The Schmidt Random Number Generator Experiments

Every psi experiment before this one used a target that a person had arranged, and that fact was a permanent weakness. A shuffled deck retains traces of its previous order; a table of random numbers has structure; a person choosing a symbol has habits. None of this requires anyone to cheat, and all of it gives a subject something to exploit unknowingly. Helmut Schmidt, a physicist at the Boeing Scientific Research Laboratories, removed the problem by taking his randomness from radioactive decay.

The moment at which a particular nucleus decays is, on every accepted account of quantum mechanics, not determined by any prior state, so there is nothing in principle for a subject to detect or infer. His machine used the timing of decay events from a small strontium source to select one of four lamps, and a subject pressed a button to predict which would light. The whole apparatus recorded its own results, removing the human scorer along with the human shuffler. He reported hit rates above chance across long runs, and published the first series as Precognition of a Quantum Process in 1969. He also ran the design in the other direction, asking subjects to push the output of a binary generator toward one side rather than to predict it, and treated the two as versions of one question, since a machine whose output matches a prediction and a machine whose output shifts toward an intention are indistinguishable from the record alone.

The design was admired well beyond the field, and its influence is easy to trace: the Princeton programme a decade later, and the whole subsequent literature of micro-psychokinesis, are variations on Schmidt's machine. What the apparatus could not do was constrain the experimenter's decisions about the experiment. Nothing in a decay-driven generator determines how many trials to collect before stopping, which subjects to keep running, whether to report a series that went nowhere, or which of several possible analyses to present. Those choices sit outside the hardware, and they are where the field's difficulties turned out to live. The point was made decisively by the 2006 meta-analysis of the whole random generator literature, which found that the reported effect size declined as studies grew larger, with the biggest deviations coming from the smallest experiments.

If the effect were physical, size of study should not predict its magnitude. If the literature is being filtered by what gets written up, that is precisely the pattern expected. Schmidt continued to publish and never accepted this reading, and the argument over the correction applied in that analysis is genuine and technical. The asymmetry itself is not disputed by anyone.