Quote from Granpa:
Averaging down will make a good trading system into an excellent one, and a bad trading system into a horrific disaster.
That said, averaging down has always been the backbone of my profits. If the timing of my first trade is wrong, but I'm still confident on the value, I'll buy 1.5x the first trade at a lower price, and if that's still off, I might buy 1.5x the first trade again at an even lower price, making a position 4x the original investment. (I used to do 1x, 1.5x, 2.5x during the psychotic .com volatility)
Simply doubling a position helps but in theory statistically, if it is a good value at the first price, its a great value at the lower price and should receive a larger bet and so on. (all else being equal that is)
The trader has to have a good system tho, a lot of mental stamina and trading maturity, a good science of what level/price to increase the original position, and a strong idea of when to say uncle (not to mention having a keen sense of frontrunning/unusual trading activity doesn't hurt.)
The one year that I didn't profit (2002), I cautiously used stops after the first opening position instead of taking more risk, and that almost put me out of business.
You would almost certainly be MORE profitable--in terms of risk/reward, and profits/drawdown--if you stopped averaging down.
Likewise, if roulette had positive expectancy (obviously, it doesn't), and there were no limits on betting size, one could Martingale (or anti-Martingale) her way into a fortune, but trading is very different from the roulette wheel. In TRADING but one could use OTHER forms of money management and do it more effectively.
