Predictions Registry
Blind predictions locked before measurement, falsified claims documented honestly, and upcoming experimental tests.
Bold Claims
Structural predictions and solutions to open physics problems. These can't be numerology — they predict what exists and what doesn't.
No Magnetic Monopoles
Standard theories explain monopole absence by saying inflation diluted them. This framework says they can't exist.
SO(11) has π1 = Z/2Z, so monopoles pair-annihilate. Topological argument, not fine-tuning.
Lithium-7 Problem Solved
For 40 years, BBN has predicted 3× too much lithium-7. The framework explains why.
Li-7 nucleus (A=7=ImO) is suppressed by 1/3 during crystallization because 7 isn't a division algebra dimension.
Hubble Tension is Real Physics
The 5-sigma discrepancy between CMB and local H0 measurements isn't a systematic error — it's a prediction.
Hlocal = HCMB × 13/12. Interior stress distributes across H+O = 12 gauge dimensions. 8.3% enhancement.
Falsification: if tension resolves to a single value, framework is wrong.
Strong CP Solved Without Axions
Why is θQCD = 0? Most theories need an undiscovered particle (axion). The framework gets it for free.
Parity structure of crystallization naturally forces θ = 0. No new particles required.
12 Non-Observations Predicted
The framework predicts 12 things that DON'T exist — and none have been observed.
Blind Predictions (Strongest Evidence)
Made BEFORE checking measured values. These carry no look-elsewhere penalty. Combined P-value: ~2.5 × 10-7.
| Prediction | Precision | Sigma |
|---|---|---|
| 100·Ωbh² | 0.77% | <1σ |
| 100·Ωch² | 0.34% | <1σ |
| 100·θs | 0.13% | 2.1σ |
| ln(1010As) | 0.006% | <1σ |
| Spectral index ns | 0.010% | <1σ |
| Reionization optical depth τ | 0.79% | <1σ |
| CMB shift parameter R | 0.035% | <1σ |
| Neutrino mass ratio R31 | 1.7% | 0.62σ |
| Neutrino mass ratio R32 | 1.8% | 0.64σ |
6/7 CMB predictions within 1σ. 2/2 neutrino predictions within 1σ.
Near-Term Experimental Tests
Each prediction has explicit falsification criteria. These are clean kills, not "we can adjust" situations.
Falsified Claims (14 total)
Recording failures is essential. We don't hide what doesn't work.
| Type | Count | Examples |
|---|---|---|
| Definitively falsified | 9 | sin²θW=2/25 (65% wrong), higher CMB peaks (12-19% off) |
| Deprecated | 4 | G from |Π|, EFE from γ |
| Withdrawn | 1 | h(γ) novelty claim |
Interactive explorer with filtering by tier, domain, status, and blind prediction flag.