Quote from smilingsynic:
You seem inebriated by your own verbosity. I have a Ph.D.--I know snow when I see it. And although the temp is in the upper 80's, there is plenty here.
I posed a few questions that help to point out altenative views to the CW.
Naturally this thread has wound up in chit chat.
Those who trade real money know that optimizing and reoptimizing canned indicators in order to find the perfect parameters to "trade" past data in shoulda, woulda, coulda form is a waste of time.
I have the view that creating ATS's comes from three autonomous sources: Snow 1, 2 and 3.
Snow 1 is the market. Snow 2 is the strategy and Snow 3 is the trading platform.
I have found that working up the development, stage by stage, for each Snow is best done autonomously, that is, in an independent manner.
The model of the market, Snow 1, for me is something, independently refined, and that comes as close as possible to delivering information on what is happening in a non stationary window using Algebraic based information where certainty is attained regarding the information created and stored.
I think of all aspects of this as the use of building blocks and the blocks have two inputs, conduct a process and produce outputs. One input is real time data (as provided) and the other input is processing oriented. Conducting a process (the contents of the box) is done with logic that in common parlance might be considered a combination of hardware and the fixed unchanging script used by the hardware to precipitate (snow).
Pragmatically, with people it looks like this. The model is in existance and perfoming mechanically. Those who wish to use it do.
To go into an business application it is necessary to take Snow 1 to a business oriented state. As the moderator who posted while this was in the strategy forum said, the model being operated mechanically needs to be validated. He meant Snows 1, 2, and 3 and their combination. He described how he thinks the path looks to get to validation. You described a path that doesn't work to get Zig ZAg David 1 or any other ATS tweaked to improve it's performance (probably based on my twaeking recommendation where you may have thought I wanted to use past data for some purpose you created).
From a model operating mechanically there are many steps to handing a validated ATS to a Business where it will be applied. It is a two way street. Specs go from the business to the admin and then the job begins.
Historically and with respect to the highlights that illustrate our differences, I will note a couple of things.
Snow 1 has the purpose of monitoring the market and produxing a set of degrees of freedom that provide processed data that enable the Snow 2 to operate with certainty at all times. Snow 2 needs complete data sets to do this at regular intervals which exceed the frequency of the interval in which the human demand resides. ATS's function to extend human capability. See the development of the sharing of flight responsibilities in modern aircraft. Aircraft built todaycannot be flown by pilots along using pilot controls. This emphatically qualitfies development of models so that businesses may use ATS's.
Snow 1 consists of software that feeds a computer who's hardware can be adjusted to carryout functions at a desired rep rate and the outputs of the functions are stored and tapped "on demand" by the Snow 2. Snow 1 is "comprehensive" as a consequence of being non inductive and non probabalistic and it operates in a non stationary manner whereby the "window" is adjustable and archiving is continuous after the window has passed.
The development of Snow 1 is complete when all aspects of the data sets required by Snow 2 are met. Snow 1 simply generates elements of a number of possible subsets of a finite set of subsets.
Lets say you and Vikana and trader666 are the crew who did this work. You pick up building blocks and you do work. Someone checks them out and gives you an EC if necessary until your work reflects the block. The books of building blocks show the interconnections. The work you did becomes part of Snow 1. It is a separate automonmous system all spec'ed to provide "on demand" what and only what Snow 2 wants.
Snow 2 is done the same way and the trading strategy is there as a separate autonomous system.
If you have ever attended Trader Expo and have seen Chris of Worden Bros do "blocks" in a prsentation or hung around their booth when we did some blocks, then you get a sense of what I said above.
You say tweaking isn't whatever. Chris intro's a tool and the audience goes aha and ewe..... Chris is cool because the tool is cool and the audience likes it too. Chris also is so cool he thinks the opposite of you. He has a P&L for the tool on the NAZ 100 and it is making money. the audience sees it making money. we all know how much money it makes.
Now comes the moment..... Jack......do you want to tweak it? This is Chris asking a member of the audience a question... Chris is rolling out a product..... He knows what tweak means.... he know my first name and I have no badge on to keep my visibility LOW.
I have a lot of Snow under my belt and so I suggest a Snow 1 to Snow 2 new connection and a new filter in Snow 2 to get more money made....
Tweak tweak..... An aha is an aha but an AHA is a lot different than an aha....
Why does the tool now WORK instead of just "work" as before? You don't know and you don't care and you don't tweak.
If I am using a "fine" level degree of freedon out of Snow 1, I can get 11.25 points in 250 reps out of block [S/S] signals of Snow 2 because I set up a Snow 1 block that allowed two degrees of freedom of to be available in Snow 1 to be sent to Snow 2. In this case, the archives show that the range during the 250 reps was 3 points. This is tweaking and has nothing to do with the use of data in a back test. What it does have to do with is looking at the logic of building blocks in trading and separately the building blocks in monitoring and putting 2 and 2 together at the Snow 1 and Snow 2 interface.
They also know that 10X atr, or even 3X atr, is not a practical daily goal. For most traders, breaking even is a goal for which most can only hope. If you have real-time trading profits that refute my contention that the goal of 3X daily atr is absurd, I welcome you to provide them. If not, then admit the truth: you are a narcissist here only for attention.
I look at this paragraph as something someone might slide across my desk for an assessment. Were i to pull out an 11x17 and my four color ballpoint pen, I would be able to sketch out the paragraph with inputs and outputs and in between a few boxes where information came in horizontally and I put arrows down into them and put timing function names on these lines. Each box would have a name and I would sketch the internal logic in a balloon and send the balloon to the box.
When I was done someone would pick it up and put it on the cartoon bulletin board.
Here is the way it is in some places. People bear down and do work. They work and get the job done. Some of these people help others. I understand that you have never been in such a place. It is very common when several businesses are run in a community or in buildings for people to occassionally look over each others shoulders. They get to observe the consequences of people working and that these people can get something done.
