You specified the following parameters as follows (in your two posts)-
1) area: 33-35N 115-125 W,
2) Magnitude range: 3+ (Ml>=3) (3-10 in search)
3) Time window: a 10 day window (from the original FFA post: http://www.earthwaves.org/wwwboard/messages/16154.html)

I identified Dr. Alan Jone's method which is accessible to the reader by way of my link to Roger's download site as well as a search engine search.

The only thing I did not mention was the time interval I used for the observed probability. This was from 1973 to 2002 (NEIC). Many people would argue (especially statistics theory) that the best estimate of observed probability comes from that record that includes the longest span of time (i.e more data points). The issue of clustering from aftershocks is addressed by Dr. Alan Jones' method.

I see that Roger has repeated the test using 118W as one boundary rather than 125 W (your original specification) and found a probability of 71.3%
This is fine since it would include only the area to the east of LA. 71.3 % is still a high probability and suggests that M3.3 are relatively common east of LA in southern CA.

Your statement:

"I can get statistics to say anything I want and the result is meaningless."

raises the question of:

"Why should anybody believe your statistics"?

------------------------------------------------

I support your trying to solve this earthquake prediction puzzle and all of your efforts to achieve that goal. I think, however, that you may wish to consider an alternative to providing support to the FFA concept/theory. A wider statistical treatment is needed (coupled with some retention of "statistics" as a tool for identifying what is "significant" and what is not ...contrary to your remark above about "getting statistics to say anything I want").

I have seen other FFA examples in the past which appear to be more interesting / convincing than the one you have cited here. Perhaps in future postings the output of the test that Roger's software performs should be considered before posting an FFA "hit" claim. I think this approach will serve everybody better.

As you well know, there are many skeptics waiting in the wings to skoff and ridicule. Let's not provide them any ammo by making claims on events which to most people would be perceived as relatively routine. I can hear a skeptic arguing that since no 3+ has happened in the last few months that the place was way (over)due for one anyway...depsite any claimed FFA effect.

Maybe we can get Roger to get his program to read in CNSS data files too which may be more accurate (from what I've read on this board).

Happy Independance Day &

Just my $.02 worth