Becoming better at X

[ThingyMajiggy]

Skc what you have listed above is what I was, probably unsuccessful, in getting at in bonkers "TA which pattern to trade" thread.

If you want to trade short term, imo, there is no "pattern" to learn. There is something far more dynamic going on because you are dealing with an uncertain market that has constant change in not just direction/trend but volatility, time, speed, volume and even where the action is.

Do you feel like what you are doing once you take a look at the progress over the years is actually quite different every say 6-12 months? But at the same time the process that leads to the change is actually the same?

Can you guys explain more on the dynamic thing? Not any "pattern" to learn etc. What do you mean by that? What are we supposed to be looking for/doing? How can you be consistent if it's ever changing, do you mean keeping track of avg volatility/volume, how fast it moves on a continuing basis or?

Some great posts here

 

[avion]

Great thread! I found the inspiration for my signature...:thankyou:

 

[Trembling Hand]

Can you guys explain more on the dynamic thing? Not any "pattern" to learn etc. What do you mean by that? What are we supposed to be looking for/doing? How can you be consistent if it's ever changing, do you mean keeping track of avg volatility/volume, how fast it moves on a continuing basis or?

Some great posts here

My low intellect I fink is making it hard to get across what I'm trying to say.:freak3: So I'll let someone with actual qualifications and experience explain it with like real science stuff,

Improving Trading Performance by Building Fluid Intelligence
An important distinction emphasized by cognitive psychologists is crystallized vs. fluid intelligence. Crystallized intelligence represents what we know, our storehouse of knowledge. Fluid intelligence represents our "how to" ability: our capacity for reasoning and dealing with new and unfamiliar problems. Interestingly, what improves crystallized intelligence does not necessarily improve fluid intelligence and may even interfere with our fluid capacities. Our ability to solve problems in unique and unfamiliar areas is not advanced by having accumulated more information, but creative activities--such as performance in the arts--may build our ability to handle new situations creatively. Physical activity also appears to be linked to improvements in aspects of fluid intelligence.

Discretionary trading is a great example of an activity that requires fluid intelligence. We can read and memorize information about markets until we have an encyclopedic database, but that does not cultivate the skill to reason and function in uncertain and rapidly changing markets. Pattern recognition is an important aspect of fluid intelligence: the ability to understand new situations based upon patterns experienced in similar situations.

My bolds.

A follow up post from the above,

How Can We Make Ourselves More Intelligent?
The recent post suggested that trading ability may be more linked to intelligence than is ordinarily acknowledged. That is because intelligence embraces both crystallized knowledge (what one knows) and fluid abilities (what one can do with what one knows). It is the latter, fluid abilities--not the crystallized information that tends to be tapped by achievement tests--that are most like trading, where reasoning in fast-moving situations is critical.

There are two big implications of this research. First, it may well be that the right kind of deliberate practice in simulation mode not only teaches trading patterns and skills, but acts as a fluid intelligence booster. By requiring traders to reason quickly in moving markets and providing them with immediate feedback regarding their decisions, simulated trading could be good brain training. Game playing could be great training for trading.

A second implication is more profound. Perhaps traditional trading psychology has had it all wrong. What if it's not lapses in discipline and emotional interference that lead to bad trading, but rather failures of fluid intelligence? If traders do not hold the right information in working memory and can't juggle that information in real time to make quick decisions, that could easily lead to frustration and other emotional interference. The cause of the trading problems, however, would be a lack of fluid thinking, not the consequent emotional reactions themselves. If that is the case, perhaps what losing traders need is brain training and not sessions with the shrink!

My bolds. Sorry about the red highlights but really please!!! That is it in a nutshell. It won't matter how many thousand static charts you know if you haven't got the skill to process unusual and fast moving information as it unfolds. My words were "If you want to trade short term, imo, there is no "pattern" to learn. There is something far more dynamic going on ", apparently thats called Fluid intelligence.

 

[CanOz]

I'm think I'm getting this now and I can relate to it. I can't recall when exactly it happened but it was more sudden than a long transition. It seemed like a matter of a couple of weeks between when I was trading what I thought could happen based on previous day's levels, channels etc., to what I could feel was happening in front of my eyes. An example is after a big sell off where I used to get killed fading the initial impulse move, later I had the patience to wait until I recognised the short covering along with new shorts that would later get squeezed out...this leaves a pattern behind which I called the 'hockey stick'. The pattern needs to be recognised as its forming to be very useful...the pattern is not just a shape but a collection of transactions that represent a certain behaviour by a percentage of the market participants...you need to recognise what's happening and who it's happening to to profit from it...in Real time.

