOLLOW ON X
RAPTURE RADAR · THE RAPTURE SIGNAL™

no probability of any prophetic event is computed here, and none ever will be

Page updated · reviewed · content · search · AI accuracy