I would like to propose, with the help of professionals in the field of seismology and statistics, an award to be presented to any individual who is able to demonstrate a prediction methodology, OR record, that results in predictive success that bests the "null hypothesis". Below are some of the considerations that would need to be addressed first.

- The award would include a monetary enhancement. I am willing to contribute $USD 5000. into an escrow account, as "seed money," in order to make this a genuine, worthwhile and interesting endeavor. It would be my hope that other individuals would contribute also. The award moneys could be used to offset the predictor's costs involved in establishing his or her success, or for any other purpose. I reserve the right to retrieve my money, principle and interest, at any time subsequent to January 31, 2017 UTC if it has not yet been awarded at that time.

- I propose that the award be titled "The Earthwaves Prediction Award," to memorialize this site's singular contribution toward an analytical approach to earthquake prediction. I welcome other suggestions.

- The Award will only be presented to individuals (or representatives of individuals) who are not employees of any governmental agency directly involved in Geology, Geophysics or Seismology.

- It is essential that a method of statistical analysis be in place first, and that it be one which does not present the predictor with discouraging prospects, nor that it be too easy to win. Here is where I need the most help. I am aware that statistical significance can be presented in a number of ways. What I propose is a level of significance that could be considered "highly likely" to be statistically significant. I understand, also, that matters such as clustering complicate the situation, but would hope that a consensus approach might provide a solution.

- Part and parcel of the above would be, I think, the necessity for predictors to present their predictions with precise parameters of magnitude (expressed as a minimum magnitude, to the nearest 1/10M, as determined no sooner than two weeks subsequent to the quake, by such an authority and on such a scale as are appropriate to the location and size of the quake), location (any method which perfectly specifies the location on the Earth's surface of the epicenter - circular, rectangular, polygonal or "within X kilometers of the surface expression of such-and-such a fault, as portrayed on such-and-such a fault map", etc.), and time (I propose time and date UTC - not so hard to provide). At the risk of being labeled ethnocentric, I propose only U.S. governmental agencies' determinations as to magnitude, time and location.

- Also intrinsic to the methodology of analysis is the requirement that such predictions as are to be part of the analysis of statistical significance must be in place, in the proper format, in an uneditable and completely public forum.

- With my limited math skills, I can foresee the problem of one prediction attaining success due solely to the high magnitude of the quake along with moderately tight parameters of time and location. As there is likely to be at least one prediction extant somewhere for even the largest earthquake, I am not sure how to address this. It might be sufficient to expect that, since only predictions that meet the specifications above will be considered, it is highly unlikely that anyone will simply "get lucky." Or, another possibility, that would also help assure both that the prediction scheme has some practical value, and that the analysis not be extraordinarily tedious, is that candidate predictions' minimum magnitudes be something like 4.0, locations be less than 20,000 square statute miles, and time spans be less than one sidereal (Skywise?) or calendar (UTC) year.

- I am quite serious about this, and am not inclined to quibble too much. Even if I lose the money, and have to eat some crow, it should be very interesting. I suspect that I will not lose the money, and that, for many years into the future, those of us who have for so long tried to plead our science-based case, will be able to point to this outstanding Award as additional evidence for the correctness of our position.

Michael F. Williams
Arroyo Grande, CA USA