This was the biggest ahhh ha moment for me after risk/reward.

CanOz

 

[McLovin]

I'm think I'm getting this now and I can relate to it. I can't recall when exactly it happened but it was more sudden than a long transition. It seemed like a matter of a couple of weeks between when I was trading what I thought could happen based on previous day's levels, channels etc., to what I could feel was happening in front of my eyes.

Doesn't knowledge come in spurts and then plateau for long periods followed by another spurt? That's how it works for me anyway. I have a big break through where everything is sharp and clear and that leads to new questions which build on the previous knowledge (hopefully!) with a lot of cognitive dissonance in between.

Self doubt is also an incredibly useful force.

 

[CanOz]

Yeah, I would agree with that for me as well McLovin. I spend some time in the DZ OR DISCOMFORT ZONE before moving onto the CZ OR COMFORT ZONE where I can use the new knowledge but learn little new knowledge and then once again moving into the DZ and learning again...sprinkled with a few moments in the PZ

 

[skc]

Skc what you have listed above is what I was, probably unsuccessful, in getting at in bonkers "TA which pattern to trade" thread.

If you want to trade short term, imo, there is no "pattern" to learn. There is something far more dynamic going on because you are dealing with an uncertain market that has constant change in not just direction/trend but volatility, time, speed, volume and even where the action is.

Do you feel like what you are doing once you take a look at the progress over the years is actually quite different every say 6-12 months? But at the same time the process that leads to the change is actually the same?

Yes. And I think that's basically the problem to most books and training. They are not dynamic enough. It's like a book of set plays in basketball - it is useful and uselss at the same time with respect to making me a good basketball coach.

apparently thats called Fluid intelligence.

Very apt terminology.

Doesn't knowledge come in spurts and then plateau for long periods followed by another spurt? That's how it works for me anyway. I have a big break through where everything is sharp and clear and that leads to new questions which build on the previous knowledge (hopefully!) with a lot of cognitive dissonance in between.

Self doubt is also an incredibly useful force.

As far as fluid intelligence is concerned, I think it's more continuous. Like driving in a new area... you are constantly looking, steeling, adjusting the accelorator and brakes etc. The "spurts and plateau" would be more akin to studying a new map area before going on a trip. It is a somewhat different process imo.

 

[So_Cynical]

The simple fact is that HFT companys spend 10's of millions even 100's of millions to position servers and write algorithms to get the tiniest of an advantage over the other HFT companys...they do this to get a clear measurable (speed) edge.

So if it was possible to obtain a consistent measurable edge by using patterns or anything else that could take advantage of the masses of data that is available, then i imagine someone would of spent x millions and done it...but of course no one has.

One could argue that no one has done it because the data is useless, making any study of data equally as useless.

 

[McLovin]

As far as fluid intelligence is concerned, I think it's more continuous. Like driving in a new area... you are constantly looking, steeling, adjusting the accelorator and brakes etc. The "spurts and plateau" would be more akin to studying a new map area before going on a trip. It is a somewhat different process imo.

Yeah, I think you're right. I can see the distinction.

 

[CanOz]

The simple fact is that HFT companys spend 10's of millions even 100's of millions to position servers and write algorithms to get the tiniest of an advantage over the other HFT companys...they do this to get a clear measurable (speed) edge.

So if it was possible to obtain a consistent measurable edge by using patterns or anything else that could take advantage of the masses of data that is available, then i imagine someone would of spent x millions and done it...but of course no one has.

One could argue that no one has done it because the data is useless, making any study of data equally as useless.

I fail to See the connection here SC

 

[lesm]

I'm think I'm getting this now and I can relate to it. I can't recall when exactly it happened but it was more sudden than a long transition. It seemed like a matter of a couple of weeks between when I was trading what I thought could happen based on previous day's levels, channels etc., to what I could feel was happening in front of my eyes. An example is after a big sell off where I used to get killed fading the initial impulse move, later I had the patience to wait until I recognised the short covering along with new shorts that would later get squeezed out...this leaves a pattern behind which I called the 'hockey stick'. The pattern needs to be recognised as its forming to be very useful...the pattern is not just a shape but a collection of transactions that represent a certain behaviour by a percentage of the market participants...you need to recognise what's happening and who it's happening to to profit from it...in Real time.

This was the biggest ahhh ha moment for me after risk/reward.

