Marsaglia's XorShift (2003) --- three shift-xor operations per step. Fast and passes most tests, but fails binary rank and linear complexity
Standard analysis sees: rich, high-entropy values; aperiodic / broadband; high-complexity (noise-like); high-dimensional / space-filling. The atlas finds no named structure, but the source is distinctively extreme on Zipf–Mandelbrot (8-bit):zipf_r_squared (-2.7z) — beyond what the standard bank predicts for it.
Zipf–Mandelbrot (8-bit):zipf_r_squared | -2.7z | bank-miss 1.8σ |









_(centered)/signed_log_z/XorShift32.png)
_(centered)/xy_path/XorShift32.png)

/barcode/XorShift32.png)
/d_curve/XorShift32.png)








/phi_spectrum/XorShift32.png)










/default/XorShift32.png)
/default/XorShift32.png)


| Nearest neighbor | Distance | |
|---|---|---|
| White Noise | 1.53 | cross-origin |
| glibc LCG | 1.59 | |
| RANDU | 1.67 |
2-adic › 2-adic:multiscale_markov_predictability | rank 304/306 | 0.0268 |
Klein Bottle › Klein Bottle:rank_deficit_max | rank 304/306 | 0.1017 |
Nonstationarity › Nonstationarity:dynamic_coupling | rank 302/306 | 1.0781 |
Predictability › Predictability:cond_entropy_k1 | rank 2/306 | 2.9977 |
Spectral Analysis › Spectral Analysis:spectral_peakedness | rank 306/306 | 0.0004 |
Symplectic › Symplectic:recurrence_rate | rank 304/306 | 0.0273 |