You are using an out of date browser. It may not display this or other websites correctly.
You should upgrade or use an alternative browser.
You should upgrade or use an alternative browser.
Do Trendlines work?
- Thread starter duard
- Start date
Quote from trefoil:
Strunk would have objected to that entire sentence, actually. It was a deliberate choice.
On the other side of Occam is a space where where you look at dividing things up into least connected regions. I'm sure you are familiar with A's Method. The point is to not have any redundancy and to only be illuminating the degrees of freedom that are required at that place in time.
I refer to it as steer and focus. In total, there are about 70 degrees of freedom and a person never requires more than 7 or so at a given moment. Steering is automatic and the extent of focus is simply a matter of sufficiency.
For example, as fair value is adjusted during the day most people know when different games kick into play. I count the drift and just use a LUT to know whose playing for that count. MLR gives me the m of the drift and I count according to the season I am in. This is managed with about 5,000 rolling data points. This isn't using probabilities, I just need to know the motion of fair value to front run given games that kick in.
Clear crisp and concise are certaily elements of style, but on the other hand you can read somewhere on ET at any time someone who has proven to himself that what I speak of is impossible, unbelievable and astonishing for those people using that vocabulary.
The fact is, that when probability math is dropped from the picture Occam is too. You didn't get it. We don't care, either.
I, in the near past, just asked if it were possible for a person to look at a trading paradigm that had removed uncertainty from the space. I really do not believe many people can let themselves do that. That is how they get stuck in unbelievable, impossible and astonishing.
I admit, if you put enough snippets together, during RTH's they make a big sucking sound. Its clear, concise and crisp as E.B. would say. Lettuce factory.
re:
liyoung (ljyoung), et alia (alii is much more friendly): a few things.
All of you need to read Strunk's (not the only style manual).
You all (might I suggest "All of you") take 10 (ten) words, big ones at that, to say what you mean, when three or four will do (strictly an opinion).
Secondly, this is for the point, if you don't know what a frequency distribution is, or a regression, you should take Stats 101 (done and more). I don't know how to be any clearer without revealing exactly what I do, and I ain't doing that for free for you guys (I don't believe anyone, other than possibly yourself, will be too upset if you keep your secret to yourself).
Thirdly, point whatever, I suppose you should be proud of knowing what a noun is (I'm not entirely sure where this came from). I'm happy for you (and me for you).
Me (as opposed to whom?), I turned down a fellowship from the University of Chicago in Linguistics for no other reason than that I was tired of school (their loss, ET's gain).
Which just means I know my declensions too.
My point, point whatever, is the same as Occam's (the razor thing?). Look it up (who doesn't know about Occam's razor?).
I find myself once again in agreement with PointOne when he observes the unpleasant, might I venture dysphoric, tone to your posts. There is a rather nasty affliction known as the arrogance of ignorance which it would seem to me that you are beset with in spades. Instead of presuming that people don't know anything about statistics, consider the possibility that others have decided, for whatever reasons not to make certain aspects of that methodology part of their trading protocol. To rabidly insist that they must incorporate your brand of statistics into their protocol, reminds me somewhat of having a conversation with an Islamofascist about the nature of the deity.
As I said before if you are making money with your trading method, cool, but don't assume that your particular view of numbers is the only one. Mandelbrot and Taleb are worth a read.
lj
liyoung (ljyoung), et alia (alii is much more friendly): a few things.
All of you need to read Strunk's (not the only style manual).
You all (might I suggest "All of you") take 10 (ten) words, big ones at that, to say what you mean, when three or four will do (strictly an opinion).
Secondly, this is for the point, if you don't know what a frequency distribution is, or a regression, you should take Stats 101 (done and more). I don't know how to be any clearer without revealing exactly what I do, and I ain't doing that for free for you guys (I don't believe anyone, other than possibly yourself, will be too upset if you keep your secret to yourself).
Thirdly, point whatever, I suppose you should be proud of knowing what a noun is (I'm not entirely sure where this came from). I'm happy for you (and me for you).
Me (as opposed to whom?), I turned down a fellowship from the University of Chicago in Linguistics for no other reason than that I was tired of school (their loss, ET's gain).
Which just means I know my declensions too.
My point, point whatever, is the same as Occam's (the razor thing?). Look it up (who doesn't know about Occam's razor?).
