The Gambler's Fallacy

Has anybody applied The Gambler's Fallacy to trading?

Briefly what this says is "The gambler's fallacy is a logical fallacy involving the mistaken belief that past events will affect future events when dealing with random activities..." etc.

The idea is that the dice or roulette wheel does not have a memory. So if red came up 10 times in a row on the roulette table then the chance of it coming up red the next time is not affected by the previous results.

I am interested to know how this applies to trading.

quote:

---

Originally posted by day trading

quote:

---

Originally posted by urbansound
...statistically proven fact that humans when given the choice will always pick the "safest" number within the realm of perception. In most cases the number 2 would always be the most common chosen value as it lies "centrally" between 1 and 3, (feels the safest)...

---

So, as a day trader, are we trading against predictable humans that will be drawn to round numbers and other "predictable" areas on a chart? Or are we trading against machines that are void of emotion?

Are there enough humans trading in the market place to cause the machines to fail often enough such that it's more likely that the human predictability will prevail.

I enjoyed your other comments!

---

Hi DT. I enjoyed the question.

I guess all I can say is we still see as much as 90% of early traders crashing up to 3 accounts and an average 50K in finding a hard road to learning the success is not about making money, but about not losing it. LOL So yeah, I think there remains considerable human influence in the market, despite automation.

Most who are programming scripts today are barely qualified as novices and so much of the automation is given of human failure that despite these considerations even the automation that works well, is a model of "au-tom-a-ton" thinking, which simply accelerates human perception and reaction.

The exceptions to the rule have to be at least well oiled hedge funds that digest much of the 90% losses in huge success as well that we not muzzle the ox which pays market makers to take the other side when they will. What do market makers and maker-bots use?.... PIVOT POINTS! (kudos mypivots.com).

Why does Fib work? Because despite winners, losers and bots, the market continues to operate on the basis of "nature". Nature will win out, every time. We've now destroyed 98% of the earth's rain forests and yet we continue to breath when they said we would not. Why? Our worlds greatest provision of O2/CO2 exchange comes from....(drumroll).... MOLD!!. who'd a thunk it?

What I learned was I could clean up in roulette using a double-down Martingale red/black by first watching for tables primed with several red or black consecutive already. $700.00 later I met my first roulette run that went 11 black in row. Nature won again. sigh. I Started out with $20.00 and spent 3 hours at it, using a pure mathematical approach on the odds. It was the human side that didn't know when to stop.

I don't know if all that answers the question or not, but in thirty years of programming I have yet to see machines do much more than accelerate what humans do; especially when it comes to repeating our mistakes. But isn't that nature's way? To err is human?

That reminds me... This is an election year, isn't it? Red... or... Blue... Let me count....

Nice Chat. See ya.

Just under 10 years ago I tried the martingale system on a $1 roulette table in Vegas. I would use the $1 that I "won" on each bet to finance putting a $1 on a single in the middle of the table. My limit was $100, if the martingale took me over that then I would have quit. Luckily that never happened and my single on the table came up after about an hour of playing and I walked away with my $35.

My justification in doing this is that I'd always wanted to play roulette in Vegas and the $100 limit was entertainment money that I expected to spend playing the game with an outside chance of winning. In my case this was pure luck and I never attributed it to skill or think that I could ever pull it off again.

No matter how you slice it, statistically, the sequence of any set of prices is considered random. The previous bar or set of bars can never predict the next bar. This has been proven so many times it would be pointless to even argue the point.

In trading the bars don't just represent price, they represent traders and trader expectations. This is why price formations work. There is psychology in every successful formation. If you cannot rationalize in terms of fear and greed why a formation works, then it probably won't. Flag formations, wedges and triangles, double tops and bottoms. 1-2-3 and ABC formations are all the result of not prices, but traders fear, greed and expectations. Being able to anticipate and quickly react to these expectations is how you win in this game. Forget the actual price, try to read the minds of the traders that bought and sold to bring price to this point.

Prices go down strong, then they rest. Some traders who missed the downturn aren't looking for a continuation, they are looking for a reversal. So they buy enough to turn oscillators in the reversed direction, that brings in a little more buying. But the volume on this reversal is small compared to the volume that drove prices down. What are you going to do, take the reversal or see the bounce as low risk entry to continue with the high volume trend?

quote:

---

Originally posted by Turk

No matter how you slice it, statistically, the sequence of any set of prices is considered random. The previous bar or set of bars can never predict the next bar. This has been proven so many times it would be pointless to even argue the point.

