Topic: TA is pseudoscience - page 2. (Read 5363 times)

I love Bitcoin, which is why I can't stand the hype that makes it volatile and completely useless as a currency. I am involved with Bitcoin on various levels and have an interest in being realistic as opposed to pumping, which is what most people here do, because I do not see a future in such a volatile 'investment'.

Bitcoin is genius and would be the cheapest, fastest way to transmit value if it wasn't so volatile. The high currency exchange risk means that only investors can currently benefit from Bitcoin. If you are truly a Bitcoin aficionado, you should not want it shoot up to $1000 anytime soon.

You sir, amaze me! To actually take the time to register and hang around on a Bitcoin forum just to bash Bitcoin all day long is really impressive. I mean, i'm all stressed out, having so much to do in my life. Can you please give me some advise on how you find so much free time to spend?

oh lol i just realized you call "support and resistance" to the last one...
(btw dont hate my for that draw, is just a joke )

Let's start again:
The first problem with TA is that there is no formal definition of any of the so-called "patterns", that's why it's so common to see people giving apologize for how they misread a chart. Even if something like a formal definition exist for a pattern there is no (real) empirical evidence that could suggest that it have some relevance in the future price.
But never mind... let's say that there are "patterns" and also "evidence", you guys choose a different method for identify a patterns each time you start doing a new analysis, and that is because you are actually using intuition, not real pattern recognition. There are no "patterns" in the market, every single time is a different situation that can't be predicted using past values.



 

I get your point but believe cynical assumption is also a pseudo-science :p

There are those (of which many simply can't analyse or recognise a trend) who believe TA has ZERO worth, this clearly isn't true and I can only speak for myself when I say understanding TA, along with other important personal traits, makes you money.

But I'll make the point again that when you check the price ticker and compare it in your head to what the last price was then you are performing trend analysis. Otherwise you would be gambling blind rather than investing or trading.

A trend needs only 2 data points to extrapolate a third or many more.

Man has achieved a lot with both mathematics and psychology, admittedly more the former than the latter!

I like to leverage that wisdom

My two satoshis:


OP, you are misunderstanding TA on a fundamental level. It is not (always) effective simply due to the self-fulfiling prophecy argument. TA is descriptive; a way of pointing out the obvious. In the same way a poker player knows what an opponent may be holding on a check, call or raise, you can gauge market sentiment and activity.

Take the symmetrical triangle, for instance.



There's lines roughly delineating support points drawn onto the graph. But their only use is to point out to the casual observer what the experienced one can see anyway: that the stock is trading within a range, and that range is narrowing, gradually eliminating swing-trade profit potential; it will only do this when all other speculative or value sentiment is too confused or uncertain to bet either way; and it will continue squeezing tighter, until like a sphere of uranium compressed too tightly it will explode; this particular triangle tells you that you will not know which direction it will explode, without outside knowledge. That is the limit of the information TA can provide you in this case. You now have a known unknown to work with. (and the corollary to that is once the pattern is broken, the immediate direction of the price will tend to be a contrarian indicator, with a dip allowing bulls to buy lower and start a snowball, or the reverse for a rise.)

TA is a way of saying "This is what people are doing right now."

Is it all so complex?

You can't use TA to predict the future, but you can use it to identify a trend or momentum on which you base your decisions to enter or exit a trade.

It's like the up/down arrow on a lift, it doesn't tell you the floor the lift is going to, it simply says what direction its going in, and sometimes how fast. It works a treat in helping you assess the available information.

This is what makes TA worthwhile.

As far as I can tell the efficient market hypothesis holds true.
Not as a absolute, but rather as a factor of randomness in the price data. This can be measured, no I won't tell you how.

i think you've got the idea. if you have a stochastic function like price, a 'pattern' is an autocorrelation, or a segment of the data which 'correlates' in some way to previous segments of data. your perspective from computer vision is spot on.

as for the EMH, how doesn't it? i was under the impression that EMH implies that the price function is a random walk.

Technical Analysis is mostly luck and self fulfilling destiny. I agree with the title of this thread. Most people that succeed are investors and not speculators; so you are making assessments based on fundamentals and market developments.  Successful investors predict the unpredictable.

For anyone interested in TA and the the hustle which is Wall St., read the Black Swan by Nassim Taleb. It's a great book (I'm sure many here have read it) and offers insight into trading psychology from a successful trader and one of the greatest minds of our time. I compare TA to those "How to Beat the House" gambling books in Vegas; interesting info, arguably effective, and distributed by the house for a reason.. and the reason isn't because the house wants you to win.  

The key thing to remember is that because something happened a certain way in the past doesn't mean it will happen that way in the future.

I used to work in computer vision r&d, specifically in real time pattern recognition. If you have enough information about some environment conditions you can tell the probability of a good match. Even when the probability is low you can use several frames and filtering to improve the recognition, since each frame (probably) saves some relation with the pasts frames...
Maybe I'm misunderstanding something: please someone explain what is a "pattern" in TA and what method you use for recognize those patterns.

@arepo: EMH does not imply that there is no way to have a better probability than just plain luck.

(btw please excuse my english today, i've got the flu and i cant think well )

+1 had not heard of the Hurst exponent, but i have multiple times talked about the autocorrelations in price data in defense of TA. now i know how to measure them

I do believe that some basic TA stands on its own, like the formation of resistance levels and stuff like that. It makes sense and it takes into account psychological drives from people, which other disciplines don't.

But I don't think there is that much to it, that's actually real. A lot of the "theory", from what I've read, mixes facts with numerology and plain pseudoscience. A bit like astrology: "these stars are really there, and these trajectories are there too. But they don't mean anywhere as much to you personally as you're implying."

I think all real TA can be explained in a rather small book. The rest is conjecture and history that won't help you predict anything at all any better.

+1

TA Is hard to do but it is a tool in your toolbox of investing

Or the lottery. Any spot on the Roulette table has vastly better odds than Bitcoin, IMO.


I do not believe this.

Might help at some markets, might help a bit at bitcoin too but never will graphs with lines and drawings guess when some huge buy or sell would come and crash/skyrocket the bitcoins.

When you see people panic selling - sell. When you see them panic buying - buy. Do both in perfect time and you're winner

It's not really a very good analogy, except in a superficial sense that markets and weather systems are irregular and highly unpredictable. I'm not even sure if the analogy can be taken further - markets are complex systems, whereas weather systems are just dynamically unstable (chaotic).
For weather forecasting, you can apply serious grunt to numerical models of the atmosphere, and you will get good predictions, though the accuracy will drop dramatically as you forecast further out. You can try the same type of modelling with markets, but it just doesn't work - if it did, then the big investment funds would simply be competing to build the biggest supercomputers to predict the market out beyond the competition. Instead, they buy servers closer and closer to the exchanges, in order to cut down on delays by microseconds, in order to profit from FOREX arbitrage and things like that.
The point is, you *can* predict the weather, with quite high accuracy in fact, if you're talking about whether it will rain in the next hour. However, you can't predict if the price of a market will go up or down in the next hour, even using huge processing power applied to numerical models.
You definitely can't predict anything by scrawling a bunch of arcane, contrived lines all over a historical price chart...

I think what he is saying is that while currently there is momentum in BTC, that could change.  Governments step in, another digital currency takes market etc...  Without a plan, any investment is worth basically nothing because every stock/currency etc does not last forever.

I guess it comes down to your personal TA or gut when to sell your investment... There are people out there that sold at 37 on the way to 266, and those that sold at 50 when they bought in at 266...  Some will buy in at $1 and still think that at $1000 they should still hold because it could become $2k eventually etc etc etc...