Frequently Asked Questions

Partially. We use the King Wen ordering of hexagrams (1100 BCE) as the input sequence. However, we're not claiming mystical I Ching properties. We're treating it as a mathematical sequence that happens to have interesting properties (non-random, correlated with astronomical cycles).

Three key differences:

1. Quaternion usage: We use the full Q8 group (mathematically rigorous), McKenna extracted only sign of w (arbitrary)
2. Validation: We test against control experiments (Fu Xi, random), McKenna didn't
3. Transparency: Full source code and methodology documentation, McKenna's implementation was opaque

Real mathematics. The Q8 quaternion group is standard group theory (taught in abstract algebra courses). The algorithm is deterministic and verifiable. The statistical correlations are testable and falsifiable. Whether the correlations are meaningful is an open scientific question.

Q8 is the quaternion group consisting of 8 elements: {1, -1, i, -i, j, -j, k, -k}. It's the smallest non-abelian group, meaning multiplication order matters (i×j ≠ j×i). This rich mathematical structure makes it ideal for the transformation.

Quaternions are 4-dimensional numbers that naturally represent rotations in 3D space. They're used in graphics, robotics, and physics. The I Ching has 4 components per hexagram (2 trigrams, each with 3 lines), which maps naturally to quaternion structure.

The norm of a quaternion q = w + xi + yj + zk is calculated as: ||q|| = √(w² + x² + y² + z²). This captures all four components in a single scalar value.

Each of the 64 King Wen hexagrams is transformed through 6 Q8 operations, producing 64 × 6 = 384 quaternion norm values. These become the "sacred numbers" that form the timewave base sequence.

The timewave is built by summing values across 3 fractal levels (64 days, 384 days, 4096 days). Each level contributes to the final novelty value at any given date, creating self-similar patterns across scales.

No. The timewave shows abstract "novelty" values. It correlates with historical events statistically (r=0.42), but does not predict specific outcomes. A peak tells you "this period tends to be eventful" not "X will happen."

McKenna's original zero point was Dec 21, 2012. Timewave2 recalculated this using rigorous mathematics. The new zero point (where timewave → ∞) is November 17, 2012. But this is a mathematical singularity, not a prophecy. The timewave calculation is valid before and after this date.

Unknown. Historical correlation (1900-2024) is r=0.42 (moderate). Whether this persists into the future is untested. Predictions are registered for falsification testing. We'll know accuracy as time passes. If <50% by 2050, hypothesis is falsified.

A peak indicates higher "novelty" - periods of rapid change, crisis, innovation, or major transitions. Statistically, peaks tend to correlate with high-novelty historical events, but not deterministically.

A trough indicates lower novelty - periods of stability, consolidation, or equilibrium. Not boring, just different energy. These are necessary counterbalances to peaks.

Check all three levels: Level 2 (long-term trend), Level 1 (medium context), Level 0 (short-term detail). A date might be a small peak locally but within a long-term trough, or vice versa.

Uncertainty grows linearly from ±10 (near term) to ±50 (2050). This reflects decreasing confidence as we project further into the future.

No. This is a research tool for exploring mathematical patterns in time, not a decision-making tool. Do not make financial, medical, safety, or life decisions based on timewave predictions.

Yes! Use the citation generator for proper attribution. The data is open access via the API. Please cite appropriately if you use the work.

GitHub issues: https://github.com/timewave2/timewave2/issues

Yes! The project is open source. Contributions welcome for code, documentation, testing, or dataset improvements. See the GitHub repository for contribution guidelines.

The web interface is a Progressive Web App (PWA) that works offline and can be installed on mobile devices. No separate native app currently.

The complete timewave (1900-2050) is 54,787 data points. At ~16 bytes per point, that's ~876 KB. We cache this for offline use (PWA). You can limit date ranges via API to reduce data transfer.

Yes! The backend is open source C code. See deployment documentation in the repository.

Python: Yes (golden reference implementation). R/Julia: Not yet. Contributions welcome!

100 requests per minute for unauthenticated users. Contact us for higher limits or API keys for research use.

Yes! CSV, JSON, Excel, and RDF formats available via the API. See API documentation.

Modern browsers (Chrome, Firefox, Safari, Edge) from the last 2 years. The 3D visualizer requires WebGL support.

Yes. Progress tracking and preferences are stored locally in your browser (localStorage). No account required, no data sent to servers except API requests.

Yes! Once loaded, the PWA caches all assets and data for offline use. You need to load it once while online.