I find myself once again in agreement with PointOne when he observes the unpleasant, might I venture dysphoric, tone to your posts. There is a rather nasty affliction known as the arrogance of ignorance which it would seem to me that you are beset with in spades. Instead of presuming that people don't know anything about statistics, consider the possibility that others have decided, for whatever reasons not to make certain aspects of that methodology part of their trading protocol. To rabidly insist that they must incorporate your brand of statistics into their protocol, reminds me somewhat of having a conversation with an Islamofascist about the nature of the deity.
As I said before if you are making money with your trading method, cool, but don't assume that your particular view of numbers is the only one. Mandelbrot and Taleb are worth a read.
lj
Mr Hershey:
re: On the other side of Occam is a space where where you look at dividing things up into least connected regions. I'm sure you are familiar with A's Method. The point is to not have any redundancy and to only be illuminating the degrees of freedom that are required at that place in time.
I refer to it as steer and focus. In total, there are about 70 degrees of freedom and a person never requires more than 7 or so at a given moment. Steering is automatic and the extent of focus is simply a matter of sufficiency.
For example, as fair value is adjusted during the day most people know when different games kick into play. I count the drift and just use a LUT to know whose playing for that count. MLR gives me the m of the drift and I count according to the season I am in. This is managed with about 5,000 rolling data points. This isn't using probabilities, I just need to know the motion of fair value to front run given games that kick in.
If you will, I have some questions about what you said immediately above:
(1) Who is "A'?
(2) How did you come up with 70 degrees of freedom?
(3) What is "LUT"?
(4) What is "MLR"?
(5) What is "m"?
(6) What does "count according to the season I'm in" mean?
Thank you for any help you might be willing to give.
lj
re: On the other side of Occam is a space where where you look at dividing things up into least connected regions. I'm sure you are familiar with A's Method. The point is to not have any redundancy and to only be illuminating the degrees of freedom that are required at that place in time.
I refer to it as steer and focus. In total, there are about 70 degrees of freedom and a person never requires more than 7 or so at a given moment. Steering is automatic and the extent of focus is simply a matter of sufficiency.
For example, as fair value is adjusted during the day most people know when different games kick into play. I count the drift and just use a LUT to know whose playing for that count. MLR gives me the m of the drift and I count according to the season I am in. This is managed with about 5,000 rolling data points. This isn't using probabilities, I just need to know the motion of fair value to front run given games that kick in.
If you will, I have some questions about what you said immediately above:
(1) Who is "A'?
(2) How did you come up with 70 degrees of freedom?
(3) What is "LUT"?
(4) What is "MLR"?
(5) What is "m"?
(6) What does "count according to the season I'm in" mean?
Thank you for any help you might be willing to give.
lj
Quote from trefoil:
No one can remove uncertainty from the space.
Degrees of freedom are needed precisely because that sentence is true.
But you knew that, didn't you?
I can't answer your question.
The group I work for did remove uncertainty from the space we are using to trade. We know that we know. The mathematical expression is: the sweep is possible. It comes from info gap theory a non probabalistic way of looking at stuff.
The maths are as I have previously described.
I am sure others have done this as well based on what is currently being published.
The deal in doing markets is taking what is offered. It gets clear, that at any time what is offered is known. Taking is done by participating and not affecting the market.
We behave as what are mistakenly called parasites. Blame the financial industry for the name.
ES, a derivative, is a slave to its master and ES because of its relative size of construct is not a swift as the primary other derivative of the primary market.
So that is that.
By only using data that has one of two values; the probability is taken out of the picture.
Conjuring is not required to do this. critical thinking is a nice thing to have in the quiver, though. It actually may be a cross of linguistics and maths. Its probably just more like the market is music.
Guy sombody wrote music of the spheres.
Faith didn't write the music of the markets; he stuck with Turtles. The Singapore subway is fun to ride, too.
The music of the markets is not probability. It is simply a record of continuing human behavior in a huge chorus too big and loud for any voice to make a difference.
Any day on ES can be played in 40 minutes without missing a beat visually. Another thread has a guy thinking he can compress the market's play by a factor to make some money not even close the what is offered.
By playing a day in forty minutes you get to see what you need to pay attention to. Actually, very few people playing a day get to see what is going on.
