Not peer-reviewed · an independent research program, shown honestly physics.magflowmeters.com

Early Universe: an independent timeline, checked against cosmology

We rebuild the history of the universe from one idea — that cosmic history is the history of increasing recordable distinction — and read off when each milestone lands. Then we lay our timeline next to standard cosmology’s and put the strongest checks first. This page is the side-by-side.

Forty-four tests, two independent routes: 30 confirmed agreements, 0 disagreements — and the 14 indeterminate split cleanly into 6 standing bets where both routes already agree and the experiment hasn’t weighed in yet, plus 8 questions that are open for all of cosmology. The whole denominator, always: 44 tests / 30 agree / 14 indeterminate / 0 disagree. Naming the open questions is the discipline that makes the agreements worth believing.
44Checkpoints
30Agree — confirmed by measurement
14Indeterminate — 6 awaiting measurement + 8 open for everyone
0Do not agree

Three results, nothing fancier

Agrees our number matches the established one, and the measurement confirms it Indeterminate honestly open — 6 rows where both routes agree and the experiment hasn’t voted, 8 open for all of cosmology Does not agree our number conflicts with data — there are zero of these

Every card below carries the same six lines — the data we’re calculating, shown whole: the number under test, both routes, the measurement, the honest agreement metric, and the source link. Nothing is graded on trust; every line opens.

The headline number: the age of the universe — 13.8 billion years

13.8 billion years Agrees — the master clock check (Test 01)

Reconstruct the age of the universe from a completely different starting question. Standard cosmology gets its timeline by solving the hot-Big-Bang equations; this framework gets its timeline from a single principle about when the universe can hold a stable, distinguishable record — what can be recorded? Run that tape forward and it keeps hitting the same marks: quarks confine at ~20 microseconds, neutrinos go free at ~1 second, the first nuclei lock in at ~3 minutes, atoms and the CMB at 380,000 years — landing on the same 13.8-billion-year-old universe today. Ask each road what time it is at every milestone, and the answers land on the same clock, epoch after epoch.

Number we’re testing
The cosmic age at each temperature milestone — the clock reading when the universe passes a given heat
Standard cosmology & measured
Standard cosmology here is the measurement itself — Friedmann expansion sets each age: neutrino decoupling at T ≈ 1 MeV (t ≈ 1 s); recombination at z* = 1089.92 (t* ≈ 3.8×10⁵ yr); BBN window t ≈ 180–1000 s · T₀ = 2.7255 ± 0.0006 K (Fixsen 2009/COBE-FIRAS); z* = 1089.92 ± 0.25 (Planck 2018 VI)
This framework (granularity)
The independent read: Each milestone lands on the same clock — same ages, same expansion physics; computed t(1 MeV) ≈ 0.73 s, t(155 MeV QCD) ≈ 2.2–2.6×10⁻⁵ s, t(0.08 MeV BBN) ≈ 2.0×10² s
Agreement
"Same timeline, same numbers" — every claimed event quantitatively consistent with the standard age–temperature relation; passed by agreement via shared physics, not two independent numbers (consistency check — shared inputs)

Check the source → Test 01

The second headline number: Neff = 3.044

Neff = 3.044 Agrees — hit by three separate tests (03 · 22 · 41)

Not three — 3.044. One second after the beginning, the neutrinos slip free a heartbeat before electron–positron annihilation pours its warmth into the photons alone. That act of timing fixes a fractional, non-obvious number, and the books close on it from both directions: standard cosmology’s entropy ledger and this framework’s record-cost ledger both total Neff = 3.044, stepping off the same cliffs at the same temperatures. The sky then votes: Planck measures 2.99 ± 0.17. Three tests in this suite hit the number from three angles — the degrees-of-freedom ledger (03), neutrino decoupling (22), and the cosmic-neutrino-background fingerprint (41), cross-confirmed by the free-streaming phase shift the CMB actually shows.

Number we’re testing
N_eff — the effective number of relativistic species: the radiation weight carried by the unseen sea of relic neutrinos (the g*(T) ledger and the T_γ/T_ν ratio)
Standard cosmology
N_eff = 3.044 — Standard-Model thermodynamics across the e⁺e⁻ threshold, including non-instantaneous decoupling and finite-temperature QED corrections (Bennett et al. 2021)
This framework (granularity)
N_eff = 3.044 — the record-cost ledger steps its species count down at the same temperatures, by the same amounts; the same (11/4)^(1/3) ≈ 1.401 photon/neutrino temperature ratio falls out of the same books
Measured
N_eff = 2.99 ± 0.17 (Planck 2018 CMB+BAO); BBN reads 2.86 +0.27/−0.28 (Fields et al. 2020); cross-confirmed by the free-streaming CMB phase shift detected at ~2–3σ and the mass-sum bound Σm_ν < 0.12 eV vs the oscillation floor ≳ 0.06 eV
Agreement
Both routes land on 3.044 — a fractional, non-obvious number — comfortably inside one standard deviation of Planck’s 2.99 ± 0.17, and hit by three separate tests in this suite (03 · 22 · 41) (consistency check — shared inputs)

Check the source → Test 03 · Test 22 · Test 41

The confirmed 30 — strongest first

Ranked by how hard each check is to fake: sharp numbers with tight measured bands first, shared conclusions and bookkeeping audits last. Every card states its result plainly and links straight to the page that shows the work.

23 · Big Bang Nucleosynthesis Light-Element Test

Agrees

Deuterium matches the measurement to about a fifth of a standard deviation and helium to less than one — and then lithium sits a factor of three high, put on the table in full view.

Number we’re testing
Primordial light-element abundances — deuterium (D/H), helium-4 (Y_p), helium-3, lithium-7
Standard cosmology
Standard BBN at Planck η_B = 6.12 × 10^-10, τ_n = 878.4 ± 0.5 s, N_eff = 3.044: D/H = (2.51 ± 0.11) × 10^-5; Y_p = 0.2470 ± 0.0002; ³He/H ≈ 1.0 × 10^-5; ⁷Li/H = (5.1 ± 0.4) × 10^-10
This framework (granularity)
No stable multi-nucleon record exists before this window; the abundances lock into a fixed record as it closes — same nuclear ledger, the framework reads it, it does not re-derive the inputs
Measured
D/H = (2.527 ± 0.030) × 10^-5 (Cooke et al. 2018); Y_p = 0.245 ± 0.003 (Aver et al. 2021); ³He/H ≈ (1.1 ± 0.2) × 10^-5; ⁷Li/H ≈ (1.6 ± 0.3) × 10^-10 (Spite plateau)
Agreement
D/H agrees within ~0.2σ (best-agreeing BBN probe); Y_p within ~0.7σ; ³He consistent (not a clean primordial probe); ⁷Li discrepant by factor ≈2.9–3 (~5σ) — the cosmological lithium problem, named openly (consistency check — shared inputs)

Check the source →

29 · Recombination Visibility Function Test

Agrees

The fog lifted in one cosmic instant at redshift 1089.92 — and both ways of asking when land on the identical number.

