What is a trend? - page 6

 
khorosh:
Where is the guarantee that the indicator does not draw? I have one with the same beautiful picture, but it redraws.

Well, we're not talking about guarantees here. Only Ostap Bender gives a guarantee ;)))

If it re-draws, it means that additional (intermediate) filtering or introduction of a dead zone is required.

 

The trend is different for everyone. There is no single definition and there cannot be one. Someone sees one thing, someone sees another on the same instrument/FT. But in spite of the fact that the trend is different for everyone, everyone classifies it in some way. And this classification is not a single criterion, but a group of parameters. Our brain automatically distinguishes these parameters and says - this is a trend. But it is very hard to explain why this is a trend. We need to learn something common for each pattern that fits the definition of a "trend" and write down those parameters the brain pays attention to "unintentionally". When a certain list is accumulated, go through the history one more time and see which parameters are most frequently present in the "trend" and which ones are less frequent. Those that occur less frequently - put them aside for now. Those that occur more often - algorithm them. And check them. If we can not algorithmize them, we can try to break down the parameters into smaller ones and algorithmize them.

In general, the process is not quick, but if you approach it carefully, the result should be there.

 
khorosh:
And where is the guarantee that the indicator does not draw? I have one with the same beautiful picture, but it overdraws.

An indicator MUST show the current state of the market! Do you work with historical data? Then what does it matter what the indicator shows on history...?

Yes, there is an opinion that re-rating is bad... But it's time to start understanding the data that is really needed to open trade orders and which are obsolete because they are not needed.

 
Serqey Nikitin:

An indicator MUST show the current state of the market! Do you work with data on the history? Then what does it matter what the indicator shows on the history...?

Yes, there's a REAL opinion that re-rating is bad... But it's time to start understanding the data that is really needed to open trade orders and which becomes obsolete as it's not needed.

Suppose the indicator has changed its direction and you entered in this direction. Immediately after that the indicator redid and returned to the previous direction. Because of this your entry was wrong and you made a loss, but you have received moral satisfaction, because you have used the current market state, shown by the indicator.)

Of course, you will now say that this is just one of many indicators in the TS, you have a lot of oscillators that filter entry. But do you really need such an indicator, which may periodically give false direction? I think it's better to let it work with lag, at least it will be expected but it won't overdraw. Or maybe you think that it is possible to make a good TS from a set of not very good indicators. Maybe so, but nevertheless the question is debatable.

 
khorosh:

Suppose the indicator changed its direction and you entered in that direction. Immediately thereafter, the indicator redid and reverted to its previous direction. Because of this, the entry was wrong and you made a loss, but you have received moral satisfaction, because you have used the CURRENT state of the market, which was shown by the indicator.)

Are we discussing about the RIGHT indicators, which are mathematically calculated, or about indicators, like flipping a coin?

If it's the latter, then I agree with you. But you hardly use such indicators... So why are we discussing this circus...?

 
If we take some indicator without the possibility of redrawing by calculating from the penultimate bar and place it on the D1 chart - we will get signals every midnight or less frequently. The same indicator on M1 will give signals much more often
 
Serqey Nikitin:

Are we discussing RIGHT indicators that are mathematically calculated, or indicators like flipping a coin?

If it's the latter, then I agree with you. But you hardly use such indicators... Then why are we discussing this circus...?

And in the bazaar, those who calculate mathematically get beaten more than those who happen to...

 

then how is the right one different from the wrong one?

 
Ivan Negreshniy:

And in the bazaar, those who calculate mathematically are beaten harder than those who accidentally...

Yes, you're right! Applying maths to trend calculations is too much..., it's better "by eye"...
 
khorosh:

Suppose the indicator changed its direction and you entered in that direction. Immediately thereafter, the indicator redid and reverted to its previous direction. Because of this, the entry was wrong and you made a loss, but you have received moral satisfaction, because you have used the current market state, which was shown by the indicator.)

Of course, you will now say that this is just one of many indicators in the TS, you have a lot of oscillators that filter entry. But do you really need such an indicator, which may periodically give false direction? I think it's better to let it work with lag, at least it will be expected but it won't overdraw. Or maybe you think that it is possible to make a good TS out of a lot of not very good indicators. Maybe so, but nevertheless the question is debatable.


The indicator that lags for a day will only show you how much you could have earned a day ago if you knew what it draws now.

Reason: