I had this question come in recently:
Strategy A has better risk-adjusted returns (measured by Annual returns/Max Drawdown, aka the MAR ratio), but lower expectancy than strategy B. It manages to achieve this by having a larger number of trades, even though the backtest period is the same for both.
Which of the two performance measures should I rely on in choosing one strategy over the other?
When dealing with expectancy it is important to not just look at it in terms of per trade figures. You must also account for the frequency of trades. In other words, it will often be best to think in terms of expectancy on time basis rather than a trade one. For example, you could think of monthly expectancy to figure out what kind of returns you would expect to see in a meaningful time frame for your trading. This would be the better way to compare two systems or strategies.