purely random numbers do not create discernable anything
the graph looks like noise
the graph looks like noise
Why do I see "Trends" in Randomly Generated Data?
- Thread starter Rahula
- Start date
Disagree. Can you post a random plot of any type of data to demonstrate this? And make sure to compound the values over time (for suitable comparison to price charts).
There is a reason why many academics model the markets as Brownian motion (granted it is not 100% gaussian random). But for purposes of visual inspection, you can apply any type of TA imaginable and see trends, oscillations, etc.. The only limit is your imagination and creativity.
P.S. here's a uniform random distribution coin toss I posted earlier (with zero state thrown in by request). Honestly, tell me if you can't intuit TA patterns in the results.
And that is a very simple type of random distribution with no cherry picking.
Coin flips go up and down. Prices go up and down. Ergo, to some, they "look the same."Quote from dtrader98:
Disagree. Can you post a random plot of any type of data to demonstrate this? And make sure to compound the values over time (for suitable comparison to price charts).
There is a reason why many academics model the markets as Brownian motion (granted it is not 100% gaussian random). But for purposes of visual inspection, you can apply any type of TA imaginable and see trends, oscillations, etc.. The only limit is your imagination and creativity.
P.S. here's a uniform random distribution coin toss I posted earlier (with zero state thrown in by request). Honestly, tell me if you can't intuit TA patterns in the results.
And that is a very simple type of random distribution with no cherry picking.
Aside from the fact that they really don't look alike (prices don't go up and down in equal increments like coin flip counts), some of us KNOW that prices are not randomly distributed. Perform a runs test on your coin test and see if you get the same emphatically-not-random results that I got for the S&P 500 over a ten-year period.
Hello guys,
really interesting topic! Bravo! From one side it looks logical the TA to work well, but from another side the random data really look as a stock chart. Amazing. I tryed to get some returns from TA since 4,5 years, studied a lot TA and the only thing I get is an average results
. The advices which I get from some gurus is: continue, do the effors and this will work. Starge. For me the TA wooks as just a random data. Can you make return out of that?
Anyway guys, I'd propose you some of you who is a guru in traging, to post 10-20 day trading deals, IN ADVANCE (before the actual trading day) and after this day we will see whether the results are random or they are OK. This can show the way.
So, is there a trader guru, who will catch up?
Cheers!.
really interesting topic! Bravo! From one side it looks logical the TA to work well, but from another side the random data really look as a stock chart. Amazing. I tryed to get some returns from TA since 4,5 years, studied a lot TA and the only thing I get is an average results
. The advices which I get from some gurus is: continue, do the effors and this will work. Starge. For me the TA wooks as just a random data. Can you make return out of that?Anyway guys, I'd propose you some of you who is a guru in traging, to post 10-20 day trading deals, IN ADVANCE (before the actual trading day) and after this day we will see whether the results are random or they are OK. This can show the way.
So, is there a trader guru, who will catch up?
Cheers!.
Are the markets composed of randomly generated data? No.
These are the results of my generated trades for the last 7 trading days in the ES (3/11 till3/20):
Number of days:7
Number of trades:11
Profitable trades:11
Avg points per trade:13.20
Avg trades per day:1.57
Do I post signals in advance: NO, the reason can be found in the following thread:
http://elitetrader.com/vb/showthread.php?s=&threadid=53761&highlight=spikey
People started to bash me even before I was able to start the thread; some negative posting were deleted.
I sometimes post signals by PM, only to people who appreciate it. Most people on ET donât.
Is it useful to post signals in real time? No, because bashers will always find reasons why your results are not representative.
Discussing this topic is a waste of energy, and most successful traders donât feel the need to convince others. On top of that daytrading and posting the signals within the minute is sometimes difficult and might lead to errors in trading because you have to do 2 things at the same time. I prefer to be concentrated on the fills of my orders and the pricing than on typing in my signals.
These are the results of my generated trades for the last 7 trading days in the ES (3/11 till3/20):
Number of days:7
Number of trades:11
Profitable trades:11
Avg points per trade:13.20
Avg trades per day:1.57
Do I post signals in advance: NO, the reason can be found in the following thread:
http://elitetrader.com/vb/showthread.php?s=&threadid=53761&highlight=spikey
People started to bash me even before I was able to start the thread; some negative posting were deleted.
I sometimes post signals by PM, only to people who appreciate it. Most people on ET donât.
Is it useful to post signals in real time? No, because bashers will always find reasons why your results are not representative.
Discussing this topic is a waste of energy, and most successful traders donât feel the need to convince others. On top of that daytrading and posting the signals within the minute is sometimes difficult and might lead to errors in trading because you have to do 2 things at the same time. I prefer to be concentrated on the fills of my orders and the pricing than on typing in my signals.
Discussing this topic is NOT a waste of energy. Additionally, this is not a thread in which one side needs to be convinced by the other...if this thread were based on opinion then I'd have to agree with you, but it's not. It's based on fact.
Concerning your initial point: yes, it's given that markets are not in themselves random, but they sure as hell act as though they are. Only in the long run is there an apparent edge by going long as opposed to short.
Second point: you can't expect anyone who's got the slightest knowledge of statistics to believe that your trading results are statistically significant given 10 days of data.
You reaction is exactly what i expected from the negativists.
I only posted 7 days of data not 10. I don't want to post more, but i daytrade for more than 13 years in futures and 5 years in forex. have been registered as CTA , and managed 2 funds. I have thousands of trades that i can show that will confirm +80% profitable trades, and the fact that the last 10 years i never had a losing month. But i don't feel the need to discuss that because your reaction is exactly what i described before: no matter what proof you give, bashers will always find excuses to deny real trades as statistically significant.
i have no reason to doubt or bash your claims. however, you must realize, your results are an abberation, picasso like performance that is likely due to some type of intuitive abilility, tape reading, etc that you developed within yourself and not objective TA--- ---with that said, is your trading automated?
there is nothing random about you needing groceries or water or heat or a roof over your head or transportation or an occasional dentist or doctor or new laundry,those are all things we need every day, there is a market for them and it fluctuates,,the weather is fluctuating ,not random, the amount of daylight fluctuates thru out the year ,every thing that happens is confined to our planet and its finite,without getting philosophical and trying to prove this ,i just except it, and the easiest way to see it(,becuase when we define it its easier to accept)is with charts, it would be hard to trade off a 20 yr chart so we narrow it down and down and down to these days tick charts, and react to them as we would a couple of drops of rain