Sad story - page 6

 
victorg:

In my previous post in this thread, I clearly overreacted. You can't determine the "absolute value" of the angle with an additional MA.

But the solution to the problem is very simple. Given that the tangent to MA-40 is a geometric interpretation of the derivative at a given point of the curve, let us find this derivative. To do it let's divide the first difference of MA-40(for example MA[1]-MA[2]) by the timeframe value (value). The arctangent of that ratio will give the required angle. But it is not necessary to calculate the arc-tangent; the ratio itself can be analyzed directly for angles close to 45 degrees. Further, if we equate the value of the current timeframe to one, the obtained ratio will become a simple difference (for example MA[1]-MA[2]). Thus, it is enough to simply compare this difference with the threshold value. The optimum value of the threshold can be determined by means of the strategy tester. This does not require any scaling of charts.

By the way, I think this solution has already been mentioned in previous posts.

In short - MACD.
 
icas:
In short - MACD.
If you are talking about the strategy I suggested, you are very much mistaken. Far from the MACD. ) If you make an indicator according to this strategy, it will show swings up and down, trends, and during a flat it will show a horizontal zero line. Read the whole thread.
 
dimeon, I'm not mocking you, I respect programmers, as you might have guessed from reading the whole thread. )
 

victorg, I think this way is a dead end. Because the MAs show us an average movement of the price chart and nothing more. And the longer the period, the smoother the movement. They kind of remove all the spikes. For example, why do I like periods of 50-60-70? Because they show market movement where it is and where it only seems to be, they show the horizon. ) So I don't think two will show the angle better...and one does it quite well. But your point is clear and I, for one, have an interest. I'll check it out.

Although if you think about it, what will two show? For example 55 and 60? During a trend they will be at a steep angle and slightly farther away from each other, while during a flat period they will be at a slower angle and closer to each other. Do you think that the points of crossovers of these averages in comparison with the points of crossovers of the short average will give a signal that the necessary angle and the trend is present? But their intersection points will be so far away from the real entry point, that I wouldn't use it...(

And about the arctangent of course it may be...but that's up to the programmers. )

 
Anyway, you certainly had fun with the sad story here, happy to join in, I love it. Had a good laugh. ))
 
uriukti:
Although if you think about it, what would two show? For example, 55 and 60? During a trend, they will be at a greater angle and further away from each other, during a flat, they will be at a greater angle and closer to each other. Do you think that the points of crossovers of these averages in comparison with the points of crossovers of the short average will give a signal that the necessary angle and the trend is present? But their intersection points will be so far away from the real entry point, that I wouldn't use it...(

And about the arctangence of course it could be...but that's up to the programmers. )

This is not for programmers - this is for mathematicians and inventors of TS))
 
uriukti:

victorg, I think this way is a dead end. Because MAs show us the averaged movement of a price chart and nothing more. And the longer the period, the smoother the movement. They sort of remove all the spikes. For example, why do I like periods of 50-60-70? Because they show market movement where it is and where it only seems to be, they show the horizon. ) So I don't think two will show the angle better...and one does it quite well. But your point is clear and I, for one, have an interest. I'll check it out.

Although if you think about it, what will two show? For example 55 and 60? During a trend they will be at a greater angle and further away from each other, while during a flat period they will be at a greater angle and closer to each other. Do you think that the points of crossovers of these averages in comparison with the points of crossovers of the short average will give a signal that the necessary angle and the trend is present? But their intersection points will be so far away from the real entry point, that I wouldn't use it...(

Perhaps I didn't make myself clear enough. Perhaps I should have written MA40[1]-MA40[2]. Where 1 and 2 are array indices, not MA number.

 
victorg, to be honest...I'm afraid to comment on that...it's definitely not for me. ) But someone might say something.
 
it's not for me either
 
kittilu:
that's not for me either.

Oh, you are making progress! Having started training your biological neural network with a single 'Y', you gradually moved on to building one word. Then two. And now you're up to a whole sentence. Congratulations! The next step is punctuation. :)

P.S. I hope you had a sense of humour already. )))