Strategy types: trend, reversion and breakout
There are hundreds of strategies, but almost all fall into three families. Understanding them saves you time: each wins in one kind of market and loses in another. None wins always.
1. Trend-following
The idea: "the trend is your friend". You buy what's rising and sell what's falling, betting the move continues. Typical tools: moving averages, crossovers, breakouts of highs. It wins in clear, sustained trends. It suffers in ranging markets, racking up small losses on false signals. Win rate is usually low (30-45%) but with high R: a few big wins pay for many small losses.
2. Mean-reversion
The opposite idea: prices overshoot and tend to return to their average. You buy the oversold and sell the overbought. Tools: RSI, stochastic, Bollinger Bands. It wins in ranges. It suffers —badly— in strong trends: "oversold" can keep falling for weeks. Win rate is usually high (55-70%) but with low R, and the danger of occasional large losses.
3. Breakout
The idea: when price breaks an important level (a range, a high), a strong move often follows. You trade the break expecting volatility. The ORB (opening range breakout) is a classic. It wins in the transition from range to trend. It suffers from "false breakouts", which are very common.
The mistake isn't picking "the best" strategy — it's running a trend strategy in a ranging market, or a reversion strategy mid-trend. That's why serious systems use a regime filter (e.g. ADX: high = trend, low = range) that switches on the right strategy for the moment. Signal-market coherence beats hunting for the magic indicator.
What about robots?
An EA isn't a fourth family: it automates one of these three (or a mix). Its quality depends on whether the logic has a real edge and is applied in the right regime — not on it being "a robot". That's why you must validate it with real ticks before trusting it with money.