You missed my point: Supercomputers don't think. Not everything about pattern recognition is known and programmed into a supercomputer somewhere, and no existing supercomputer is going to create such currently unknown wisdom on its own. I recently made a discovery in money management which is shockingly obvious to me in hindsight but if anyone else has made the same discovery, he or she has not publicly disclosed it. In fact I go to a quant site like Wilmott and they are less aware of some aspects of position sizing than we are here in ET. Certain training and knowledge can be eye-opening or it can be a straitjacket, depending on the individual. Neural networks were supposed to do what you think genetic programming is capable of doing now. The final rule for supercomputers is still Garbage In Garbage Out. They are very useful tools for those who fully understand their capabilities and limitations, they are not "partners" who are going to help us issue in some golden age of ultimate wisdom.
Yeah Bass wrote "The Predictors". 309 pages. Keep in mind UBS got sucked into buying the company. slip to page 273 of 309. All the original employees but one have quit. LOL. Why? The boss was fucked up. Now back to page 273. The books reads: At the same time every day ........ Yes these shithead quants are doing trading with real money at the "same time every day" .. Bass has written a book. He finds out something very late in his research. He proves he is totally unaware of markets and trading just as these shithead quants are. Now we look at two things" 1. Marketsurfer's reference. 2. When the shithead quants find out their data feed does not do bid or ask identification. They see data and do not even know if it is bid or ask data. Are these guys in the book dumber than ET members. Yes and No. Most members are as dumb as marketsurfer who is really baked clay. But there are some members that know MS is totally full of shit. How far have things progressed? For surf they will never progress. He still believes the bullshit Bass wrote about some shithead losers. Today we can use the leading independent variable (volume) of the dependent variable (price) to determine 15 to 30 minutes ahead just when one of four types of trends is going to make a turn. Just use the Mordrian table ingredients. TA is complete and objective. there are no flaws, no noise and no anomalies.
Ok, man, we agree on lots of points. Thank you for the clarification. A question, how do you account for claims made by nodoji, Cornix and others that they only use objective TA and rarely have a losing day??? You state you can not use TA by itself succefully. What's the disconnect here? Thanks. Surf
Ok, thanks. Care to share your discovery? Surely a money management angle could be shared without diminishing its value. surf
My view is that handling 56 set elements divied into sub sets, has such a low latency that it is not comparable in the manners in which you have chosen. Further, there is the matter of "lock in" of the "failsafe" protection used in trading. Statisitically significant data elements all lock in before the end of the data parcel.
I don't know how NoDoji, Cornix and others define "Objective TA" (e.g. buy if today's volume > yesterday's volume). Yet, I do know there's also "subjective TA" (e.g. trendlines). Yet, I have seen them say they use "pure TA" as if to imply they're using nothing else. In contradiction, I have seen them say in other threads how important money management, discipline, trade management, trading experience is to them. Thus, they obviously aren't using TA all by itself in their trading plan. I say that "discretionary traders" can not profitably use TA alone. In contrast, if it was automated...they could then say its "pure TA" and nothing else. Simply, I don't believe anyone that's a "discretionary trader" saying they only use "objective TA" especially when they are already on record saying there are other components in their trading plan that's important to them. Too many problems with the semantics of these debates. Folks don't define what they call TA, pure TA, objective TA, subjective TA, discretionary, automation and then if they do define such...others don't agree with the definition. That's probably why these TA debates have continued for +20 years and will continue for as long as there are markets due to the fact most can't agree on the definitions. As to the issue of "rarely" having a losing trading day. TA alone by itself won't accomplish such. Most of such will be due to the experience of the trader and a great trading plan with all the components working well together along with some luck.
Alright, but no different in uselessness or usefulness? P. S. That's how I tested TA vs. random entries. It's easy for me cause I already have a working TA approach. I simply took all the same trade management tricks and nuances of my TA entries and applied them to random entries. According to your and Dr. Maestro's hypothesis result should be the same as success is to be attributed to trade management and not the entry. But results were MUCH different. Which tells me there's still some difference and entries do matter.
Actually I don't disprove TA, I trade myself based on TA patterns and absolutely agree with you successful TA patterns are slightly different in application than what's taught in books. If it was that simple as see a douple top and enter then it would be too easy for the money trading offers. Having said that, DTs/DBs and other classic patterns still do work, but within proper context.
For the same simple reason robots still can't substitute humans in most professions: too many inputs. Some day I believe they will do it.
The Predictors did not start making money til they started understanding the markets and trading the short term patterns they found. Then they pounded the hell out of them til they stopped working. I knew someone who was there. One of the more interesting minds in trading at the time. Patterns found via algo or data mining or patterns found by looking at a chart... its all t/a.