Every bold idea about the universe owes its readers one answer: what would show it is wrong? Ricardo Maldonado, author of GOD PLAYS DICE, built his book around that question. His idea, HD-Blast, is an Author’s hypothesis. It says our hot early universe may be the aftershock of a brief blast in a larger space with more dimensions.
An idea like that has to meet the sky somewhere, and the book picks the place. It points to a faint, slow hum of gravitational waves. These are ripples in space itself. Astronomers listen for them with pulsar timing arrays. A pulsar is a dead star that spins and sends radio pulses toward Earth, as steady as a clock. When a slow ripple passes, some pulses arrive a tiny bit early and others late, in a shared pattern across the sky. In June 2023, the NANOGrav team reported evidence for such a hum in 15 years of data on 67 pulsars.
Here is the bet. Picture a graph of how loud the hum is at each rhythm, from slow to fast. On the kind of graph scientists use, the simplest black-hole story is a straight line. HD-Blast draws a smooth hill with a bend, or “knee,” near one cycle per year. In December 2025, the author froze five numbers that fix that curve in a public record. Freezing matters. It stops anyone from moving the target after new data come in.
So the clearest way to prove this version wrong is easy to state. Wait for more years of pulsar data, fully analyzed by the teams that collect it. If the hum stays a straight line through that range, the benchmark fails. If different teams, after matched analyses, find no common bend, the shared-knee version weakens too.
The test is lopsided, and the book says so. Pairs of giant black holes are the leading explanation for the hum, and they can bend the curve too. So a bend would not prove a blast. But no bend would count hard against this version.
How do pulsars let us hear slow ripples in space?
Established. A gravitational wave is a ripple in space and time. As it passes, it stretches and squeezes distances by a tiny amount. Detectors like LIGO catch fast ripples from colliding black holes. But some ripples are so slow that one wave takes years to decades to pass. No detector built on Earth is big enough for those, so astronomers use the galaxy itself.
The tools are pulsars: collapsed stars that spin and sweep radio beams past Earth like lighthouses. The fastest spin hundreds of times each second, and their pulses are very steady. Teams time dozens of them for many years with large radio telescopes. For each pulsar they predict when every pulse should arrive, then study what is left over. A passing slow wave delays some pulses and advances others. The key is the pattern. For a background of gravitational waves, pairs of pulsars should drift together in a way that depends on how far apart they sit on the sky. That pattern is called the Hellings–Downs curve, after the two scientists who described it in 1983.
Supported, still debated. In June 2023, NANOGrav reported evidence for this pattern using 67 pulsars from its 15-year data set. The strength was about 3 to 4 sigma, depending on the method. Sigma measures how surprising a result would be if no real signal were there; 5 sigma is a common bar for a discovery. Teams in Europe, India, Australia and China reported evidence at varying levels. When the International Pulsar Timing Array compared its member teams’ results, their estimates of the hum agreed within about one sigma.
What exactly does HD-Blast bet on?
Author’s hypothesis. HD-Blast says a brief, violent event in a higher-dimensional space may have set up our hot early universe. Its surviving mark today would be part of this slow hum. The book turns that picture into a curve. Imagine a graph with slow rhythms on the left, fast ones on the right, and loudness going up. The HD-Blast curve rises, bends at a “knee,” and falls. The knee sits at 3.16 × 10⁻⁸ hertz, a rhythm of about one cycle per year.
Five numbers fix the curve: where the knee sits, how loud it is there, the two slopes, and how smooth the bend is. They were frozen in a public Zenodo record dated December 17, 2025. Zenodo is an open research archive run at CERN, Europe’s particle-physics lab. A record there fixes a date and its contents. The book states that it is not peer review.
There is an honest catch. The numbers were chosen after the author had already looked at the summary points bundled with his analysis. The book calls this calibration, not prediction. Matching the same points you tuned to shows nothing new. The fair test starts with later, independent data, or with a full analysis written down before anyone looks.
Which future results would count against it?
The book lists its own failure points. Here they are in plain words.
- A straight line. If more years of data show the hum following a simple straight line (a “power law”) through the knee region, the benchmark weakens. The book’s glossary says a later straight spectrum would reject the version tested in the book.
- The bend fades under a full analysis. The hint of a bend comes from a simplified screen the author ran on summary points. If the pulsar teams’ own full analysis, with all its noise modeling, finds no preference for a curve, the case weakens.
- The teams disagree. If careful, matched analyses by different teams put the knee in different places, one shared knee is hard to defend. The book compares it to three doctors taking your temperature and getting three different numbers.
- Black-hole pairs explain everything. NANOGrav expects finer details of the spectrum, how evenly the hum is spread across the sky, and signals from single black-hole pairs standing out above it to reveal more about the hum’s origin as data grow. If giant black-hole pairs account for all of it, the book’s glossary says the blast idea loses, and calls that fair.
- Other guardrails. If the curve came from the early universe, its energy must stay small enough not to upset the first minutes of cosmic history or the oldest light. Its tail must also stay quiet where LIGO, Virgo and KAGRA listen, and where the LISA space mission, planned for launch in 2035, will listen. The author’s archived calculations put the curve far below today’s limits and below LISA’s planned sensitivity, which suggests these checks are less likely to be the deciding test.
- Black holes ringing off-key. The simplest version assumes gravity today follows Einstein. If merged black holes ring, again and again, at notes that differ from his theory’s predictions, that assumption breaks. So far they match: GW250114, which LIGO scientists call the clearest gravitational-wave signal observed so far, matched general relativity in every test, within current precision.
What is the strongest objection?
