Cathryn, Michael and Canie,

Could be Canie meant "allude." I do not know.

In Roger's post to Michael he said "My past objections have been based on my experience with predictors who really don't want to be pinned down. The great Japanese debate for a recent example" and it looked like an indirect reference to me and my criticism of prediction evaluation concerning Jim Berkland. If so, I disagree, but no longer have interest in debating what constitutes pinning someone down, and so on, especially if probability based on historical catalogs gets dragged into the discussion.

Then came Roger's comment that I am always just "stirring the S__t" which could be construed as me being "outright insulted" as Cathryn rightly suggested, though I myself did not feel insulted. In fact I did not consider Todd saying I was "full of S__t" an insult either. In the latter case I just thought it was ironic that Todd claimed some words are taboo and his examples were "dangerous" and "tension" and then he went on to use S__t, which is an OK word with me, but often in American culture it is a forbidden word. Some of my favorite comedy records are George Carlin and his analyzing of banned words.

Then later John came on and said that "tempers flared" but by then I was unable to post any reassurance that my temper had not flared, nor to voice my wonder why mentioning tension is taboo but not flaring tempers? (Cultural note: "tension" is a word used in Japan about 36 times a day; in fact there are several words for it; "dangerous" is used everywhere, even in recorded advice given to people riding escalators.)

Now, to be topical, regarding the Prediction Registry threads, I agree most with Canie and Michael Tolard. In particular, Tolard's ideas are very similar to mine even in the details. Specifically Tolchard hit the nail on the head by saying that a prediction covering the whole world is acceptable, but the usefulness is a judgement call. That is what I had once been saying too, in my own sweet way.

The only thing I would add is that if software was developed to automate the various inputs of predictions, more people would find the service reliable and useful, and would probably use it. For instance, if they used the mouse and clicked on a map to create an area, which then automatically recorded the relevant coordinates. This way, for example, they can decide in their own precise way what they mean by "San Francisco." A sophisticated version could even give them an instant assessment of their area according to a complete historical catalog, which is more or less up Roger's alley.

Risking repetition, I would say that the Registry should just be a way to record predictions in a reliable, uniform way. With software tools, inputting into the Registry could be easy, fun and educational. Various optional ways of scoring points to grade the predictions could then follow, but only after the service is popular. In reverse, debating the scoring of predictions first, is futile or irrelevant.