Number we’re testing
Recombination redshift z* — the visibility-function peak where the fog clears and light streams free — plus the last-scattering thickness
Standard cosmology
z* = 1089.92 ± 0.25, from the Saha–Peebles ionization calculation
This framework (granularity)
z* = 1089.92 — the same moment, read as when the atom-and-photon record first becomes stable and public (a consistency reading on the same measured inputs, not an independent forced number)
Measured
z* = 1089.92 ± 0.25 (Planck 2018); last-scattering thickness Δz ≈ 195; T(z*) ≈ 2973 K; θ* = (1.04109 ± 0.00030) × 10^-2 rad
Agreement
Both roads land on the identical redshift, 1089.92, with the same measured inputs and the same last-scattering thickness; the Saha suppression factor of ~53 below the naive 13.6 eV expectation checks out (consistency check — shared inputs)

Check the source →

28 · Matter–Radiation Equality Timing Test

Agrees

Take the textbook density ratio and you get 3422; ask the microwave sky directly and it answers 3402 ± 26 — the same fifty-thousand-year turning point to better than a percent.

Number we’re testing
Redshift of matter–radiation equality, z_eq — the moment matter density first overtakes radiation density
Standard cosmology & measured
Standard cosmology here is the measurement itself — z_eq ≈ 3402 ± 26 — a firmly settled, directly measured number (Planck's own direct fit to the CMB) · z_eq = 3402 ± 26 (Planck 2018 direct fit); cross-checked by the matter power spectrum turnover scale (eBOSS DR16)
This framework (granularity)
The independent read: z_eq ≈ 3422 — computed from the density ratio 1 + z_eq = Ω_m/Ω_r using Planck 2018 parameters (Ω_m h² = 0.1430, Ω_r h² = 4.177 × 10^-5 with N_eff = 3.044)
Agreement
3422 vs 3402 ± 26 — a match to 0.6%, well inside the measured uncertainty (consistency check — shared inputs)

Check the source →

31 · CMB Acoustic Peak Geometry Test

Agrees

The first acoustic peak sits at ℓ₁ ≈ 220, and the framework lands on exactly the same peak, in exactly the same place — there is no daylight between the two.

Number we’re testing
Position of the CMB TT acoustic peaks — first peak multipole ℓ₁ ≈ 220 (with ℓ₂ ≈ 537, ℓ₃ ≈ 810), plus peak-height ratios set by Ω_b h²
Standard cosmology & measured
Standard cosmology here is the measurement itself — Peak sequence from the ΛCDM fit (Planck 2018: sound horizon r* = 144.43 ± 0.26 Mpc, 100θ* = 1.04092 ± 0.00031), giving ℓ₁ ≈ 220 · ℓ₁ ≈ 220, a firmly detected peak (BOOMERANG → WMAP → Planck 2018; Planck Collaboration V 2020, A&A 641, A5), pinned to sub-percent precision by 100θ*
This framework (granularity)
The independent read: Same peak geometry, ℓ₁ ≈ 220 — the framework uses the identical acoustic calculation and reproduces the whole sequence rather than re-deriving it
Agreement
The number matches the number — ℓ₁ ≈ 220 both ways, pinned to sub-percent precision by 100θ* = 1.04092 ± 0.00031; 'no daylight between the two' (consistency check — shared inputs)

Check the source →

33 · CMB Lensing and Growth Consistency Test

Agrees

There was no freedom to arrange it — no separate knob to turn, so the theory had one shot and hit it.

Number we’re testing
CMB lensing amplitude A_lens — the smearing of the microwave-background pattern by all the matter its light later passed through
Standard cosmology
A_lens = 1 — a parameter-free consequence of the same base-ΛCDM model that already fits the primary temperature and polarization maps
This framework (granularity)
A_lens ~ 1 — the framework reads lensing as later cosmic structure re-encoding the earlier photon record, inheriting the same amplitude of 1 with no free dial to move it
Measured
A_lens^φφ = 1.011 ± 0.023 (Planck 2018 lensing reconstruction, multipoles 8–400) — a direct detection at greater than 40σ
Agreement
Fractional deviation from the parameter-free prediction of only ~1.1%, well within 1σ, on a >40σ detection; the separate A_L parameter runs high (1.180 ± 0.065, ~2.8σ) in the unlensed-spectra-only fit but returns to 1 with the lensing map (consistency check — shared inputs)

Check the source →

21 · Baryon-to-Photon Ratio Consistency Test

Agrees

One number the universe wrote down twice, in two different eras with two utterly unrelated pens — and the two copies match to about three percent.

Number we’re testing
Baryon-to-photon ratio η (as Ω_b h²), read independently from BBN deuterium and from CMB acoustic peaks
Standard cosmology & measured
Standard cosmology here is the measurement itself — Ω_b h² = 0.02237 ± 0.00015 (CMB, Planck 2018) and, read off completely separately, Ω_b h² = 0.02166 ± 0.00015 (BBN deuterium, Cooke et al. 2018, D/H = (2.527 ± 0.030) × 10^-5) · η_10 ≈ 6.12 (CMB) vs ≈ 5.94 (BBN) — the two roads agree to ~3% (fractional difference 3.2%)
This framework (granularity)
The independent read: Reads the baryon count as a single stable recordable quantity carried unchanged from nucleosynthesis to recombination — it consumes the measured η, it does not derive it
Agreement
Agreement to about 3%, a mild 1.5–2σ combined-uncertainty tension — framed in the literature as concordance, not contradiction; no late entropy or baryon-number injection required (consistency check — shared inputs)

Check the source →

07 · Inflation Scalar Spectrum Test

Agrees

The sky handed us a tilt to four decimal places — 0.9649 ± 0.0042 — and the geometry's window [0.9643, 0.9679] closes right around it; proving the roads had to meet there is the work still ahead.

Number we’re testing
The scalar spectral index n_s — the tilt of the primordial ripples (with amplitude A_s as companion input)
Standard cosmology
Single-field slow-roll: nearly scale-invariant, slightly red-tilted spectrum, n_s a little below 1 (n_s ≈ 1 − 6ε + 2η)
This framework (granularity)
n_s ∈ [0.9643, 0.9679] — the σ-inflaton exponential-plateau candidate run through standard slow-roll with the historical λ² = 1/6 slope (N* ≈ 50–60), carried forward strictly as a phenomenological fit, not a forced prediction
Measured
n_s = 0.9649 ± 0.0042 (Planck 2018 VI, TT,TE,EE+lowE+lensing, 68% CL); A_s ≈ 2.10×10⁻⁹ (ln(10¹⁰A_s) = 3.044 ± 0.014) at k₀ = 0.05 Mpc⁻¹
Agreement
'The predicted band brackets the measured tilt' — measured 0.9649 falls squarely inside [0.9643, 0.9679]; a match, not yet a lock (slope not geometrically forced: genuine slopes are 8/3, 4, 22/9) (consistency check — shared inputs)

Check the source →

35 · BAO Standard-Ruler Consistency Test

Agrees

One ruler, read twice across 13.8 billion years, refuses to disagree with itself.

Number we’re testing
The baryon-acoustic-oscillation standard ruler — the sound horizon at the drag epoch, r_d, a fixed length frozen into cosmic structure
Standard cosmology & measured
Standard cosmology here is the measurement itself — r_drag = 147.09 ± 0.26 Mpc (Planck 2018 base-ΛCDM, CMB-inferred); galaxy-survey distance ratios (D_M/r_d, D_H/r_d) at z = 0.15–2.33 agree with this to ~1–2% · r_d = 147.09 ± 0.26 Mpc — the one ruler, recovered by BOSS DR12 (~1% level), eBOSS DR16, and DESI DR1 (r_d h ≈ 101.8 Mpc, consistent at ~1–2σ per tracer)
This framework (granularity)
The independent read: The galaxy-scale ruler is a late-time fossil of the very same acoustic physics that set the CMB peaks — the framework recovers the same r_d, inherited without modification
Agreement
BAO distance ratios consistent with the Planck-calibrated r_d = 147.09 Mpc to within the quoted 1–2% measurement errors across the full z = 0.15–2.33 range (consistency check — shared inputs)

Check the source →

19 · QCD Confinement / Hadronization Test

Agrees

At a temperature the lattice pins near 156.5 MeV the strong force snaps shut — the instant ordinary matter, the stuff of stars and people, first becomes possible.

