That is not the issue. What you stated is a red hearing. Trading is all about probabilities. If someone has a sample of 1,000 trades and associated win rate, confidence intervals can be calculated for future win rate. This is all you need. Knowing the future has nothing to do with trading.
Statistical Question
- Thread starter Safetrading
- Start date
That is not the issue. What you stated is a red hearing. Trading is all about probabilities. If someone has a sample of 1,000 trades and associated win rate, confidence intervals can be calculated for future win rate. This is all you need. Knowing the future has nothing to do with trading.
If you have a 86.6% win rate on a 1:4 risk/reward setup... then obviously, you have positive expectancy on every trade you place. And that means, you should trade as frequently possible.
Note: the markets are memory-less. No one is tracking the fact that you've won 4 in a row... the statistical probability of you winning the next trade remains exactly the same as if you've lost in a row.
It's no different than if you were playing blackjack. Even if you've lost 100 hands in a row (or won 100 hands in a row), the statistical distribution for possible outcomes of your next hand remains exactly the same.
So, again, *IF* (and that's a huge if) your numbers are correct... you should trade as frequently as possible.
Note: the markets are memory-less. No one is tracking the fact that you've won 4 in a row... the statistical probability of you winning the next trade remains exactly the same as if you've lost in a row.
It's no different than if you were playing blackjack. Even if you've lost 100 hands in a row (or won 100 hands in a row), the statistical distribution for possible outcomes of your next hand remains exactly the same.
So, again, *IF* (and that's a huge if) your numbers are correct... you should trade as frequently as possible.
Unless there is a clear pattern in your trades i would not cherry pick them. Its possible youve found a small nitch for yourself. Trade it pure and see if it really holds.
once you have a live set of statistically significant trades under the belt you can then try to improvise..
my 2 cents
-kcgoogler
once you have a live set of statistically significant trades under the belt you can then try to improvise..
my 2 cents
-kcgoogler
A 86.6% win rate on a 1:4 risk/reward over 100 trades does not necessarily indicate a positive expectancy.
If he has positive expectancy, the statistical probability of him winning the next trade remains exactly whatever the numerical expression of the positive expectancy is.
It depends on what kind of blackjack you are playing. If you are playing *basic strategy* and also effectively counting cards on a shoe less than 6 decks, then you have positive expectancy. Otherwise, you'll have X negative expectancy per hand. If you know how many decks you are dealing with, and you know how effective your counting method is (on top of basic strategy), you can put a numerical value on your positive expectancy. Your next hand in blackjack will have X positive expectancy for Y positive count. The higher the positive count, the higher the numerical expression of your positive expectancy. These numbers are always the same.
That's true only if he has positive expectancy, which is still an unknown. If he doesn't know how to use math, then he can set up a model in a random environment and generate a million pseudo-trades with a 1:4 reward/risk ratio. It will generate a win% rate of, i'm guessing between 80%-90%. If it generates a win% ratio of 80%, then possibly at this point (100 trades) he has a 6.6% positive expectancy. This means that for every $1,000 put at risk, he should be making $66 dollars. If he want's to make $200 a day, then he needs to risk 200/.066 = $3030. To find out how many trades he needs to make per day, he would calculate 3030/average-risk-per-trade. He would need to proceed cautiously because with only 100 trades, he could easily have variance of +- 6.6% that could vanish (go to zero) over a larger set of data.