Every strategy tester report and every verification service shows a block of summary figures. Each compresses the record into a number, and each leaves something out. Used together, they screen; used alone, they mislead.
Profit factor
Gross profit divided by gross loss. A profit factor of 1.5 means winners earned one and a half times what losers cost. Values between 1.3 and 2 are the ordinary range for real strategies over long records; values above 3 on a long record are rare, and on a short one are a sign of a favourable period or held losses. A grid system shows a very high profit factor until its losses are realised, because they are not in the gross loss yet.
Recovery factor
Net profit divided by maximum drawdown. It answers the same question as the Calmar ratio without annualising, and it grows with the length of the record, so it compares records of similar length only. A recovery factor of 3 over two years means the strategy earned three times its worst drawdown in that time.
System quality number
Van Tharp's SQN is the average trade in R, divided by the standard deviation of trades in R, multiplied by the square root of the number of trades. It rewards consistency and sample size together. Readings below 1.6 are poor, 2 to 2.5 average, 3 and above good, with the caveat that it rises with trade count regardless of edge, so a high SQN on thousands of tiny trades can describe a strategy whose edge is smaller than its costs.
The others
Expectancy, average trade and win rate have their own pages. Largest winning trade as a share of net profit shows dependence on one event. Average holding time classifies the strategy. Maximum consecutive losses is a check on the risk setting. And the number of trades governs how much any of the others can be trusted.
Two records. The first: profit factor 4.1, recovery factor 12, win rate 96 per cent, 1,800 trades, average holding time nine days, maximum drawdown on balance 6 per cent. The second: profit factor 1.6, recovery factor 2.8, win rate 43 per cent, 340 trades, average holding time three days, maximum equity drawdown 14 per cent. The first is a grid whose held losses are absent from every figure. The second is a strategy. The table flattered the wrong one.
Read the table as a set of screens. A profit factor between 1.3 and 2.5 on a long record, a recovery factor consistent with the record's length, an SQN above 2 on a few hundred trades, and no single trade dominating the profit, together describe a strategy worth reading further. Extreme values in any of them describe a short record, a favourable period, or a method that holds losses. In every case the equity curve and the trade list are where the statistics are checked, and the statistics are never a substitute.