Number we’re testing
QCD confinement crossover pseudo-critical temperature T_c — the moment quark soup condenses into hadrons
Standard cosmology
T_c = 156.5 ± 1.5 MeV — continuum-extrapolated 2+1 flavor lattice QCD, roughly 20 μs after the Big Bang
This framework (granularity)
First separately-recordable strong-force objects (color-neutral hadrons) switch on precisely at the confinement crossover; the temperature itself is taken from lattice QCD, not independently computed
Measured
T_c = 156.5 ± 1.5 MeV (Wuppertal-Budapest, Phys. Lett. B 730 (2014) 99; HotQCD, Phys. Rev. D 96, 074510 (2017)); downstream fossil record η_B = (6.12 ± 0.04) × 10^-10, Ω_b h² = 0.02237 ± 0.00015 (Planck 2018)
Agreement
Consistency, not a second independent measurement — the framework lands where lattice QCD points; Γ_QCD exceeds H(T_c) by ~18 orders of magnitude (equilibrium hadronization), baryon number exactly conserved (consistency check — shared inputs)

Check the source →

20 · Lattice QCD Equation-of-State Transition Test

Agrees

Come at the most violent phase change of the early universe from record-cost bookkeeping and you trace the identical curve decades of lattice QCD nailed down — with no room left to nudge it.

Number we’re testing
QCD equation of state across the crossover — p(T), ε(T), s(T), trace anomaly (ε−3p)/T⁴, and g_*(T), anchored at T_c
Standard cosmology
Continuum-extrapolated lattice EoS curves (HotQCD, Bazavov et al. 2014; Wuppertal–Budapest, Borsanyi et al. 2014), with T_c = 156.5 ± 1.5 MeV (Bazavov et al. 2019)
This framework (granularity)
Record-cost reading tracks the same lattice curve — g_* ≈ 61.75 (hot quark-gluon soup) down to g_* ≈ 17.25 (hadron gas); it reads the curve, it does not re-derive it
Measured
T_c ≈ 156.5 ± 1.5 MeV, cross-checked at 155 ± 9 MeV (Borsanyi et al. 2020); trace-anomaly peak (ε−3p)/T⁴ ≈ 3.5–4 near 200 MeV
Agreement
Consistent with the lattice curve end to end — both endpoints match known dof counts, trace-anomaly peak in the same place; two independent lattice collaborations agree within a few MeV in T_c and a few percent in p/T⁴ (consistency check — shared inputs)

Check the source →

05 · Entropy Conservation and Scale-Factor Consistency Test

Agrees

The gap between the 2.7255 K photon sky and the 1.945 K neutrino sea is one sharp number — the cube root of eleven over four — and both roads balance the books to exactly 1.401.

Number we’re testing
The photon-to-neutrino temperature ratio T_γ/T_ν
Standard cosmology
T_γ/T_ν = (11/4)^(1/3) ≈ 1.401 (comoving entropy conserved across e⁺e⁻ annihilation; g*s 5.5 → 2)
This framework (granularity)
Reads the same moment as photons and neutrinos becoming two separately recordable thermal ledgers — and uses the same entropy-conservation bookkeeping to get the same 1.401
Measured
Today's photon temperature T₀ = 2.7255 ± 0.0006 K (Fixsen 2009) fixes the neutrino background at T_ν,0 ≈ 1.945 K; consistent with N_eff ≈ 3.044
Agreement
'Both roads land on the identical ratio, by the same conservation law ... the agreement is exact' — (11/4)^(1/3) = 1.40102… (consistency check — shared inputs)

Check the source →

16 · Electroweak Symmetry Breaking Mass Turn-On Test

Agrees

Rewind to a hundred-billionth of a second and nothing has weight — then the cosmos cools through one specific temperature and mass switches on across the particle world.

Number we’re testing
Electroweak crossover temperature T_c — the temperature at which W, Z and fermion masses turn on
Standard cosmology
T_c ≈ 159–163 GeV — finite-temperature Standard Model on the lattice (D'Onofrio & Rummukainen 2016)
This framework (granularity)
T_c ≈ 159–163 GeV — mass becomes a real, stable, recordable distinction only after the Higgs settles into its broken value: the same crossover
Measured
Zero-temperature anchors both roads land on: v_0 ≈ 246.22 GeV, m_W = 80.3692 ± 0.0133 GeV, m_Z = 91.1876 ± 0.0021 GeV, m_t = 172.57 ± 0.29 GeV (PDG 2024)
Agreement
Same observable, same value — two roads meet at a single temperature (T_c ≈ 159–163 GeV); timing/order-of-magnitude check, crossover width ΔT/T_c ~ few percent (consistency check — shared inputs)

Check the source →

32 · CMB Polarization and Reionization Optical-Depth Test

Agrees

Same number, arrived at from two directions — value meets value, inside the error bar.

Number we’re testing
Reionization optical depth τ — the fraction of CMB light re-scattered after the first stars switched the cosmic fog back on
Standard cosmology & measured
Standard cosmology here is the measurement itself — Ordinary hot-Big-Bang reionization: first luminous sources imprint a large-scale polarization bump whose height sets τ ≈ 0.054 · τ = 0.0544 ± 0.0073 (Planck 2018, TT,TE,EE+lowE+lensing) — a real detection of the low-ℓ E-mode reionization bump at ℓ ≲ 10, not an upper limit
This framework (granularity)
The independent read: τ = 0.0544, with reionization centered near z_re ≈ 7.7 — the framework inherits the same reionization physics and reproduces the measured depth as a self-consistency check
Agreement
τ = 0.0544 both ways, landing 'dead-center' inside the ±0.0073 error bar; z_re ≈ 7.7 (68% range ~6.5–8.9) consistent with quasar end-of-reionization z ≈ 6–7 (consistency check — shared inputs)

Check the source →

34 · Primordial Power Spectrum to Large-Scale Structure Test

Agrees

Two descriptions of the same physics, meeting at one number — a turnover wavelength of about 590 million parsecs.

Number we’re testing
Turnover scale of the matter power spectrum, k_eq ≈ 0.0107 Mpc⁻¹ (a ripple ≈590 Mpc across), and the broad shape of P(k)
Standard cosmology
Planck's measured primordial spectrum (n_s = 0.9649 ± 0.0042, A_s ≈ 2.10×10⁻⁹) run through the ordinary radiation-then-matter transfer function with Ω_m h² = 0.1430 ± 0.0011 → turnover at k_eq ≈ 0.0107 Mpc⁻¹
This framework (granularity)
The same early density distinctions, read out later as an independent record — the same turnover, k_eq ≈ 0.0107 Mpc⁻¹, and the same broad shape
Measured
Galaxy and weak-lensing surveys (DES Y3, KiDS-1000, eBOSS DR16) map exactly this shape and turnover; amplitude comparison: S₈(Planck) = 0.832 ± 0.013 vs DES Y3 0.776 ± 0.017 and KiDS-1000 0.759 +0.024/−0.021
Agreement
Both roads land on the identical turnover k_eq ≈ 0.0107 Mpc⁻¹ and the same broad shape within reported uncertainties; the S₈ amplitude is an open inherited ΛCDM tension (~2.4σ vs DES Y3, ~2.6σ vs KiDS-1000), left honestly unresolved and routed to Test 36 (consistency check — shared inputs)

Check the source →

38 · Early Galaxies and First-Star Formation Test

Agrees

Planck pinned the reionization midpoint at redshift 7.68, give or take 0.79 — and the framework's cooling timeline lands first light inside those error bars.

