Probability of success due to chance alone?

Quote from logic_man:

One way you can do a quick check of your results against chance is to take all of your trades and calculate the average gain and the standard deviation, then divide the gain by the deviation and multiply by the square root of the number of trades you've made or by 10, whichever is smaller. If the number you get is greater than 1.65, you're profitable beyond some level of chance. Otherwise, it's probably just noise.

Van Tharp calls that "system quality number" or something similar. According to his telling of it, the best traders in the world have a number near 5 with hundreds or thousands of trades under their belts.
More...

I sincerely thank you! You bring great happiness to my evening with new information to learn. I will begin to look up all that you tell me (b/c I don't know how to calculate std dev of my trades - I guess I will try excel).

I will also look up van Tharp and system quality number and make a decision for myself whether all this is worth a lot, a little, or nothing at all. It seems like it will help at least a little bit in analyzing my trading performance.

I will report number here. But as I'm on vacation in Boston and still have to learn how to make these calculations, it might be a few days before I can report it.
 
Quote from intradaybill:

hmmm....I flip a coin every 1 min and if it shows heads I buy 1 ES with 2 points target and 2 points stop.

At the end of the day I probably have a few hundred - I don't know how many - trades with an average win of 2 points and standard deviation close or equal to zero.

My "system quality number" is a huge number.

Clap...clap...clap...
More...

Knock yourself out.

And, actually, I said "average gain" not "average win". If the trades split 50/50 (and I don't know why they wouldn't. Just because you flip a coin and get heads doesn't mean your trade will be a winner), your average gain is zero and your standard deviation is 2.

I realize after only seeing a couple of your posts that you have a compulsive need to be smarter than everyone, but you might slow it down a bit and r-e-a-d the posts before trying to prove it. I was only trying to give the OP something a bit more to go on in his quest. Don't have a conniption over it.
 
Quote from anesthesiaman:

I sincerely thank you! You bring great happiness to my evening with new information to learn. I will begin to look up all that you tell me (b/c I don't know how to calculate std dev of my trades - I guess I will try excel).

I will also look up van Tharp and system quality number and make a decision for myself whether all this is worth a lot, a little, or nothing at all. It seems like it will help at least a little bit in analyzing my trading performance.

I will report number here. But as I'm on vacation in Boston and still have to learn how to make these calculations, it might be a few days before I can report it.
More...

The easiest way to do it in Excel is to put the % gains and loss on each trade into a column and use the "average" and "stdev" functions on that column.

It's just a quick and dirty figure that makes some assumptions about normality of the distribution, but I wouldn't worry about that for the moment.
 
Quote from logic_man:

Knock yourself out.

And, actually, I said "average gain" not "average win". If the trades split 50/50 (and I don't know why they wouldn't. Just because you flip a coin and get heads doesn't mean your trade will be a winner), your average gain is zero and your standard deviation is 2.

I realize after only seeing a couple of your posts that you have a compulsive need to be smarter than everyone, but you might slow it down a bit and r-e-a-d the posts before trying to prove it. I was only trying to give the OP something a bit more to go on in his quest. Don't have a conniption over it.
More...

hey idiot man, you now say "average trade" you retard. Average gain is the average of winners you shi* head. Let me help you:

Average winner (gain) = average of winners

Average loser (loss) = average of losers

Average trade = average of trades (Avg. Profit/Loss in Amibroker, Avg. profit per Trade in Metastock)

Get your terminology correct while I put you on ignore sh*thead.

and yes you shi*head, it is possible that someone enters 100 random trades using a random generator and wins in all of them 2 points you dumb as* while your magic metric shows it is a non-random system. Think before you spew crap.

You deserve no more replies you compuslive responder.
 
Quote from logic_man:

One way you can do a quick check of your results against chance is to take all of your trades and calculate the average gain and the standard deviation, then divide the gain by the deviation and multiply by the square root of the number of trades you've made or by 10, whichever is smaller. If the number you get is greater than 1.65, you're profitable beyond some level of chance. Otherwise, it's probably just noise.

Van Tharp calls that "system quality number" or something similar. According to his telling of it, the best traders in the world have a number near 5 with hundreds or thousands of trades under their belts.
More...



In which book did you find this ?
 
Quote from intradaybill:



Get your terminology correct while I put you on ignore sh*thead.

and yes you shi*head, it is possible that someone enters 100 random trades using a random generator and wins in all of them 2 points you dumb as* while your magic metric shows it is a non-random system. Think before you spew crap.

You deserve no more replies you compuslive responder.
More...

If you put me on ignore, that will only improve my experience here, I'm sure.

Yes, it's possible that someone wins every trade with a random trade generator and it's also possible that Gisele will leave Tom Brady and come to live with me. Both are about equally likely.

I didn't say the metric was "magic" and, in fact, it has limitations like not accounting for maxDD, but, seeing as how it's basically the same as a t-test in statistics, which has been around for quite a while and somehow managed to become a standard part of the statistics curriculum, it must have something going for it.
 
Quote from Peternam:

In which book did you find this ?
More...

I was online looking for system metrics and found it, not in one of his books, although I'm sure it's in one of them.
 
Quote from logic_man:

Yes, it's possible that someone wins every trade with a random trade generator and it's also possible that Gisele will leave Tom Brady and come to live with me. Both are about equally likely.
More...

False analogy you loser. You have no understanding of probability. Gisele will never leave Brady for an idiot like you. This is an impossible event. You cannot distinguish probability from experiments in probability. There is no possible experiment that can produce the event {Gisele leaves Brady for an idiot loser trader who posts in ET}. Probaility = 0, impossible event.

On the contrary, there is a well-defined experiment with a finite probability of producing a sequense of 100 random winning trades of 2 points each in ES such that your metric will have infinite value you compulsive idiot.

Total failures like you do not know the difference between an experiment in probability and the probability of a random variable in an experiment.
 
Quote from logic_man:

Van Tharp calls that "system quality number" or something similar. According to his telling of it, the best traders in the world have a number near 5 with hundreds or thousands of trades under their belts.
More...

This is the guy that tried to get a trademark for "position sizing" and "system quality number". Here it is for your laughs:

http://www.tradingblox.com/forum/viewtopic.php?p=32570

His methods are for losers.
 
You must log in or register to reply here.
Back
Top