The conventional view, that Einstein was wrong to believe that quantum physics is local and deterministic, is challenged. A parametrized model, `Q', for the state vector evolution of spin-1/2 particles during measurement is developed. Q draws on recent work on `riddled basins' in dynamical systems theory, and is local, deterministic, nonlinear and time asymmetric. Moreover, the evolution of the state vector to one of two chaotic attractors (taken to represent observed spin states) is effectively uncomputable. Motivation for considering this model arises from speculations about the (time asymmetric and uncomputable) nature of quantum gravity, and the (nonlinear) role of gravity in quantum state vector reduction. Although the evolution of Qs state vector cannot be determined by a numerical algorithm, the probability that initial states in some given region of phase space will evolve to one of these attractors, is itself computable. These probabilities can be made to correspond to observed quantum spin probabilities. In an ensemble sense, the evolution of the state vector to an attractor can be described by a diffusive random walk process, suggesting that deterministic dynamics may underlie recent attempts to model state vector evolution by stochastic equations. Bell's theorem and a version of the Bell-Kochen-Specker quantum entanglement paradox, as illustrated by Penrose's `magic dodecahedra', are discussed using Q as a model of quantum spin measurement. It is shown that in both cases, proving an inconsistency with locality demands the existence of definite truth values to certain counterfactual propositions. In Q these deterministic propositions are uncomputable, and no non-algorithmic mathematical solution is either known or suspected. Adapting the mathematical formalist approach, the non-existence of definite truth values to such counterfactual propositions is posited. No inconsistency with experiment is found. As a result, it is claimed that Q is not constrained by Bell's inequality, locality and determinism notwithstanding.