The Best TA Book Ever Published!!

[PoundTheRock]

Quote from marketsurfer:

i own the book.

i will ask again--does he DEFINE the patterns?

surf

No, he doesn't. And pattern recognition is a matter of programming, and as we know software is a precise set of instructions. Several years ago, a couple of guys sold a program called PatternSmasher that recognized head-and-shoulder formations, cup-and-handles, etc. I was able to talk to one of the developers and unfortunately their recognition algorithm was brute force and based on pivot strength. So, what defines the base of the cup, and what is the pivot strength of the low that forms the handle? These patterns are non-trivial to program and have many variations, so a scanner is only able to recognize a certain subset of all possible patterns. Further, do these patterns exhibit any consistency across time frames? While I appreciate Bulkowski's work, I would take his statistics with a grain of salt.

So yes surf, by not defining the patterns algorithmically, Bulkowski left out the key to the map and any possibility of objective testing.

 

[marketsurfer]

Quote from PoundTheRock:

No, he doesn't. And pattern recognition is a matter of programming, and as we know software is a precise set of instructions. Several years ago, a couple of guys sold a program called PatternSmasher that recognized head-and-shoulder formations, cup-and-handles, etc. I was able to talk to one of the developers and unfortunately their recognition algorithm was brute force and based on pivot strength. So, what defines the base of the cup, and what is the pivot strength of the low that forms the handle? These patterns are non-trivial to program and have many variations, so a scanner is only able to recognize a certain subset of all possible patterns. Further, do these patterns exhibit any consistency across time frames? While I appreciate Bulkowski's work, I would take his statistics with a grain of salt.

So yes surf, by not defining the patterns algorithmically, Bulkowski left out the key to the map and any possibility of objective testing.

exactly. thank you, pound.

regards,

surf

 

[HolyGrail]

Just because something can't be computer tested, it certainly can be manually tested, and that is exactly what he did.

 

[HolyGrail]

I'll tell you what Market Surfer. Let's play a game. We both start out with 50,000 paper accounts. You trade based on what you learned from "the greatest ta book ever published" and I will trade nothing but patterns in Bulkowski’s book that are highly suspect because they are not evidence based.

I guess it will take some time to develop your evidence-based system so PM when you have something.

 

[ACM Trader]

I think it is bed time for the children !!!

 

[JimmyJam]

Quote from HolyGrail:

Just because something can't be computer tested, it certainly can be manually tested, and that is exactly what he did.

My point exactly.

The patterns can be manually tested, traded and proven.

You've promoted a book as the best technical analysis book ever, written by a guy who by his own statement, does not have proven success as a trader.

What's the point?

Besides selling a book, that is?

And please, don't go into the abstract, keep it concrete.

Give us concrete examples of how this book impacts a traders ability to increase their profitability in the markets.

This request has already been made, BTW.

Regards,

JJ

edit: ... and I don't think, under the circumstances, that this is an unreasonable request or by making it I (or anyone else) is being disrespectful in any way, shape or form, to either yourself or the author, surfer.

 

[Thunderdog]

Quote from JimmyJam:

...Give us concrete examples of how this book impacts a traders ability to increase their profitability in the markets.

I may be speaking out of turn here, especially since I have not read the book. But to the extent possible, a trader armed with statistical tools could properly evaluate testable systems and methods to determine their viability. This would go a long way in definitively exposing people like Larry Williams for the charlatans that they are, provided that they offer testable systems and methods. In these circumstances, it might not necessarily increase the trader's profitability per se, but it could certainly curtail unnecessary losses. That has to count for something.

 

[HolyGrail]

Thunderdog I agree, but by the authors own admission nothing really came out of this.

What I was really trying to offer in the book is first to make the case against subjective TA and to make the case for objective TA and then evaluating those results with a specific set of statistical tools.
I would have loved to be able to point to one of the rules and say --- gold!. It was not to be. So a third point is that it is very hard to find real gold and much of what is offered in other books is either subjective and thus can't be tested or is objective but has not been scrutinized so that we know if it is good or not.

 

[steve46]

Quote from JimmyJam:

My point exactly.

The patterns can be manually tested, traded and proven.

You've promoted a book as the best technical analysis book ever, written by a guy who by his own statement, does not have proven success as a trader.

What's the point?

Besides selling a book, that is?

And please, don't go into the abstract, keep it concrete.

Give us concrete examples of how this book impacts a traders ability to increase their profitability in the markets.

This request has already been made, BTW.

Regards,

JJ

edit: ... and I don't think, under the circumstances, that this is an unreasonable request or by making it I (or anyone else) is being disrespectful in any way, shape or form, to either yourself or the author, surfer.

Hello JJ;

If you are waiting for Marketsurfer to reply, you may want to bring a lantern and your lunch.

The simplest answer is as follows;

First, manual testing is fine. Problem is logisticis. It is difficult to manually test enough data to obtain a significant sample size. So what is likely to happen if you test manually is that the result may deceive you into thinking you have a profitable edge, when you really don't. The issue of "bias" is also important. If you test without screening out the bias in the work, the result isn't going to be viable.

Second, I would bet that without reading the book, no one here has a clue as to what White's "Reality Check" process is, even though it has been around a while now. The benefit of the first part of the book is that it puts a trader on the right track as far as completing useful research. Something they can actually make a buck with. If you compare the result of properly done research with the crap that comes out of most Tradestation reports, then you would see what I mean. Problem is (as I have said before) most folks here don't have the skills to properly evaluate the results. So they just take whatever they get and run with it.

Really what the book does is point a finger in the right direction and bring people up to date as to how research should be done. Its up to you to decide whether you want to keep on the way you were, or try to upgrade your performance.

Good luck
Steve

 

[JimmyJam]

Quote from Thunderdog:

I may be speaking out of turn here, especially since I have not read the book. But to the extent possible, a trader armed with statistical tools could properly evaluate testable systems and methods to determine their viability. This would go a long way in definitively exposing people like Larry Williams for the charlatans that they are, provided that they offer testable systems and methods. In these circumstances, it might not necessarily increase the trader's profitability per se, but it could certainly curtail unnecessary losses. That's got to count for something.

I couldn't agree with you more T-Dog.

But it is a given that traders are going to keep accurate records of their performance/methodology so that they can constantly imporve their returns / minimize their drawdowns.

But you definitely don't need this book to do that. You just need to keep good records so you can run your analysis on them, and you don't need an MBA to do that, either.

Just because a trader did not succeed at trading using a sub par system and then decides to write a book that says TA doesn't work, attempting to use statistics to prove his erroneous point(s) which are PREDICATED on erroneous assumptions to begin with (as ProfLogic has pointed out) doesn't make it so.

Regards,

JJ