At 11:52 PM 8/5/2011, Harry Veeder wrote:
From: Abd ul-Rahman Lomax <[email protected]>
No, I'm trying to *not* draw conclusions from unpublished results.
Are you trying to remain indifferent about unpublished results?
No. Impossible. However, "unpublished results" here is more drastic than, say, some conference paper. There are very few fully-published results relevant to the Rossi reactor. There are what amount to news reports, at most, with some informal compilations of data. In the case of the February test, the "18-hour test," there is even less, there are some summary conclusions presented with only a couple of data points.
One of the Langmuir's characteristics of "pathological science," and oft-abused concept, but nevertheless of some value, is that results become less as precision of measurement increases. So I was noticing here that as more data was collected and published on a demonstration, the results became less dramatic. The most dramatic results, and most convincing if truly confirmed, were from the test where the least amount of data was available.
Jed has often pointed out that a boiler test form only has a couple of data points, and this is correct, but a commercial boiler has very well-known operating characteristics, and what is being done is to simply verify that it's operating within specificiations. That just takes a spot check. This simple testing would never be done to develop initial assessment of a radically new design, operating based on unknown physical principles!
Yes, yes, the 18-hour test is the most convincing, and that's precisely the point. The least data is the most convincing?
No, it's the most *interesting.*