CanOz

Dance with the market. Learn that and you will have another ahhh ha moment.

You have now discovered why lagging patterns are what they are and you need to see them starting to form rather than waiting for them to form. Much too late in a dynamic environment, even worse in a fast moving situation.

 

[DeepState]

Yeah, I think you're right. I can see the distinction.

I see it this way. In order to acquire useful, breakthrough, knowledge and ability, you need to acquire a disparate range of things and put it together before something meaningfully new and better is actually created to add to your existing skillset. Up until the last useful piece is added, not much goes on. Learning to ride a bike requires a bunch of things to work. It isn't until the last bit happens/clicks that you manage to ride off without scraping your knees by the third pedal. Up until then, you aren't really riding. It is a breakthough because several component skills came together and combined to produce something special. Plateaus and spurts. Generally, the more bits that need to combine, the longer the plateau. So it is with learning something useful or learning a complex skill. These things do not move in straight lines, although it might be argued that, once somewhat mastered, further skill acquisition is more proportionate to outcome than some big "Ah-huh" moment.

Fluid intel is like situational awareness. Some people just see things and relationships much more quickly. There are many ways that high performers do this and I think the research points to the skills not being transferable across domains. You can be a genius in a fighter jet whizzing all over the place and have no idea whatsoever about Kasperov's deep strategy. Both require a lot of fluid intel.

Crystal is about having a library to hand. It's value degrades by the number of books on loan and not returned.

 

[ThingyMajiggy]

So basically we're talking about having the feel for the market based on experience and being able to quickly adapt and make decisions based on what you've seen before, and that pretty much practice makes perfect?

So I'll keep watching market replays on 10x speed then, sounds like the way forward, get my mind used to and trained in what to see, like speed reading, get with the flow.

That's really all it is? Just a certain set of skills that differentiates the big boys from the failures. Seems too simple to be true.

 

[CanOz]

Dance with the market. Learn that and you will have another ahhh ha moment.

You have now discovered why lagging patterns are what they are and you need to see them starting to form rather than waiting for them to form. Much too late in a dynamic environment, even worse in a fast moving situation.

That's the great thing about the DOM, it only lags as much as your latency!

 

[So_Cynical]

I fail to See the connection here SC

Connection is that if data/patterns were a predictor of anything an algorithm could be written to take advantage of that...since no such algorithm exists we can assume with great confidence that there is nothing to be learned from data/patterns.

 

[Trembling Hand]

Connection is that if data/patterns were a predictor of anything an algorithm could be written to take advantage of that...since no such algorithm exists we can assume with great confidence that there is nothing to be learned from data/patterns.

You cannot be serious?

For more than twenty years, Simons' Renaissance Technologies hedge fund, which trades in markets around the world, has employed complex mathematical models to analyze and execute trades, many of them automated. Renaissance uses computer-based models to predict price changes in easily-traded financial instruments. These models are based on analyzing as much data as can be gathered, then looking for non-random movements to make predictions.[7] Some also attribute Renaissance performance to employing Financial signal processing techniques such as pattern recognition.

Renaissance employs staff with non-financial backgrounds, including mathematicians, physicists, astrophysicists and statisticians. About a third of the 275 employees at the East Setauket office have Ph.Ds. http://en.wikipedia.org/wiki/Renaissance_Technologies

James Harris "Jim" Simons (born 1938) is an American mathematician, hedge fund manager, and philanthropist. Considering his background as a code breaker and an expert in pattern recognition[4] some also consider him as one of the pioneers of financial signal processing.

Simons taught mathematics at Stony Brook University on Long Island in New York.

In 1982, Simons founded Renaissance Technologies, a private hedge fund investment company based in New York with over $15 billion under management. Simons retired at the end of 2009 as CEO of one of the world's most successful hedge fund companies.[5] Simons' net worth is estimated to be $12.5 billion.
http://en.wikipedia.org/wiki/James_Harris_Simons

 

[CanOz]

Connection is that if data/patterns were a predictor of anything an algorithm could be written to take advantage of that...since no such algorithm exists we can assume with great confidence that there is nothing to be learned from data/patterns.

Oh no, here we go again...

 

[DeepState]

1. So basically we're talking about having the feel for the market based on experience and being able to quickly adapt and make decisions based on what you've seen before, and that pretty much practice makes perfect?

So I'll keep watching market replays on 10x speed then, sounds like the way forward, get my mind used to and trained in what to see, like speed reading, get with the flow.

