Monday, December 27, 2010


One very highly regarded technician (name withheld) said:

"I am always surprised by the discussions of the Hindenburg Omen as it fails what is for me the first test of an indicator, robustness. A good indicator or system must, in my view, be insensitive to small changes in
parameters. Any indicator failing that test, and the HO fails it without question, cannot be trusted."

Another also highly regarded technician (Tom McClellan) responded:

"By that same reasoning, a railroad crossing sign would also fail as an indicator due to too many false positive indications. But we still use RR XING signs because of the observation that most car-train encounters occur at points where streets cross tracks. The sign is not the sole determinant that a train is there; we also can use confirming indications like the horn, or like seeing it.

If the HO identifies the preconditions for each and every great decline, crash, or whatever you want to call it, plus has a bunch of extra false indications at times when crashes don't come, I can live with that. I'd rather know, and accept the possibility that it may turn out to be a false positive reading."

MK opines - I love being around really smart people (both of these gents). Critics who said it predicted 10 of the last 3 crashes should attempt to learn how to use technicals properly.

No comments: