|
PureBytes Links
Trading Reference Links
|
On Wednesday, May 23, 2001, 8:00:37 AM, Ira Tunik wrote:
IT> I have read all the drivel and supposed humor about the moon, oysters
IT> and other things mentioned. I am not into astro, but I do believe there
IT> is a repetitive cycle that occurs. If it can be tied or anchored to the
IT> moon, to a fictitious count, to one planet interacting with another or
IT> tarot cards, I will investigate and say yes it is there or no it isn't
IT> there.
Good idea. I would agree entirely. You can often save a bit of time by
checking the literature. A number of cycles are well known and do not
need additional verification.
IT> I bought the Delta book and read it and believed that it had
IT> some validity ,but not enough to spend the money for the software or to
IT> join the society.
Likewise bought the book, and did a bit of checking before going
deeper.
IT> but if you allow +/- a couple of days, I believe that you will
IT> find a high correlation.
This is where we may differ a bit. Let's do a little thinking
experiment (the numbers are not exactly correct, though). With +/- 2
days, we are allowing a band 5 days wide around each prediction. If
the bars are, say, about 28 days apart, about 18% of all days are
'acceptable'. With the additional bands and conditions that are
allowed, we end up with, say, 25% of all days as 'acceptable'.
Now, if we just ask if there are highs and lows in any of the bands,
the answer is "of course". There are, in fact, quite a few. If this is
your criterion for acceptance, then you will believe it works. Most
people do not think statistically, and I think many will use this
criteria for deciding.
But I suggest that a better question, and one that will be kinder to
your trading acount, is to ask if the number of H&L's are *greater*
than would be expected by chance. Another words, if 25% of all days
are acceptable, does the technique show *significantly* more than 25%
of the H&L's within the bands. If so, then it is better than chance.
If not, then it is not better than a coin, so to speak.
When I ran such an analysis for Delta, it did not do better than
chance. If anyone knows of a stat analysis that comes to a different
conclusion, I'd very much like to know about it. When I wrote to WW
about it, he did not have a different result either. Given that, I
would still view Delta as no better than chance.
IT> You look at the highs and lows of the various bands and see if
IT> the match up.
Try to look VERY carefully. Many will find that their eye is tempted
to include a H/L as 'OK' even when it is more than +/- 2 bars away.
The eye is quite easily fooled sometimes. :-) This is one case where
the enforced discipline of computer analysis is useful, IMHO.
IT> I never cast someone else's work out
IT> without seeing if there is any thing there that might improve my
IT> trading system without making it more complicated.
I agree.
IT> You be the judge, Newton discovered gravity because an apple fell.
IT> I am glad that he wasn't on this list or a lot great physics
IT> formulas would have never been made public.
I believe Newton would have given some *very* good reasons to the list
for why his ideas were valid, and would have answered scientific
questions quite well. If anything, he would have convinced the list he
was right.
IT> Oysters are far more complicated and who knows what great
IT> discovery might come from them.
Actually, we have learned a lot from oysters regarding biological
clocks. But, so far, I have not seen a good explanation for *how*
oysters are able to move the stock market.
ztrader
To unsubscribe from this group, send an email to:
realtraders-unsubscribe@xxxxxxxxxxxxxxx
Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
|