|
PureBytes Links
Trading Reference Links
|
Logical fallacy: argumentum ad verecundiam.
That is, quotations from authorities instead of refuting anything I
wrote doesn't prove a point.
In particular, what you just quoted (below) conflates angles and
slopes, two different things. The statement is really about angles,
not slopes.
The fact remains, that CALCULATED slopes are invariant with scale
changes in the axes on a chart. VISUAL slopes (in other words,
angles) will change, and are indeed arbitrary.
I doubt that Eckhart or Schwager would disagree, since this is math
they should have learned in middle school. Otherwise, well, in my
life I've encountered many idiots with PhD's.... but in this case I
suspect we're not in disagreement, just stating the same thing from
different points of view.
-Alex
>Alex: Whatever. By the way Bill Eckhardt's doctoral disertation at the U.
>of Chicago was on mathematical logic. When discussing slope with Jack
>Schwager, Jack's comment was "I've always been amazed by how many people are
>either oblivious to the scale-dependent nature of chart angles or
>unconcerned about its ramifications. My realization of the inherent
>arbitrariness of slope-of-line mthods is precisely why I've never been
>willing to spend even five minutes on Gann angles or works by the proponents
>of this methodology." So I guess there are at least three of us who are
>mistaken about slope and angles. But if it works for you, bless you.
>Regards, Jack.
>
>----- Original Message -----
>From: "Alex Matulich" <alex@xxxxxxxxxxxxxx>
>To: "jack zaner" <jzaner@xxxxxxxxxxxx>
>Sent: Wednesday, December 10, 2003 10:41 PM
>Subject: Re: TL Angle function
>
>
>> Jack:
>> >Sorry Alex: If you change the scale differentially, you change the
>slope.
>> >Nuff said.
>>
>> I don't know why I'm arguing about this when it's so easy to verify
>> for oneself...
>>
>> If I calculate a slope from bar 20 at a price of 900 to bar 30 at a
>> price of 1000, I get a slope of 10 price points per bar. It doesn't
>> matter if I'm plotting it using a vertical axis scale of 1 price
>> unit per grid line or 1254.3 price units per grid line. Or if I
>> plot 5 bars per gridline or 20 bars per gridline. Visually the
>> slope will look different at the different axes scales. But the
>> slope will still calculate to the same value, and that's all that
>> matters.
>>
>> My last paragraph in my previous message below is correct. Maybe
>> I don't know the context of what Bill Eckhardt's statement, but in
>> the context I see it in your mail, he's blowing smoke. Maybe he
>> was referring to visual differences. In THAT context he's correct.
>> However, mathematically, the visual differences caused by different
>> axis scales are irrelevant, the math is all that matters.
>>
>> Naturally, if you decide to change your price units from points to
>> Euros or whatever, you'll end up with something different. Any use
>> of slope should be done in the same context, comparing relative
>> slope values (a single slope value is meaningless by itself, it only
>> has meaning compared to other slope values). Comparing slopes is
>> essentially what divergence engines do, comparing momentum of one
>> time period with a previous time period.
>>
>> Momentum is directly and exactly proportional to slope over a fixed
>> interval, and if you think momentum doesn't matter, I can point you
>> to some folks who trade for a living who would disagree with you.
>>
>> -Alex
>>
>> >----- Original Message -----
>> >From: "Alex Matulich" <alex@xxxxxxxxxxxxxx>
>> >To: <omega-list@xxxxxxxxxx>
>> >Sent: Wednesday, December 10, 2003 10:06 PM
>> >Subject: Re: TL Angle function
>> >
>> >
>> >> Jack:
>> >>
>> >> >Alex: The slope is dependant upon scale. To quote Bill Eckhardt
>> >> >in the "New Market Wizards", ". . . it would become abundantly
>> >> >clear that the slope value depends directly on the choice of units
>> >> >and scales for the time and price axes." Thus a change in these
>> >> >values could result in different slopes (and angles) between any
>> >> >two points.
>> >>
>> >> No, that's blatantly incorrect, no matter who said it.
>> >>
>> >> Price is price. A dollar of price is a dollar of price no matter
>> >> how you're plotting it. A day of time is a day of time no matter
>> >> how you're plotting it. You can stretch or compress the axes as
>> >> much as you want, but as long as you're working with price units and
>> >> time units, you get the same slopes between two points, in terms of
>> >> price change per unit time.
>> >>
>> >> Granted, you'll get different slopes if you decide price will be in
>> >> ticks per 5-minute bar instead of dollars per week. But as long as
>> >> you use a consistent price and time unit, it doesn't matter one whit
>> >> what choice of scales you use on your axes.
>> >>
>> >> -Alex
|