Experiments / V2.602
V2.602
Precision Cosmological Tests COMPLETE

V2.602 - Tension Autopsy — The V2.599 3σ Was a Methodological Artifact

V2.602: Tension Autopsy — The V2.599 3σ Was a Methodological Artifact

Motivation

V2.599 reported 3σ tension between the framework and combined BAO+CMB+SNe data. Before accepting this as a real challenge to the framework, we must ask: is the tension physics, or statistics? The compressed CMB likelihood (R, l_A) used in V2.599 constrains H₀r_d at 0.26% precision — 3× tighter than Planck’s actual marginalized H₀ constraint (0.8%). If the compressed likelihood over-constrains H₀ when Ω_m is fixed, it could inflate the tension.

Method

Fit the framework (Ω_m = 0.3123 fixed) and ΛCDM (Ω_m free) to DESI DR1 BAO under 8 different prior schemes:

SchemeCMB priorOther
ANoneBAO only
A’NoneBAO with D_M–D_H correlations
BPlanck Ω_m = 0.315 ± 0.007
CPlanck H₀ = 67.4 ± 0.5
DPlanck Ω_m + H₀Correct combined
EPlanck R, l_A (compressed)V2.599-like
FPlanck Ω_m + H₀ + Pantheon+Full proper
GPlanck R, l_A + Pantheon+V2.599 method

Results

Framework tension under each scheme

SchemeFW χ²ΛCDM χ²Δχ²Tension
A: BAO only14.514.1+0.30.6σ
A’: BAO (correlated)14.212.9+1.31.1σ
B: BAO + Planck(Ω_m)14.614.6+0.00.2σ
C: BAO + Planck(H₀)16.515.9+0.60.7σ
D: BAO + Planck(Ω_m,H₀)16.616.1+0.60.8σ
E: BAO + Planck(R,l_A)31.619.2+12.43.5σ
F: BAO + Planck + SNe18.116.9+1.21.1σ
G: BAO + R,l_A + SNe (V2.599)33.024.2+8.83.0σ

The critical comparison

Using proper marginalized priors (scheme D): 0.8σ. Using compressed likelihood (scheme E): 3.5σ.

The compressed likelihood inflates the tension by 2.7σ. The 3σ reported in V2.599 was a methodological artifact.

Framework Ω_m vs each dataset independently

DatasetΩ_m±Framework tension
Planck CMB0.3150.007−0.4σ
DESI BAO0.3030.015+0.6σ
Pantheon+ SNe0.3340.018−1.2σ
Planck combined (official)0.3110.006+0.2σ
Inverse-variance combined0.3130.004−0.2σ

The framework’s Ω_m = 0.312 is within 0.2σ of Planck’s own combined constraint and within 1.2σ of every individual dataset.

Diagnosis: Why the Compressed Likelihood Fails

The shift parameter R = √(Ω_m) · D_M(z*) / (c/H₀) is a derived quantity measured at 0.26% precision. When Ω_m is fixed (as in the framework), R becomes a constraint on H₀ alone — at 0.26% precision. This is 3× tighter than Planck’s actual marginalized H₀ constraint (0.8%).

The reason: R doesn’t account for degeneracies with Ω_b h², n_s, and τ that soften the true constraint on H₀. The compressed likelihood was designed for models near the Planck ΛCDM best-fit where these degeneracies are properly explored. For a model with fixed Ω_m, it over-constrains.

The result: the compressed likelihood forces H₀ = 66.8 km/s/Mpc, 2% below the BAO-preferred value of 68.3. This 2% shift costs 15 points of χ² in BAO — the entire source of the V2.599 tension.

With proper priors (H₀ = 67.4 ± 0.5 from Planck), H₀ = 68.0 — only 0.5% from the BAO optimum, costing just 2 in χ².

Corrected Assessment

TestV2.599 (compressed)V2.602 (proper)
BAO alone0.6σ0.6σ
BAO + CMB3.5σ0.8σ
BAO + CMB + SNe3.0σ1.1σ
Ω_m (Planck combined)0.2σ

The true tension is 1.1σ, not 3σ.

Remaining Honest Tensions

Even with the corrected analysis, these tensions exist:

  1. DESI prefers lower Ω_m (0.303 vs framework’s 0.312): 0.6σ today, could grow with DESI DR2.
  2. Pantheon+ prefers higher Ω_m (0.334 vs 0.312): 1.2σ. SNe and BAO pull in opposite directions.
  3. D_M–D_H correlations slightly increase tension (0.6σ → 1.1σ) because anti-correlated pulls at LRG1 are more anomalous than independent pulls.
  4. w₀w_aCDM hint from DESI: the framework predicts w = −1 exactly, and DESI hints at evolution. This is the existential threat regardless of Ω_m.

What This Means for the Science

The framework is not under 3σ pressure. The correct combined tension is 1.1σ with proper treatment of Planck constraints. The framework’s Ω_m = 0.312 is 0.2σ from Planck’s own combined value — essentially perfect agreement.

The V2.599 result was an important stress test, but the conclusion was overstated due to the compressed likelihood methodology. The corrected picture: the framework’s zero-parameter prediction passes all current cosmological tests at ≤1.2σ individually and 1.1σ combined.

The real threat remains the DESI evolving dark energy hint (w₀ ≈ −0.55), which challenges both the framework AND standard ΛCDM. DESI DR2+ is the decisive test.