Number we’re testing
Onset of the first galaxies — the atomic-cooling halo mass threshold — and the timing of cosmic reionization (midpoint z_re)
Standard cosmology
First galaxies switch on once halos can cool: M_vir ≈ 5×10⁷ M_sun at z = 15, rising to 1.7×10⁸ M_sun by z = 6 (Barkana & Loeb 2001 scaling); enough light up to reionize with a midpoint near z ≈ 7.7
This framework (granularity)
The atomic-cooling threshold read as the first publicly, remotely recordable signature — reading the timing off that threshold places first light at cosmic dawn, early enough to hit the observed reionization midpoint (no independent number)
Measured
Reionization midpoint z_re = 7.68 ± 0.79 (from Planck 2018 τ = 0.0544 +0.0070/−0.0081, tanh model) — a real detection with error bars; galaxies spectroscopically confirmed out to z = 14.32 (Carniani et al. 2024)
Agreement
Framework timing reproduces the measured midpoint inside the error bars ('number meets number'); cooling-threshold halo masses sit comfortably below the JWST-confirmed galaxies out to z = 14.32 — no timing violation (consistency check — shared inputs)

Check the source →

30 · CMB Blackbody and Spectral Distortion Test

Agrees

The oldest light in the sky is the most perfect blackbody ever measured — a receipt saying nothing since the beginning has managed to smudge the record.

Number we’re testing
Size of any μ- or y-type spectral distortion in the CMB — the fingerprint of energy dumped into the photons after they thermalized
Standard cosmology
A near-perfect blackbody at T = 2.72548 ± 0.00057 K, with only a tiny Silk-damping distortion floor: μ ~ 2 × 10^-8, y ~ 4 × 10^-9 (Chluba & Sunyaev 2012) — about four orders of magnitude below detectability
This framework (granularity)
The framework adds no new energy to the photon bath, so it expects no distortion beyond the standard floor — the blackbody is the clean fossil of a thermalization that finished successfully
Measured
COBE/FIRAS sees no distortion, only ceilings: |μ| < 9 × 10^-5 and |y| < 1.5 × 10^-5 (95% CL, Fixsen et al. 1996)
Agreement
A shared prediction of silence: predicted floor sits ~4 orders of magnitude under the FIRAS bounds (μ_pred/μ_max ≈ 2 × 10^-4, y_pred/y_max ≈ 3 × 10^-4) — both roads far below both ceilings, no tension (consistency check — shared inputs)

Check the source →

44 · Decoherence versus Hubble Time Cosmic Record Test

Agrees

Two roads, one answer — about 100 to 1 — at the one moment that made the CMB permanent.

Number we’re testing
How fast a quantum pattern loses its quantumness versus how fast the universe expands — the ratio Γ/H at the moment the CMB record is imprinted
Standard cosmology
Compton coupling to the photon-baryon fluid runs about 100× faster than expansion right through recombination — Γ_T/H ≈ 1.3×10² (range ≈114–146 across z = 1000–1200); inflationary squeezing-driven decoherence exceeds H by factors ~e^{2N}
This framework (granularity)
A record locks in exactly when decoherence beats expansion — the criterion Γ/H ≫ 1; applying the same rule to the same citable physics lands on the same Γ/H ≈ 10²
Measured
Decoherence is textbook, laboratory-tested physics; the recombination redshift z* = 1089.92 ± 0.25 is fixed by the CMB itself (inputs: H₀ = 67.36 ± 0.54 km/s/Mpc, Ω_b h² = 0.02237 ± 0.00015, σ_T = 6.6524587×10⁻²⁵ cm²)
Agreement
Same ratio, both roads — roughly 100 to 1 (Γ_T/H ≈ 1.3×10²) right where the sky's pattern freezes; the criterion Γ/H ≫ 1 is met by two orders of magnitude (consistency check — shared inputs)

Check the source →

01 · Age–Temperature Expansion Consistency Test

Agrees

Run the tape of the early universe forward and it keeps hitting the same marks — quarks confine at ~20 microseconds, neutrinos go free at ~1 second, nuclei at ~3 minutes, atoms at 380,000 years.

Number we’re testing
The cosmic age at each temperature milestone — the clock reading when the universe passes a given heat
Standard cosmology & measured
Standard cosmology here is the measurement itself — Friedmann expansion sets each age: neutrino decoupling at T ≈ 1 MeV (t ≈ 1 s); recombination at z* = 1089.92 (t* ≈ 3.8×10⁵ yr); BBN window t ≈ 180–1000 s · T₀ = 2.7255 ± 0.0006 K (Fixsen 2009/COBE-FIRAS); z* = 1089.92 ± 0.25 (Planck 2018 VI)
This framework (granularity)
The independent read: Each milestone lands on the same clock — same ages, same expansion physics; computed t(1 MeV) ≈ 0.73 s, t(155 MeV QCD) ≈ 2.2–2.6×10⁻⁵ s, t(0.08 MeV BBN) ≈ 2.0×10² s
Agreement
"Same timeline, same numbers" — every claimed event quantitatively consistent with the standard age–temperature relation; passed by agreement via shared physics, not two independent numbers (consistency check — shared inputs)

Check the source →

13 · GUT Symmetry-Breaking Relic Test

Agrees

Sweeping the monopoles away needs only ~16 e-folds of inflation; smoothing the sky independently demands 50–60 — the numbers don't fight, they nest.

Number we’re testing
Whether the flood of magnetic monopoles made at GUT symmetry breaking is thinned out below the density of ordinary matter (e-folds needed vs available)
Standard cosmology
GUT breaking overproduces monopoles by ~20 orders of magnitude (Ω_mono h² ~ 6×10¹⁹ undiluted); inflation's 50–60 e-folds stretch them away to nothing
This framework (granularity)
Same GUT-scale moment, riding the same inflationary stretch; the ~16 e-folds it independently calls for sit well inside that stretch — same conclusion, no monopole overrun
Measured
Zero monopoles ever detected (flux ≲ 10⁻¹⁶ cm⁻²s⁻¹sr⁻¹, MACRO); observed matter density Ω_m h² = 0.1430 ± 0.0011 (Planck 2018)
Agreement
N ≳ 16 e-folds needed vs 50–60 independently required — 'those numbers don't fight, they nest'; 'both roads erase the monopoles below detectability' (consistency check — shared inputs)

Check the source →

39 · Primordial Black Hole Constraint Test

Agrees

It is agreement on an absence — nobody has ever caught a primordial black hole, and neither picture needs one.

Number we’re testing
The fraction of dark matter locked in primordial black holes, f_PBH(M), across ~17 factors-of-ten in mass (evaporating micro-holes to ~30-solar-mass objects)
Standard cosmology
No such population has ever been found; every survey only sets a ceiling — primordial black holes are allowed, but not required, to exist
This framework (granularity)
Introduces no mechanism that would boost small-scale ripples into black holes — so it, too, expects no primordial-black-hole population and nothing that would breach the observed ceilings
Measured
Zero primordial black holes detected; f_PBH held below 1 across the whole mass range (microlensing, CMB accretion, LIGO/Virgo/KAGRA merger rates, gamma-ray limits; Carr et al. 2021 compilation)
Agreement
Both roads answer the same question with the same word — none: no population required, no ceiling breached; PBH formation would need P_ζ ~ 10⁻² at small scales, ~7 orders of magnitude above the measured A_s = 2.1×10⁻⁹, and neither picture supplies that boost (consistency check — shared inputs)

Check the source →

43 · Complexity Capacity Bound per Epoch Test

Agrees

The bill clears every time, with fifteen to thirty-four orders of magnitude to spare.

Number we’re testing
The information actually required by each epoch's cosmic records versus the hard storage ceiling (Bekenstein and holographic bounds) for the same volume
Standard cosmology
Two established ceilings: the Bekenstein bound for the CMB photon gas in the Hubble volume, S_Bek ≈ 1.2×10¹¹⁸ k_B, and the holographic bound for today's Hubble volume, S_hor ≈ 2×10¹²² k_B
This framework (granularity)
A consistency reading, not a new number: the records relied on are highly compressed summaries — a few numbers per epoch (S_γ ~ 10⁸⁸, S_ν ~ 10⁸⁹ k_B) — so the information required sits far below the ceiling
Measured
Entropies read off measured parameters (H₀ = 67.4 ± 0.5 km/s/Mpc, T₀ = 2.7255 ± 0.0006 K; Egan & Lineweaver 2010 budget: S_γ ~ 10⁸⁸, S_ν ~ 10⁸⁹, S_SMBH ~ 10¹⁰³ k_B); required information falls short of capacity by 15 to 34 orders of magnitude
Agreement
Every record clears the capacity bounds by 15–34 orders of magnitude at every epoch checked — the books never overflow (consistency check — shared inputs)

Check the source →

02 · Interaction Rate versus Hubble Rate Decoupling Test

Agrees

About one second in the neutrinos let go — the framework's recordability crossing lands between 1 and 2 MeV, bracketing the textbook 0.8–1 MeV by the very same Γ = H rule.

Number we’re testing
Neutrino decoupling temperature T_d — the Γ = H crossing where the weak-interaction rate falls below the expansion rate
Standard cosmology
T_d ≈ 0.8–1 MeV for ν_e (Γ_ν ~ G_F²T⁵ vs H ~ T²; Dodelson & Schmidt 2020, full matrix elements)
This framework (granularity)
The Γ = H crossing for when a neutrino first becomes separately recordable lands between 1 and 2 MeV (order-of-magnitude evaluation), bracketing 0.8–1 MeV
Measured
Downstream relic imprint: T_ν,0 ≈ 1.95 K and N_eff = 3.044, 'both confirmed by the CMB and light-element abundances' (Bennett et al. 2020)
Agreement
Computed crossing (1–2 MeV) brackets the literature precision value (0.8–1 MeV) — 'a bracket, not yet a percent-level fit' (order-unity prefactor dropped) (consistency check — shared inputs)

Check the source →

04 · Causal Horizon and Record Propagation Test

Agrees

Ten thousand sky patches that never talked wear the same temperature to five decimal places — both roads draw the same 1–2° coin on the sky and reach for the same escape hatch: inflation.

Number we’re testing
The angular size of a causally-connected patch on the last-scattering sky (degrees)
Standard cosmology
Without inflation, causal patches at last scattering subtend only ~1–2° (particle horizon ~0.2–0.3 Mpc physical)
This framework (granularity)
The record-affordability rule reaches the identical ~1–2° patch — 'not an independently forced number; the same causal structure, read a second way'
Measured
CMB uniform to ~1 part in 10⁵ across ~10⁴–10⁵ never-connected patches; θ* ≈ 0.596° (100θ* = 1.04109 ± 0.00030, Planck 2018); r_s = 144.4 ± 0.3 Mpc
Agreement
'Same patch size, same diagnosis' — 'our number and the textbook number are the same because we are both describing the same physics'; both point to inflation as the named channel (consistency check — shared inputs)

Check the source →

11 · Reheating Entropy Dilution / Relic Reset Test

Agrees

Sixty e-folds crush the leftover monsters by ~78 orders of magnitude while the 6.12×10⁻¹⁰ baryon surplus — written after the great dilution, not before — survives to build everything we see.

Number we’re testing
The entropy-dilution ledger — how thoroughly expansion wipes out unwanted relics (~78 orders of magnitude) while the matter–antimatter surplus η_B survives
Standard cosmology
Dilution e^(−3N); for N ≳ 60 e-folds, e^(−180) ≈ roughly 78 orders of magnitude in number density (vs ~14–20 orders needed for GUT monopoles); η_B survives because it is made after reheating
This framework (granularity)
The same ledger, reread: reheating as a 'reset' erasing sub-resolution leftovers from the recordable record while η_B stays — 'it reproduces the standard books; it derives no new number of its own'
Measured
η_B = (6.12 ± 0.04)×10⁻¹⁰ survives (Planck 2018 via Ω_b h² = 0.02237); no primordial monopole or other diluted relic has ever been detected — exactly the silence the suppression predicts
Agreement
'Same dilution ledger, consistently reread; the framework inherits the number, it does not re-derive it' — same suppression, same survivor, null relic record consistent (consistency check — shared inputs)

Check the source →

14 · Monopole Production and Dilution Test

Agrees

Seventeen orders of magnitude of monopole overproduction, diluted to Ω_M h² ≈ 2×10⁻⁶¹ — indistinguishable from none, which is exactly the silence every detector reports.

Number we’re testing
The leftover density of magnetic monopoles after inflation, Ω_M h²
Standard cosmology
Kibble mechanism gives Ω_M h² ~ 3×10¹⁷ undiluted (~17 orders of magnitude overclosure); ~60 e-folds thin by e^(−180) ≈ 6.7×10⁻⁷⁹ → Ω_M h² ≈ 2×10⁻⁶¹
This framework (granularity)
Runs the identical defect-formation estimate and the identical ~60-e-fold dilution → the same Ω_M h² ≈ 2×10⁻⁶¹ — 'not a fresh derivation; the same calculation, read as records too sparse to ever register'
Measured
No monopole has ever been observed; flux ceilings only: MACRO ≲ 1.4×10⁻¹⁶, Parker ≲ 10⁻¹⁵, IceCube ≲ 10⁻¹⁸ cm⁻²s⁻¹sr⁻¹ — 2×10⁻⁶¹ sits far beneath every one
Agreement
'Both roads land on the same vanishing density, by the same dilution ledger' — and 'agree by construction, because they are doing the same arithmetic from two vantage points' (consistency check — shared inputs)

Check the source →

15 · Cosmic Strings and Topological Defects Test

Agrees

A GUT-scale string network at Gμ ~ 2.7×10⁻⁶ would be roughly twenty times over Planck's ceiling — too heavy to hide, which is exactly why the sky shows none.