Almost no one is displaying what is going on, so,logically, thee is no point is playing it back since it is not seen in the first place.
To get down where the rubber meets the road, the NOW is not enough even if it is coupled to the very very near past to make vectors. By moving the NOW over to the left (in the direction of the past) and looking at what NOW is becoming, is when you get your first look at the market.
I assume that very few people do this or have a combination of platforms that make it possible.
As a person gets acquainted with the markets he finds that displays aren't where it is at in the final analysis.
In monitoring, analysis, decision making and action. The operating place is sort of between decision and action. NOW is to the left.
I speiled on the four games and gave the players colored jerseys a while back. Units, tens, hundreds and thousands place holders on the cars.
Seeing "control" is where "the knowing that you know" is played.
At some point we will introduce advanced intermediate and expert trading. Then this operating place will finally show up in ET.
The market operates on facts. A finite set of facts. The sets of fact "pour" out. Better still the facts play out just as an orchestra plays. The elements have timbre, etc....
I was at the MERC recently and I commented on two days of trading occasionally. i made it a point to run out 20 to 25 minutes so it would be easier for my colleagues to enjoy the moments of the trading. No surprises that way. It is like turning the pages of the orchestra score. The notes are facts on each line of the composition.
Saying with your finger on a screen a level of price and a time coming up makes it much better for a trader to have the turn noted as the music comes to that value and time for the next turn.
It is case of moving a little ahead of NOW and only as far as the next profit taking and reversal.
It can be done by writing coordinates on a sheet of paper and skipping the display. The best way is to just have a thing like a power plant control panel. The degrees of freedom show as required and values show too. The music plays.
Stay in your Occam place and just keep noticing that the choice presented to you now is not the choice for next year or 10 or 20 or 50 years from now. Maybe you will drop the Occam thingy at some point, even.
I lucked out. I didn't get into probabilites starting day one (1957). Then, I just used sheets of paper as a matter of fact. I used Stunk and White too. 31 books worth. Now I am knocking off several more.
Its funny you are stuck in regression and distributions and I took the psychological boundaries approach thank God the boundaries overlap and we know ahead of time.
You can see the difference between your emotions and mine; they are consequences of the space being occupied.
I'm not saying that Mandelbrot isn't good. Merely to be clear about what it is about him that is good.
And, considering that fractals are harder than Stats 101, would fractals reward you with more money for the extra effort than Stats 101 would?
For instance, and someone else somewhere in this forum pointed this out, Wilder's RSI I have found, gives good info when it goes from below 50 to above 50, and vice versa. It also sometimes gives good info when it goes into the oversold/overbought areas that are normally defined, below 20 and above 80. But not often enough to make up for the increased trading you'd have to do to take advantage of that information, I've found.
I've also tried a few of the other ways that have been suggested for its use, and found pretty much the same thing. Some give good info, but the increased trading cost added to the increased risk eats up the incremental advantage over just using the 50 crossover to trade.
Now all of that is clear. Your results may vary from mine on this indicator, and we can go back and forth about that. By contrast, none of what Mr. H posts is at all clear about what he's actually doing. Therefore, debate is closed off, and the only thing left is to either accept what he says, or not. But none of it is either testable or verifiable.
That's what I object to.
And, considering that fractals are harder than Stats 101, would fractals reward you with more money for the extra effort than Stats 101 would?
For instance, and someone else somewhere in this forum pointed this out, Wilder's RSI I have found, gives good info when it goes from below 50 to above 50, and vice versa. It also sometimes gives good info when it goes into the oversold/overbought areas that are normally defined, below 20 and above 80. But not often enough to make up for the increased trading you'd have to do to take advantage of that information, I've found.
I've also tried a few of the other ways that have been suggested for its use, and found pretty much the same thing. Some give good info, but the increased trading cost added to the increased risk eats up the incremental advantage over just using the 50 crossover to trade.
Now all of that is clear. Your results may vary from mine on this indicator, and we can go back and forth about that. By contrast, none of what Mr. H posts is at all clear about what he's actually doing. Therefore, debate is closed off, and the only thing left is to either accept what he says, or not. But none of it is either testable or verifiable.
That's what I object to.
<i>The group I work for did remove uncertainty from the space we are using to trade.</i>
Clearly, this is logically impossible.
<i>NOW is to the left.</i>
Physically impossible. One cannot both be in the future and in the present.