2. That's really all it is? Just a certain set of skills that differentiates the big boys from the failures. Seems too simple to be true.

1. There is no concept of sustained perfection in this business. There really are just a set of ever-changing perfects. Dynamic markets. [If interested, refer to Andrew Lo "Adaptive Markets Hypothesis" for a pretty realistic idea of what actually goes on]. So the game is partially about practicing to become perfect in whatever happens to be working at this time.

However, there is a very large advantage to those who can rapidly identify what will work and skill up or build infrastructure to capture it. Whilst practice does make perfect, learning bloody fast about the implications of change makes you rich (or however you want to define success) in the endeavor of choice. It turns out that the difference in abilities in financial markets has very large consequences. It is somewhat less in dentistry.

I could read a book at 10k words per minute. My comprehension would pretty much be zero. There is more to it than speed, although it is clearly a feature. This concept of fluid intel is a sense of just knowing what to do in a situation. Some call it intuition, some call it guided experience, some call it street-smarts. As I mentioned before, this ability seems situationally dependent. I could practice all day long, watch match replays at 10x speed for years, and I can tell you my chances of making the Wallaby tour are essentially indistinguishable from a hole in the ground. The same could not be said in other fields.


2. Those skills are broadly defined. Many combinations of skills are successful. Some include knowing the kinds of people that help you figure out what's going on. No-one says you have to acquire it yourself. There are other similar concepts. There just happen to be a whole lot more that are not successful.

Then there is environment. An awesome investor in one period can look like an ass in another. Sometimes the skills do not align with the environment. We can go into organizational behavior and other societal matters if you wish.

So it is not as simple as that. But your chances of success are definitely enhanced if you can recognize things a little bit faster or more accurately and capitalize on it. You don't need to be good, only better. We live at the margins here.

 

[DeepState]

You cannot be serious?

Simons is scraping and parsing our conversation right now, classifying it and embedding it in predictive regressions via kernel learning algo. Very formidable. And that stuff is old. All algos are certainly not created equal...and neither are discretionary traders/investors. You know that - partly because you actually are serious.

 

[Weatsop]

Do you like trading? If no, stop trading. You're done.

If yes, then it's no big ask to practice stuff. Stopped out for today? Or the market is bad for how you trade? Hopped on early? JUST STARTING? Then hurray, fire up the paper account and practice something, or pop open historical data and go candle-by-candle.

Practice one thing. For the love of god, practice one thing and ONLY one thing, and know what that one thing is. (Seriously - what's the thing? Write it down!) Maybe one kind of entry, with one kind of exit. Sure, when you trade live you've got a bunch of tools and tricks, and use them all. But when you intentionally practice, have ONE thing to focus on. Otherwise, how can you clearly see whether it was working, or whether you're getting better, or if you are PROFITABLE, when it's all mixed up in a bunch of maybes?

Be honest with yourself. Don't cheat. Log the trades, the losses and wins and the commissions and slip. Don't just eyeball old data and tell yourself you'd have made a motza. That's not practice. That might be ok to get some basic ideas, but reality doesn't work like that.

Note when it's going good. Print the charts. Stick 'em on your wall, mark the trades. DO THE SAME WHEN IT'S BAD. Notice anything different? Is the price action looking different? Is there some way to avoid those bad trades, but keep the good ones? Can you see the jaggy stupid wandering crud that tells you to stay the hell out of the market? Can you recognise that? Practice NOT trading that stuff, because seriously, staying out of bad conditions is something you need to do.

And that noting what's good and bad? Do it with the real trades! Real trades are the best practice, but it's messy. It's not just one thing you're practicing, so especially at the time you made the trades, you'll miss some of the details - you'll miss the forest for the trees. So on the next day, get on your computer a couple of hours early. Look over yesterday's trades. What was good? What was bad? Mark all the trades, and what you were thinking when you did them. Does REASON X keep failing? Does EXIT Y keep missing something obvious?

Don't - this is important - don't try to get all the pips. If you could, you'd be the richest person on the earth. You won't get them all. Don't kick yourself for "mistakes" that are only clear in hindsight. Practice, instead, being PROFITABLE, not being RIGHT. Can you introduce a new "rule", under some conditions, that would have made your old trades better? Fine, try that rule - and practice ONLY that rule for a while. See if it works when the bars are moving.

If all of that isn't fun, then quit. There are plenty of jobs on earth. Go do one you like.

There's no perfect trade. Ever. That's what makes this fun. Because there is ALWAYS something to do.