Number we’re testing
Whether a stable network of GUT-scale cosmic strings could have survived to today — 'the yes/no conclusion, not the tension number'
Standard cosmology
A naive GUT-scale string network would be too heavy to have escaped detection, so some non-defect-forming or diluting mechanism must apply
This framework (granularity)
From a generic GUT scale (η ~ 2×10¹⁶ GeV) the string tension works out to Gμ ~ 2.7×10⁻⁶ — above the bounds, so a naive stable network is likewise ruled out
Measured
No cosmic strings seen. Bounds: Gμ < 1.3×10⁻⁷ (Planck 2018, 95% CL); Gμ ≲ 10⁻¹⁰ (NANOGrav 15-yr, loop-model dependent)
Agreement
Naive estimate exceeds the Planck bound by ~21× and NANOGrav by ~2.7×10⁴× — 'the over-the-limit number is the reason for the agreement'; agreement 'on the conclusion (no stable strings survive), not on a matched number' (consistency check — shared inputs)

Check the source →

42 · Recordability / Distinguishability Ledger Test

Agrees

The framework passes its own honesty test — nothing in the suite is quietly treated as a discovered object when it is really a theoretical placeholder.

Number we’re testing
Not a number — an honesty ledger: for every object the other 43 tests lean on, does it count as recordable (coupled, encoded, stabilized, retrievable) at its claimed epoch?
Standard cosmology
The same sorting the field already uses (directly observed records vs indirect relics vs purely theoretical variables), anchored to the Particle Data Group, the Planck Collaboration, and standard BBN reviews
This framework (granularity)
Sorts every object by one rule — coupling, encoding, stability, retrieval — into a class from 'directly observed public record' down to 'theoretical placeholder'; produces no new number
Measured
Paradigm case: the CMB at z ≈ 1089 classed as a directly observed public record on both sides; dark-matter gravity recordable but the WIMP identity not; ledger uses N_eff = 2.99 ± 0.17 (Planck 2020), r < 0.036 (BICEP/Keck 2021)
Agreement
Same classifications, every epoch — the two ledgers agree everywhere checked; nothing theoretical is smuggled in as observed (consistency check — shared inputs)

Check the source →

The standing bets — both routes agree, the measurement is pending (6)

These six are not weaknesses — they are the suite’s forward edge. On each one the two routes already agree; the experiment simply hasn’t voted yet. The sharpest of them is a genuine framework-native prediction: the tensor band r ∈ [3.5, 36]×10⁻³, which LiteBIRD (~2030) can test outright.

08 · Inflation Tensor Bound Test

Routes agree — measurement pending

Standing behind r ∈ [3.5, 36]×10⁻³ is real; having the sky agree is a moment that hasn't arrived — LiteBIRD around 2030 gets the deciding vote.

Number we’re testing
The tensor-to-scalar ratio r — strength of primordial gravitational waves relative to density ripples
Standard cosmology
Single-field slow-roll expects a small but nonzero r; the exact value depends on the potential and is not fixed; most well-motivated models land below current sensitivity
This framework (granularity)
r ∈ [3.5, 36]×10⁻³ (union of four normalization branches; operative branch [3.5, 10]×10⁻³) — a prediction awaiting a measurement, not a confirmed match
Measured
r < 0.036 (95% CL, pivot k = 0.05 Mpc⁻¹) — BICEP/Keck 2021 (BK18) + Planck + BAO; an upper bound only, no detection, r = 0 not excluded
Agreement
Routes agree — measurement pending (LiteBIRD, ~2030). Every branch of the predicted band r ∈ [3.5, 36]×10⁻³ already sits strictly under today’s ceiling, r < 0.036; LiteBIRD tests the full band outright. (the band is the framework’s own number; slope not yet geometrically forced)

Check the source →

37 · 21-cm Dark Ages / Cosmic Dawn Granularity Test

Routes agree — measurement pending

The epoch matches, the depth is undecided, and we say so plainly.

Number we’re testing
The cosmic-dawn 21-cm signal — the redshift where the first stars light up the neutral hydrogen fog (z ≈ 17.2, 78.1 MHz) and the depth of the absorption dip they carve into it
Standard cosmology
Dip centered at z ≈ 17.2 (78.1 MHz); thermal ceiling limits its depth to about −230 mK
This framework (granularity)
Same epoch, z ≈ 17.2 — reproduces the standard adiabatic-cooling physics; contributes no independent number for the dip's depth
Measured
EDGES (2018) reported a dip at 78.1 MHz but −500 +88/−50 mK deep — more than twice the ceiling; SARAS 3 (2022) later found no such feature
Agreement
Routes agree on the epoch — z ≈ 17.2 (78.1 MHz) from both directions; the trough depth is the pending measurement: EDGES reports −500 mK against the −230 mK standard ceiling, and SARAS 3 finds no feature — unresolved, said plainly. (consistency check — shared inputs)

Check the source →

12 · Primordial Isocurvature Constraint Test

Routes agree — measurement pending

One clock ticking through inflation means no isocurvature stragglers — and Planck's β_iso < 0.038 line in the sand found exactly nothing on the far side.

Number we’re testing
The primordial isocurvature fraction β_iso — the share of early-universe density fluctuations that is not the common adiabatic mode
Standard cosmology
Single-field slow-roll inflation expects essentially purely adiabatic fluctuations, β_iso ≈ 0
This framework (granularity)
A single geometric modulus carries the same one-clock condition → negligible isocurvature (β_iso ≈ 0 in the single-field limit) — 'inherited from the single-field setup, not independently derived'
Measured
β_iso < 0.038 (95% CL, Planck 2018 X, uncorrelated CDM isocurvature) — a one-sided upper bound; no nonzero value detected, fully consistent with exactly zero
Agreement
Routes agree — β_iso ≈ 0 from both directions, well inside Planck’s β_iso < 0.038; measurement pending — a one-sided bound consistent with exactly zero leaves the decisive detection still ahead. (consistency check — shared inputs)

Check the source →

27 · Axion / ALP Misalignment and Defect Test

Routes agree — measurement pending

The standard axion recipe works cleanly — but only after you write down, by hand, how far off-center the field started.

Number we’re testing
The cold-dark-matter density an axion population would have to supply (Ω_c h²) via misalignment + string/domain-wall defects, plus the isocurvature bound
Standard cosmology
Misalignment-and-defect production is consistent with the data once θ_i is put in by hand; post-inflationary DM-saturating window f_a ≈ (0.5–4) × 10^10 GeV (m_a ≈ 40–500 μeV); pre-inflationary branch requires H_inf ≲ 10^7 GeV (f_a/10^11 GeV)^0.4
This framework (granularity)
No specific axion-like field, decay constant, or starting angle has been built yet — no distinctive number of its own; reads the misalignment angle θ_i as the same empirical input everyone uses
Measured
Ω_c h² = 0.120 ± 0.001 (Planck 2018); isocurvature β_iso < 0.038 (95% CL, Planck 2018); r < 0.036 (95% CL, BICEP/Keck 2021)
Agreement
Routes agree — the whole axion viability window is open and consistent on both roads; measurement pending — no axion detected yet, and the misalignment angle θᵢ remains an empirical input. (consistency check — shared inputs)

Check the source →

09 · Inflation Non-Gaussianity Test

Routes agree — measurement pending

Planck found the primordial ripples quiet to −0.9 ± 5.1; the framework matches that silence by class membership, not by a number of its own — an open question kept in plain sight.

