[Sophistifunk]

> how can a deterministic PRNG have less entropy (0) than that of its seed

If there's any way two seeds can eventually lead to identical internal states would be one example.