Here is my thesis:
1. Because of the upward bias of the stock market, day trading should have a slightly better than 50/50 chance of making a winning trade. In fact I tried this back test sometime ago: Took the daily historical prices of SPY, bought at open and sold at close every day. At the end of each several 10 year periods I tested (including 2008-09), ignoring commissions, I was actually profitable! However, the odds of positive outcome was only slightly better than 50%.
2. Then why would 90%+ of traders lose money? I think most of them got wipe out by unfortunate random events of huge drawdowns which statistically is unavoidable and trading sizes.
3. So, I am beginning to appreciate those that expounded keeping their trading amount small (1%-2%) as risk management to reduce the risk of a wipe out.
4. The other day, in the post on Kelly Criterion, I finally found the mathematical support of those that said one should trade small: For a "coin toss" that has a ~ 51% winning (about what I found in my SPY daily back test), the optimal amount of each "toss/trade" should be = 0.51 - 0.49 or 0.02 of capital => no more than 2% of capitals should be placed on each trade!
So, it is not luck but mathematics.