Number we’re testing
f_NL^local — primordial non-Gaussianity, how far the primordial ripples depart from a perfect bell curve
Standard cosmology
f_NL^local ≈ 0.01–0.02 — single-field consistency relation f_NL = (5/12)(1 − n_s) ≈ 0.0146 (Maldacena)
This framework (granularity)
No independent number is computed — the surviving inflation candidate is single-field, so it inherits the same near-zero expectation by class, not by its own calculation
Measured
f_NL^local = −0.9 ± 5.1 (Planck 2018 IX, SMICA T+E, 68% CL); also f_NL^equil = −26 ± 47, f_NL^ortho = −38 ± 24 — a wide bound consistent with zero
Agreement
Routes agree — the single-field class prediction (≈0.015) sits ≈0.18σ from Planck’s −0.9 ± 5.1; measurement pending — today’s precision is ~two orders of magnitude too coarse to discriminate. (consistency check — shared inputs)

Check the source →

10 · Reheating Temperature and Thermalization Test

Routes agree — measurement pending

Between the 4 MeV nucleosynthesis floor and the 10¹⁶ GeV gravitational-wave ceiling lies a corridor nineteen orders of magnitude wide — both pictures fit inside it, and neither pins the number.

Number we’re testing
The reheating temperature T_reh — how hot the universe becomes when inflation's energy converts into an ordinary particle bath
Standard cosmology & measured
Standard cosmology here is the measurement itself — Not one value but an allowed window: above ~4 MeV (BBN floor; ~4.4 MeV at 95% CL in refined treatments) and below ~10¹⁶ GeV (tensor-ratio ceiling from r < 0.036) · No direct measurement of T_reh exists; only the two edges of the window are pinned by data (BBN/N_eff below — N_eff = 2.99 ± 0.17 Planck 2018 — gravitational-wave limits above)
This framework (granularity)
The independent read: The moment unified energy first becomes a recordable particle bath — lands inside the same window, but the framework supplies no decay mechanism, so it names no specific temperature of its own
Agreement
Routes agree on the same window — 4 MeV to 10¹⁶ GeV, both roads inside it; measurement pending — no direct measurement of T_reh exists, only the window’s two edges are pinned by data. (consistency check — shared inputs)

Check the source →

Open problems stay open (8) — for all of cosmology

Eight questions have no settled answer from anyone — not from standard cosmology, not from this framework. We reproduce the standard calculation, mark the question open, and keep it on the board in full view. Showing these eight is what makes the thirty worth believing.

36 · Sigma8 / S8 Growth-Tension Gate

Open (for all of cosmology)

Two numbers, a real gap between them, and nobody — us included — yet knows whether the fix is new physics or an unmodeled systematic.

Number we’re testing
Late-time clustering amplitude S₈ ≡ σ₈√(Ω_m/0.3) — how clumpy today's universe is
Standard cosmology & measured
Standard cosmology here is the measurement itself — S₈ = 0.832 ± 0.013 (Planck 2018 CMB-inferred, from σ₈ = 0.8111 ± 0.0060 and Ω_m = 0.3153 ± 0.0073) · Weak lensing measures lower: KiDS-1000 S₈ = 0.759 +0.024/−0.021 (Asgari et al. 2021); DES Y3 3×2pt S₈ = 0.776 ± 0.017 (Abbott et al. 2022)
This framework (granularity)
The independent read: Same growth equations, no separate late-time mechanism of its own — the propagated distinction predicts the identical S₈ ≈ 0.83 from the CMB epoch
Agreement
A real ≈2.6–2.8σ mismatch (page's own recomputation): Planck 0.8315 vs KiDS-1000 0.759 → 2.8σ; vs DES Y3 0.776 → 2.6σ — consistent with the 2–3σ range in the tension-review literature (Abdalla et al. 2022)

Check the source →

25 · Dark Matter Thermal Freeze-Out Test

Open (for all of cosmology)

Both roads hit the same amount of dark matter, 0.12, almost exactly — and neither can tell you what the particle is.

Number we’re testing
Dark-matter relic density Ω_DM h² from thermal (WIMP) freeze-out — plus whether any candidate survives detection bounds
Standard cosmology
Ordinary freeze-out reproduces Ω_DM h² = 0.12 for ⟨σv⟩ ≈ 2.5 × 10^-26 cm³ s^-1 (canonical WIMP range) at mass ~10–1000 GeV
This framework (granularity)
Runs the same freeze-out bookkeeping and lands on the same target amount — it reproduces Ω_DM h² = 0.12, it does not derive it and names no particle
Measured
Ω_DM h² = 0.1200 ± 0.0012 (Planck 2018); LZ 2024 excludes σ_SI ≳ 2.2 × 10^-48 cm² at m_χ ≈ 36 GeV; Fermi-LAT excludes canonical ⟨σv⟩ below ~100 GeV in b-bbar
Agreement
Back-of-envelope freeze-out reproduces the canonical WIMP-miracle cross-section and the measured density; but the simplest weak-scale candidates are squeezed by direct detection and no particle is named — identity open for everyone

Check the source →

24 · Lithium-7 Tension Handling Test

Open (for all of cosmology)

The very same calculation that nails deuterium and helium insists the cosmos should have made about three times more lithium than we can find.

Number we’re testing
Primordial lithium-7 abundance — ⁷Li/H (A(Li)), predicted vs the Spite-plateau reading in old halo stars
Standard cosmology
SBBN at the measured baryon density predicts A(Li) ≈ 2.6–2.7, i.e. ⁷Li/H ≈ (5.1–5.6) × 10^-10 — roughly three times too much
This framework (granularity)
Uses the identical nuclear-reaction network, so it reproduces the same over-prediction — and brings no separate leverage on the lithium channel; reports the tension rather than papering over it
Measured
Spite plateau in metal-poor halo stars: A(Li) ≈ 2.1–2.2, i.e. ⁷Li/H ≈ (1.5–1.6) × 10^-10 (Sbordone et al. 2010)
Agreement
Over-prediction by a factor ≈3.4 (range ~3–5), ΔA(Li) ≈ 0.4–0.5 dex — a many-σ discrepancy given plateau scatter of ~0.05–0.1 dex; unresolved for the whole field

Check the source →

17 · Electroweak Sphaleron and Baryogenesis Test

Open (for all of cosmology)

Something tipped the scales by about one part in a billion — that tiny surplus is everything solid you have ever touched, and nobody can derive it from first principles.

Number we’re testing
Origin of the baryon asymmetry η_B — whether known electroweak-era physics can generate the matter–antimatter imbalance
Standard cosmology
Can't close it: a 125.20 GeV Higgs makes the transition a smooth crossover (Sakharov condition 3 fails); CKM CP violation (Jarlskog J ≈ 3×10^-5) falls ~10 orders of magnitude short
This framework (granularity)
Doesn't close it either — reproduces the settled timing (T_c ≈ 159 ± 1 GeV crossover, sphalerons freezing out at ~130–140 GeV) but supplies no new source for the asymmetry
Measured
η_B = (6.10 ± 0.04) × 10^-10 (Planck 2018 via Ω_b h² = 0.02237 ± 0.00015); equivalently Y_ΔB ≈ 8.7 × 10^-11
Agreement
Sphaleron bookkeeping agrees quantitatively (Γ_sph ≫ H above T_c); the CP-violating out-of-equilibrium source of η_B is missing — a ~10-orders-of-magnitude shortfall shared by the whole field

Check the source →

18 · CP-Violation Sufficiency Test

Open (for all of cosmology)

Feed the measured CP violation through the arithmetic and the most it can produce is roughly one part in 10^20 — ten orders of magnitude shy of the surplus we actually inherited.

Number we’re testing
Whether the measured CP violation can build the observed baryon asymmetry η_B
Standard cosmology
Known CP violation (Jarlskog J ≈ 3.0 × 10^-5) can make at most η_B ≲ 10^-20 — about ten orders of magnitude too little; new physics is required
This framework (granularity)
Agrees on where and why the books must balance, but supplies no new CP source of its own — inherits the same open shortfall
Measured
η_B = (6.12 ± 0.04) × 10^-10 (Planck 2018, via Ω_b h² = 0.02237 ± 0.00015), cross-checked by BBN (η_10 = 6.09 +0.42/−0.39)
Agreement
SM-CKM ceiling ~10 orders of magnitude below the observed η_B, and the required first-order phase transition is excluded by m_H = 125.25 GeV — no mechanism in any framework closes the gap

Check the source →

26 · Dark Matter Freeze-In / Nonthermal Relic Test

Open (for all of cosmology)

Imagine a relic so shy it was never in the crowd at all — it just trickles into existence through a whisper-thin portal and quietly adds up to exactly the amount we measure.

Number we’re testing
Dark-matter relic density Ω_DM h² via freeze-in — a feeble portal coupling slowly building a never-thermalized relic
Standard cosmology
Generic freeze-in benchmarks: portal coupling y ~ 10^-11–10^-12 and mass keV–TeV reproduce Ω_DM h² = 0.120 (Hall et al. 2010; Bernal et al. 2017), with Γ_portal ≪ H(T) satisfied by many orders of magnitude
This framework (granularity)
Uses the identical freeze-in machinery — it reproduces the same 0.120, but brings no separate calculation and specifies no field, portal, coupling, or mass of its own
Measured
Ω_DM h² = 0.120 ± 0.001 (Planck 2018); warm-dark-matter free-streaming floor m_WDM ≳ 5.3 keV (Lyman-α); ΔN_eff ≲ 0.3
Agreement
Mechanism class viable and unexcluded — benchmark points satisfy relic-density, free-streaming, N_eff and BBN bounds by construction; no named candidate exists to compute an actual prediction from

Check the source →

06 · Planck Boundary / Quantum-Gravity Terminal Test

Open (for all of cosmology)

Run the clock backward and both roads walk to the same cliff edge at 5.4×10⁻⁴⁴ seconds — and past it, for anyone, in any theory: fog.

Number we’re testing
Where known physics runs out — the Planck-time wall, t_Pl
Standard cosmology & measured
Standard cosmology here is the measurement itself — t_Pl ≈ 5.391×10⁻⁴⁴ s (from ħ, G, c) — no validated theory beyond it · t_Pl = 5.391×10⁻⁴⁴ s (CODATA 2018) — 'a fixed constant, not a fit'; nearest directly probed physics is ~33 orders of magnitude later (electroweak, t ~ 10⁻¹¹–10⁻¹² s)
This framework (granularity)
The independent read: Same wall: below the Planck scale a distinction can no longer be coupled, encoded and stabilized, so the recordability rule simply stops applying (t_Pl = 5.391246×10⁻⁴⁴ s, computed the same way)
Agreement
'A shared limit, not a matched prediction — open for everyone'; agreement on where the wall sits, computed the same way by both; graded Indeterminate to avoid dressing shared ignorance as shared discovery

Check the source →

40 · Primordial Magnetic Field Test

Open (for all of cosmology)

A real window is waiting — about seven orders of magnitude wide — but no one, on either road, has a number to place inside it.

Number we’re testing
The present-day strength of any primordial magnetic field spread across intergalactic space — its comoving amplitude B₀ on Mpc scales
Standard cosmology
Not a single value — a window: roughly 3×10⁻¹⁷ G at the bottom (blazar/gamma-ray non-detections) up to 0.69 nG at the top (Planck 2018 CMB + BBN)
This framework (granularity)
A field would be a recordable fossil of a symmetry-breaking event, but no sector yet computes a coupling, so there is no amplitude of its own to check
Measured
The window itself: lower edge ≈3×10⁻¹⁷ G (blazar cascades, Acciari et al. 2023); upper edge <0.69 nG (Planck 2018, Paoletti et al. 2022) — about seven orders of magnitude wide
Agreement
The two roads do not meet at a number here, because neither road has reached one — no mechanism means no amplitude to check; nothing conflicts with any measurement

Check the source →

The cosmic tape — every test in its epoch

The same 44 tests, pinned to cosmic time — earliest first. Green = confirmed, blue = routes agree/measurement pending, amber = open for everyone.

The backbone
10⁻⁵ s → 380,000 yr
Planck wall
5.4×10⁻⁴⁴ s
Inflation & reheating
~10⁻³⁴ s
GUT relics & defects
~10⁻³⁶ s
Electroweak
~10⁻¹¹ s
QCD & baryons
~20 µs
Neutrinos & BBN
~1 s – 3 min
Dark matter & axions
~10⁻¹⁰–10⁻⁶ s
Equality → CMB
50,000–380,000 yr
Large-scale structure
to 13.8 Gyr
Dark ages → first stars
~100–400 Myr
Relics, PBHs & fields
early → today
Ledger & capacity
all epochs

What this scoreboard shows

Start with what is genuinely surprising: two independent reconstructions of the early universe agree. Standard cosmology gets its timeline by solving the hot-Big-Bang equations. This framework gets its timeline from a single principle about when the universe can hold a stable, distinguishable record. They did not have to match. They do — at the microsecond of quark confinement, at the one-second neutrino freeze-out, at the three-minute forging of the first nuclei, at the 380,000-year release of the CMB, and all the way to a 13.8-billion-year-old universe today. Agreement across two methods, thirty times over with zero conflicts, is how a picture earns being taken seriously.

The 14 Indeterminate rows are the discipline that makes those thirty believable — and they split cleanly in two. Six are standing bets: both routes already agree and the experiment hasn’t weighed in yet — LiteBIRD’s tensor band, the disputed 21-cm trough depth, the isocurvature null, the axion window, non-Gaussianity, the reheating window. Eight are open for all of cosmology: what sourced the matter–antimatter asymmetry, what dark matter is, why lithium-7 runs high, the S8 clumpiness tension, the primordial magnetic field, and the Planck wall itself. We reproduce the standard calculation, mark the question open, and show the whole denominator — no rounding up, ever.

Where we can say something new: near and before the Big Bang

Standard cosmology's timeline runs backward until it hits the Planck wall at ~10⁻⁴³ seconds and stops — the equations stop meaning anything there. That wall is where this framework has something of its own to say. In the recordability picture, the Big Bang is simply the first moment the universe can carry a recordable distinction — the origin of the timeline, not an explosion inside a pre-existing one. And “before” the Big Bang is not an earlier time at all; it is the absence of any record for time to be measured against. Asking what happened before is like asking what is north of the North Pole. That reframing doesn't need new data to be useful — it dissolves a question that the standard timeline can only leave blank, and it turns the Planck boundary from a dead end into a definite statement about where recordable history begins.

The inflation gate

The one place the framework's own prediction is genuinely on the hook: whether the geometry forces the inflaton slope. The measured spectrum agrees; the derivation is still open. LiteBIRD (~2030) can test the sharp version, r ∈ [3.5, 36]×10⁻³.

How we keep it honest

The rule that makes agreement mean something: compute our number first, then compare — never load the observed value before a prediction exists.

Back to Early & Distant Universe

The section overview: what “the beginning” means here, and the two open frontiers — inflation, and the cosmological constant and expansion history.

Show calculation · two independent methods, one number