Thoughts on randomness and systems

Martingale increases the risk of ruin without improving expectancy. I've posted XLSs demonstrating how MM has no effect on expectancy (see posts #31, #32 and #43 in this thread).

If the market is a random walk...... then ALL systems have a statistical expectancy of zero (actually negative, if we consider costs - so, given enough time, every trader would eventually lose). Even if the market has moved to an extreme, the probability of a move either way remains 50/50. To believe that history affects a random event is Gambler's Fallacy (Google it).

If grid systems like Nanningbob's are long term profitable, then they demonstate that price reverts to a mean more often than not (i.e. is not a random walk), and/or the larger bet sizes coincide with greater probability of a mean reversion, all of which creates positive expectancy. By limiting the number of Martingale doubles to the account margin, the possibility of a 'death trade' is eliminated, but then the recovery from a loss (after which one reverts to the original bet size) is crippled proportionally. And, as with any MM, a long enough sequence of losses will eventually result in ruin. The greater the number of permitted doubles, the shorter the losing sequence needed.

There are no MM shortcuts, no free lunch. If you're not net long while prices are rising, and/or net short when they're falling, frequently enough to overcome costs, then you will join the 95% (or whatever the number is) on the forex scrapheap. Nothing is surer.