You say these things, and expect, nay demand, to be taken seriously. You cover up the craziness and impossibility of what you're saying with big words.
Me, I find it difficult not to just laugh.
This was a thread with some promise until it turned into a debate all about you. If you want the true source of the anger you detect, that's it. Your ego sucks the oxygen out of this place.
Clearly, this is logically impossible.
<i>NOW is to the left.</i>
Physically impossible. One cannot both be in the future and in the present.
You say these things, and expect, nay demand, to be taken seriously. You cover up the craziness and impossibility of what you're saying with big words.
Me, I find it difficult not to just laugh.
This was a thread with some promise until it turned into a debate all about you. If you want the true source of the anger you detect, that's it. Your ego sucks the oxygen out of this place.
Quote from ljyoung:
Mr Hershey:
re: On the other side of Occam is a space where where you look at dividing things up into least connected regions. I'm sure you are familiar with A's Method. The point is to not have any redundancy and to only be illuminating the degrees of freedom that are required at that place in time.
I refer to it as steer and focus. In total, there are about 70 degrees of freedom and a person never requires more than 7 or so at a given moment. Steering is automatic and the extent of focus is simply a matter of sufficiency.
For example, as fair value is adjusted during the day most people know when different games kick into play. I count the drift and just use a LUT to know whose playing for that count. MLR gives me the m of the drift and I count according to the season I am in. This is managed with about 5,000 rolling data points. This isn't using probabilities, I just need to know the motion of fair value to front run given games that kick in.
If you will, I have some questions about what you said immediately above:
(1) Who is "A'? alexander; he's over at Berkley
(2) How did you come up with 70 degrees of freedom?Carefully. I let you know soon
(3) What is "LUT"? Look up table
(4) What is "MLR"?multilinear regression
(5) What is "m"?m stands for slope. it is the slope of the MLR line and it serves to show the change over time of the drift as measured in terms of "noise"
(6) What does "count according to the season I'm in" mean?trading varies over the seasons as a function of vacations of rich people. It is a tradition to do certain things at certain times. The New York Yatch Club criuse is one of these. Opening summer homes is also part of the schedule. This shows up in breakpoints on a lot of degrees of freedom. so a person has to deal with this to keep at optimum. As it turns out the ball park for me is the fourth significant figure when I am eyeballing. that means my snippets of software for all of these indepentent things have to be pretty good to make them worthwhile as a BOT of me. i have very good BOTs
Thank you for any help you might be willing to give.
lj
We can get into how differnet things are important at different times. It is easy to reason through the notion that a person has to be able to see the markets and know what is going on. We all recognize that.
The shift from EMH to AMH was an advent that put behavioral finance on the map at the turn of the century. The fact that EMH is not arbitraged fully means that the quants are not on it from their stats point of view (stats 401, 501 and 601). Trefoil shows us how the behavioral factors dominate him even in his prose and at his level of stats (101 plus..).
The work around for getting an optimal pool extraction paradigm running is to not have to react (this is a behavioral term that is thought to be handled by arbitrage quants).
The market is huge and it only migrates as a consequence of alternatives being blocked and, thus, leaving but one path of change. I'll do a John Nash on this someday.
Counting down the alternatives being closed off in a multi-dimensional matrix is a fun and entertaining way to observe. Personally, I do not believe this has come to mind for a lot of people as yet.
There is another aspect of this. When it is realized that you are required to look in the right place at the right time (Trefoil drones along with the same purview all the while, for example.), then, you get to understand that the matrix is capable of telling you when something is no longer important for the time being since it is blocked.
The logic of "steering" can appear at this point of critical thinking. It is a snippet consideration.
Markets and how they do their thing is largely counter intuitive in terms of human group psychology. behavioral finance is not looking at this, instead they focus on the extent to which traders are damaged by the traits that trefoil is kind enough to exhibit for us this evening. The damage is statisitcally significant in terms of the bottom line.
I think it may even be contagious from my observations in the Merc trading rooms.
For sure thee is a line that, once crossed, is not reversable. See lizard syndrome.
By now many of the degrees of freedom are coming to the surface for you. 70 is no big deal. There is no probability involved. That has to be anihilated from the get go.
It is simply not going to happen for mostly everyone who is part of the financial industry. This is a real relief for me as you can imagine.