The only flaw is in your brain that can't distinguish between the probability of your decision method and the historical returns of a market. In this case the SP500. If you haven't got it yet, I don't care.
Breakeven in expectation over the very long run if you had no costs. If you go bust in the short run, then the long run does not apply, does it?
If you go bust in the short run, you are still bust in the long run and you don't beat 95% of traders. Expectancy is breakeven minus costs only when the probability of going bust is vanishingly small. You need a large trading account in addition to your big pile of coins.
i have a system like that - i think you guys would love it. you will struggle to accept its possibilities, maybe.
... actually the market's expectancy isn't zero; check the options' model ... somebody took the Nobel for it ... and no, his name doesn't start with "O" ... in long run your coin toss approach will loose more than you've estimated, function of interest rates ... the market isn't random, and you can't approximate it with a coin toss even for a large number of samples ... I guess you knew all of these and just wanted to have fun on ET's crowd expense