Quote from: defxor on September 03, 2011, 10:12:39 PM

Quote from: the joint on September 03, 2011, 10:04:13 PM

About 90% of my trades have been profitable ones. I've made approx. 30-40 trades on tradehill. Try flipping a coin 30-40 times and see how often you get 90% heads.

Since you're the one claiming it would be out of the ordinary, why don't you do the math

You of course know how to do basic statistical calculations.

(PS: Traders talking about their success rates after the fact, using words like "about" and "approx" don't ever show up in scientific research. Can you guess why that would be?)

Ugh fine.
Trials = 40
40*.9 = 36
What is the chance that I flip AT LEAST 36 heads (at MOST 4 tails) -- That is, if the choices are profiting vs. not profiting on a trade, what is the likelihood that I will profit AT LEAST 36 times?

Total possible outcomes = 2^40 = 1024^4 = 1099511627776
nCr (40,4) = 91,310 possible 4 tail outcomes
nCr (40,3) = 9880 possible 3 tail outcomes
nCr (40,2) = 780 possible 2 tail outcomes
nCr (40,1) = 40 possible 1 tail outcomes
nCr(40,0) = 1 possible 0-tail outcomes
= 102011 possible outcomes or 4 tails or less
(91,310/1099511627776) + (9880/1099511627776) + (780/1099511627776) + (40/1099511627776) + (1/1099511627776) ~ 9.279^-8

.00000009279 = .000009279% chance I will flip at least 36 heads (i.e. trade positively at least 36/40 times)

.000009279 * x = 100
x = 100/.000009279 = 10777023.38614075

Yep, I guess I got really lucky! There was only ABOUT a 1 in 10,777,023 chance that I was as least as successful as I was!

That is because, as you know...I profited purely by chance.

Edit: Note this calculation is an oversimplification of the problem and impossible to truly demonstrate. There are fees, there is a possibility the market will not move at all, or will not move enough to compensate for the fees. Simply put, it is impossible to create a formula for calculating the 'chances' of profiting x number of times after y trades because there are too many (key word) UNKNOWN variables. These unknown variables are what make you think the market is due to chance, but in reality they are simply that -- unknown.

Edit 2: You do realize that the chances of trading at 90% profit become less likely the more times I trade...right? I hope you weren't implying that my chances of doing this across 40 trials were good, were you?