In trading the bars don't just represent price, they represent traders and trader expectations. This is why price formations work. There is psychology in every successful formation. If you cannot rationalize in terms of fear and greed why a formation works, then it probably won't. Flag formations, wedges and triangles, double tops and bottoms. 1-2-3 and ABC formations are all the result of not prices, but traders fear, greed and expectations. Being able to anticipate and quickly react to these expectations is how you win in this game. Forget the actual price, try to read the minds of the traders that bought and sold to bring price to this point.

Prices go down strong, then they rest. Some traders who missed the downturn aren't looking for a continuation, they are looking for a reversal. So they buy enough to turn oscillators in the reversed direction, that brings in a little more buying. But the volume on this reversal is small compared to the volume that drove prices down. What are you going to do, take the reversal or see the bounce as low risk entry to continue with the high volume trend?

---

I told someone the charts are brain waves. If you can read
the chart it will give insight as to what traders are thinking.
For me I gain more understanding when I think of the market as
people not this big "market" that is running wild.

I am agreeing with you Turk.

quote:

---

Originally posted by Turk

No matter how you slice it, statistically, the sequence of any set of prices is considered random. The previous bar or set of bars can never predict the next bar. This has been proven so many times it would be pointless to even argue the point.

---

Unfortunately, the first two statements above, really have little to do with each other. In order to understand why certain statistics cannot work on markets has to do with the lack of adequate range variation between least significant, parent population, contiguous samples or "minimum granularity".

A bar being a 4 hour bar, may well hold more than adequate data to identify the properties anticipated on the next bar. Not from a typical view of statistical analysis, but rather more of an attribute confirmation surrounding confluence of Fib levels, identified Pivot Points and any given slice of time they were found relevant.

If the markets were even CLOSE to something we could call "Random", then Fibonacci would not be capable of locating specific points which correlate reliably with Pivot Points, nor would market celebrations ever repeat as they surely do.

Fear, greed, system or mechanical influence, whichever you'd like to choose, are simply playing on the results of the combined and coincidental moves of Market Makers placing and withdrawing at celebrated levels of known Pivot Points, in order to attract and create motion in the market.

Since the Pivots are consistently evolved from celebration points of the prior day's action, each successive day has a relationship with regard to Pivot Points, from the prior day. It is this consistent model which causes the market to behave relative to the Fibonacci pattern and exactly why those who come to understand these relationships are able to reliably gage "Probabilities" that are relevant in kind and therefore create a measure of "predictability".

Therefore, any bar may have enough data to suggest probable levels that are discernible and predictable on the next bar. It becomes more a question of how you would do that, if you will accomplish it or not.

One also does not need to "predict" the market to profit from it either, which is one of the reason's certain arbitrage / hedge systems have created innumerable wealth for some. ;)

The market is a reflection of human behavior. In many cases this behavior is nothing more than a set of habits, especially so for successful traders and the same can be said of algorithms and black boxes. A successful trader over time becomes so by developing habits and behaviors which lead to success. Conversely, failed traders exhibit habits which are not desirable and lead to failure. We translate these behavioral tendencies of market participants into probabilities. For example, we might anticipate that if a certain Support or Resistance level is broken through, traders will have a specific reaction to that event. Where we get into trouble is making the leap of faith from a probability to a certainty.

The other side of the coin is, there is nothing certain about what the market will do next. There is no way for a human (or computer) to predict what the next trader will do, or the next 100 market orders will be. We cannot predict with any reliable consistency when the next 1000 lot buy order will rip through the offer side of the order book.

So your left with the choice to either anticipate what is likely to happen next and jump in front of the train or wait and react after it starts moving.

I submit the market is random, and at the same time has tendencies which repeat because it is a reflection of human behavior. Unlike dice or a roulette wheel, traders being human do have memory, have the capacity to learn new behaviors leading to greater success and are willing to adapt. A trader that is successful tends to stick with the behaviors which they have learned will give them the reward they desire. Comparing a trader with a pair of dice or a coin toss is an over simplification of the reality of trading and the markets.

UrbanSound, you and I are saying the same thing. Simply using price bars you cannot predict the next bar. The predictability of future bars is in the patterns. Such patterns include S/R, Fibs and such. It is the formation that gives predictive value because from that you are able to see what other traders are thinking and hoping for.