300 is probably too low but 3000 is excessive. Let's say you want to determine your system's winrate. For proportions (like winrate) the error associated with sample size is 1/√n. This is for 95% confidence. So a sample size of 100 is an error in the proportion estimation of ±10%. To cut this error in half (±5%), you need a sample size of 400. The typical survey sample size is a minimum of 625, giving an error of ±4%. Eckhardt suggests a minimum sample size of 1800 trades.
As Handle123 explained testing requires sampling every type of market that will be encountered. Sample size can be reduced to a very small number in cases where you understand how a model works in it's ideal climate and it's worst case climate. When testing is approached from this angle all that time spent on traditional backtesting processes can be diverted to directly looking for intelligent ways to improve the model. This more direct approach is simply ...better.
Unless he's doing hfts (hundreds of trades per day), he's going to encounter every type of market over 1000 or so trades. 3000 is fine as a personal choice but not as a recommendation to newbies who are still trying to get a starting strategy.
Sample size is an overrated and meaningless statistic. What's important is testing the most extreme types of markets both good and bad to determine typical success in best case ideal environment ...and worst case failure in adverse climate. Along with the over importance too many people place on these useless stats there's an overabundance of denial of the importance of understanding why a strategy works. When you focus on understanding why then you know how to improve both the strategic concepts behind the strategy and also how to improve the few stats worth considering. Yet nearly everyone swears they don't need to know why something works, they just need to know that it works. There's such obvious errors in that line of thought ..... it's no wonder all the time wasted on useless stats and hardly any spent on important stats.
You've got to look for the nuggets and ignore the rubbish. This pattern is OK although the stock isn't or the time scale:
A good trader has research skills. Everything you mentioned in your quote above is already discussed in great detail at this forum alone. You just simply need to type into the upper right corner in the "search box" to access those topics you mentioned. Therefore, due to the fact its all said here at this forum...maybe you're not observant enough, maybe these topics really aren't of interest to you...resulting in you only watching them for a few days before you move on to another topic, maybe you're making the classic mistake of trying to find trades or trade signals when you have not learned how to do the analysis first that would allow you to understand the price action prior to looking for trade signals ? My point is this, if you been watching the S&P 500 Emini ES futures for almost a year...have you really been studying it or have you been trying to trade it...the latter is a mistake. Analysis ------> Trade Signal (correct path) The analysis should take many months or a few years prior to doing any real trades. In comparison, its like a real education (college)...you study for many years before actually getting a chance to apply it in real conditions. I'm always amazed by how many traders think they can be a professional trader in a few weeks or few months. Seriously, would you send your kids to a school where the teachers learned how to teach in only a few months, would you let a doctor operate on you if he/she learned how to do it in a few months, would you let your pharmacist prepare your medicine when he/she learned how to do it in a few months...something to think about...right ? Simply, many traders say they been watching or observing but in reality they have not...instead...they've been trying to trade it during that time span in hopes that they'll be able to figure it out as they go. Thus, someone that's truly observing, studying, watching a trading instrument for one year...would not have taken one single trade until they knew what was going on with that particular trading instrument...such may take up to several years or more prior to the first trade. Don't believe the myth that its better to start trading small and then learn as you go...worst mistake ever and most do it and most fail.
Hi wrbtrader.... Great reply! I am exactly THAT GUY....the one who traded on a sim platform, made a bunch of Play Dough on a bunch of BS trades, thought this was easy as pie and then jumped right in! And its surprising, I'm really not that idiotic! Usually I am quite reserved and patient and measured about things....especially about matters of money. So.....this is what has led me here. I know there is a wealth of info here and many MANY very experienced traders with all the advice and insight they can possibly offer. I guess at this point I don't even really KNOW what I don't know... Not at the stage yet where I can even ask intelligent questions about topics.... Plus the fact that on any one post there ends up being an enormous amount of contradictory opinions... I am reading...I am studying and I agree wholeheartedly that this will take a very long time to get a grip of and a lifetime to master, if ever.... Thank you again for your advice and to all who have offered guidance thus far.... On it goes. switchgear666