The strongest objection is that a bend is not a fingerprint. Giant black-hole pairs can bend the curve too. NANOGrav’s paper on black-hole binaries says realistic models can reproduce both the loudness and the shape of the hum, though matching the loudness means pushing many model settings to the edges of their expected values, or a few well away from standard expectations. Its 15-year paper adds that binaries interacting with nearby stars and gas could weaken the hum at the slowest rhythms. A simulation study by Kelley and colleagues found that above about one cycle per year, the slope of a black-hole background carries information about how many binaries there are. That is right where the knee sits.
This makes the test lopsided. Finding the knee would not prove a blast. It would move the idea up one rung, from a curve on paper toward a feature in real data. Missing the knee would hurt it badly. The book accepts that trade and says a lasting bend would not identify a higher-dimensional blast by itself.
A second objection is sharper. No working blast mechanism yet produces the curve. The book says so, and the author’s newest HD-Blast record, from September 2026, lists a hot-Big-Bang mechanism as not established. A critic could ask: if the knee fails, won’t a new curve simply replace it? The book’s answer is a set of rules. Keep the old record. Date every change. Say what failed and why. A later dataset is not a fair test if the target is quietly adjusted after the new points are seen.
What do people often get wrong?
- “Evidence” is not “detection,” and neither is “source.” NANOGrav reported evidence for a background. That is a big result. It does not yet say which source made it.
- “New physics fits better” is not new physics. NANOGrav’s own new-physics paper found that many early-universe models fit the signal, some with Bayes factors of 10 to 100 over the black-hole model. (A Bayes factor is a ratio of how well two models explain the data.) The team warned that these results depend strongly on assumptions about the black holes and “should not be regarded as evidence for new physics.” The same caution applies to HD-Blast.
- A DOI is not a stamp of approval. The book states that a Zenodo DOI preserves a public record. It does not mean peer review, independent checking, or endorsement by CERN or any pulsar team.
- Failing this test would not end every extra-dimension idea. The book says a rejection would reject the canonical version tested there, not every possible higher-dimensional cosmology. That is scope, not an escape hatch. The version on the table would be gone.
- The author’s life story is not evidence. The book’s first chapter is memoir. It explains why the question matters to him. It adds no weight to the hypothesis.
How can you check it yourself?
- Open NANOGrav’s 15-year evidence paper and find the plot of the spectrum, often shown as a row of separate points. Look at the right-hand, faster end. That is where a knee near one cycle per year would show.
- Read the authors’ most careful sentence. The paper says the signal fits black-hole-binary expectations, but that more unusual sources cannot be excluded.
- Open the December 17, 2025 Zenodo record and write down the five numbers. When new pulsar results arrive, check whether the same numbers are being tested, or new ones.
- Ask the book’s five questions of any headline. Did the feature last as data improved? Did the teams agree after matched analyses? Does the energy fit the early universe? Does the curve stay under limits in other bands? Are the black-hole tests based on real events, or on a demo?
- Watch for the International Pulsar Timing Array’s third data release. It had not come out when this page was checked in September 2026. It is designed to combine data from pulsars timed by more than one team.
Deep dive (optional)
The five frozen numbers
The benchmark is a “smooth broken power law”: one slope below the knee, another above it, and a smooth join. The archived point is fk = 3.16 × 10⁻⁸ Hz (the knee), Ωk = 8.00 × 10⁻⁹ (the curve’s height at the knee, written as a fraction of the universe’s energy density), α₁ = +3.0 and α₂ = −2.0 (shape exponents) and Δ = 2.0 (smoothness). In the book’s sign convention, the curve grows like f² well below the knee and falls like f⁻³ well above it.
Where the knee sits in the data
One cycle per year is about 3.17 × 10⁻⁸ Hz, because a year lasts about 31.6 million seconds. NANOGrav’s standard 15-year analysis used 14 frequencies from about 2 to 28 nanohertz. The knee, near 31.6 nanohertz, sits just above that range. The book notes that the highest-frequency pulsar-timing point carries the most weight for the apparent curvature. If that point shifts in later releases, the case could change.
How strong the author’s screen really is
In a simplified screen run on bundled summary points, the knee curve scored about 24 units of BIC (a model score where lower is better) ahead of a fixed black-hole template, and about 13.6 ahead of a free straight line. (The record gave 26.21 for the first gap; the book, using one comparison method throughout, gives about 24.) A separate curvature check scored about 8. The book calls these internal screening scores that cannot be promoted to publication-grade evidence.
The December 2025 record’s own summary used stronger words. It called the knee curve “strongly favored,” quoted about 3.2 sigma for the curvature check, and marked a cross-team overlap score as a pass. The book walks those labels back. It says the curvature score is not a 3.2-sigma detection, and that the overlap score came from a workflow demonstration with stand-in inputs, not from teams measuring the same knee. That is the Chapter 11 lesson in action: a number can survive while the sentence around it has to shrink.
The guardrail numbers
For the assumed curve, the author’s calculation gives an extra early-universe radiation of ΔNeff ≈ 7 × 10⁻⁴; the record calls this well below typical limits of about 0.06 to 0.3. It also gives Ω ≈ 2.3 × 10⁻³⁵ at 25 hertz. LIGO, Virgo and KAGRA found no background there and set limits of about 2 to 3 × 10⁻⁹, depending on the assumed spectrum. Both figures are conditional calculations under stated assumptions, not measurements.
THREE THINGS TO REMEMBER
- The clearest test is the shape of the slow hum: a bend near one cycle per year, or none.
- Finding the bend would not prove a blast, because black-hole pairs can bend the curve too.
- A failed benchmark would sink this version of HD-Blast, and the book commits to saying so.