What 117 real backtests say about the two numbers a strategy report leads with. Win rate counts how often a trade closed with a gain; profit factor weighs what the winning trades made against what the losing trades lost. 6 of 37 runs that won at least half their trades still lost money, 11 of 56 that won under 40% made money, and every run with a profit factor of 1 or less lost
The win rate is the share of closed trades that ended with a gain after costs. The profit factor is the gains of all the winning trades added together, divided by the losses of all the losing trades added together: above 1, the winners made more than the losers lost. A win rate says how often; a profit factor says how much, which is the part that decides the result.
Across the 117 runs on the strategy result pages, the ones that made money won a median 50.4% of their trades and the ones that lost won 36.0%, so the win rate is not useless. It just does not decide: 6 of the 37 runs that won at least half their trades lost money, and 11 of the 56 that won fewer than 40% made money. The profit factor drew the line almost exactly: all 51 runs at 1 or less lost money, and 58 of the 66 above 1 made it.
Each dot is one backtest from the result pages: how often it won, against how much its winners made over what its losers lost.
117 runs. Every run at or under the line lost money; 8 of the 66 above it lost too, all with a profit factor of 1.20 or less. Among runs with 30 trades or more, runs that made money and runs that lost it sit side by side at every win rate from 25.9% to 54.7%.
It depends on the size of its wins against its losses: the payoff.
Divide the average winning trade by the average losing trade and you have the payoff. A strategy breaks even when it wins one trade in (1 + payoff): with wins the size of its losses it needs 50%, with wins twice the size a third, with wins half the size two thirds. Above that line it makes money, below it it loses, whatever the win rate looks like on its own.
| Average win against average loss | Win rate needed to break even |
|---|---|
| Wins half the size of losses | 66.7% |
| Wins the same size as losses | 50.0% |
| Wins 1.5 times the size of losses | 40.0% |
| Wins 2 times the size of losses | 33.3% |
| Wins 3 times the size of losses | 25.0% |
| Wins 5 times the size of losses | 16.7% |
The break-even win rate calculator does this for your own numbers, and the expectancy calculator turns a win rate and a payoff into what one more trade is worth.
Each from its result page; the averages come from the run's own trade list, after costs.
| Run | Trades | Win rate | Average win | Average loss | Needed to break even | Profit factor | Net |
|---|---|---|---|---|---|---|---|
| Bollinger Bands 20, 2.0 SPY daily, long and short | 138 | 54.4% | +1.6% | -2.4% | 59.5% | 0.81 | -28.2% |
| RSI(2) 10/90, 200-day trend, 5-day exit SPY daily, long and short | 113 | 67.3% | +1.1% | -1.7% | 60.7% | 1.33 | +20.3% |
| Trailing stop 9/21 cross, 3 ATR trail SPY daily, long and short | 94 | 39.4% | +5.1% | -2.1% | 28.7% | 1.61 | +79.8% |
| 200-day SMA SPY daily, long only | 29 | 31.0% | +12.7% | -1.8% | 12.7% | 3.10 | +91.0% |
| 50-day EMA SPY daily, long and short | 191 | 31.4% | +3.2% | -1.5% | 31.2% | 1.01 | -19.5% |
| Golden cross 50/200 QQQ daily, long only | 4 | 100.0% | +38.0% | none | any | none lost | +253.1% |
Bollinger Bands 20, 2.0 on SPY (long and short) won more than half its trades and lost money. It won 54.4% of 138 trades with an average win of 1.6% and an average loss of 2.4%, so it needed 59.5% to break even, and it lost 28.2% after costs.
RSI(2) 10/90, 200-day trend, 5-day exit on SPY (long and short) won two trades in three and made money, with small wins. It won 67.3% of 113 trades with an average win of 1.1% and an average loss of 1.7%, so it needed 60.7% to break even, and it made 20.3% after costs.
Trailing stop 9/21 cross, 3 ATR trail on SPY (long and short) lost most of its trades and made money. It won 39.4% of 94 trades with an average win of 5.1% and an average loss of 2.1%, so it needed 28.7% to break even, and it made 79.8% after costs.
200-day SMA on SPY (long only) won fewer than one trade in three and made money. It won 31.0% of 29 trades with an average win of 12.7% and an average loss of 1.8%, so it needed 12.7% to break even, and it made 91.0% after costs.
50-day EMA on SPY (long and short) had a profit factor above 1 and still lost money. It won 31.4% of 191 trades with an average win of 3.2% and an average loss of 1.5%, so it needed 31.2% to break even, and it lost 19.5% after costs.
Golden cross 50/200 on QQQ (long only) won every trade, on too few trades to judge. It won 100.0% of 4 trades, never closed one at a loss, and made 253.1% after costs. The tester's held-out check said: too few trades to judge.
The profit factor adds up each trade's return; the account compounds them. A gain of 10% followed by a loss of 10% has a profit factor of exactly 1 and leaves the account 1% down, because the loss came off a bigger balance. With the whole account in every trade, a profit factor close to 1 can lose money over many trades: the 50-day EMA rule on SPY (long and short) had a profit factor of 1.01 over 191 trades and lost 19.5%. 8 runs here did that, all at 1.20 or less.
| Run | Trades | Profit factor | Net |
|---|---|---|---|
| SuperTrend ATR 10, factor 3 on SPY, long and short | 93 | 1.05 | -17.8% |
| SuperTrend ATR 10, factor 3 on QQQ, long and short | 88 | 1.16 | -8.0% |
| MACD 12/26/9 on SPY, long and short | 231 | 1.01 | -4.5% |
| MACD 8/17/9 on SPY, long and short | 278 | 1.03 | -1.0% |
| Parabolic SAR 0.02 / 0.02 / 0.2 on NVDA, long and short | 233 | 1.001 | -82.6% |
| 50-day EMA on SPY, long and short | 191 | 1.01 | -19.5% |
| 50-day EMA on QQQ, long and short | 183 | 1.08 | -21.5% |
| Trailing stop 9/21 cross, 3 ATR trail on NVDA, long and short | 106 | 1.20 | -32.1% |
13 runs had fewer than 20 trades, and 4 of them won 75% or more, like the Golden cross 50/200 rule on QQQ (long only): 100.0% of 4 trades. On a handful of trades a perfect win rate or a profit factor in double figures is what luck looks like, which is why the tester judges a held-out third with too few trades as too few to judge rather than as a pass.
Describe it in your own words or paste the script. The Free plan writes the Pine v6 script with the AI, runs it on any US stock or ETF with costs charged, and reports the win rate, the profit factor, the drawdown and the held-out verdict, with every trade on the chart. No card.
Every dot is a run on a strategy result page, with every trade and the script as it ran. The same runs also show what a strategy's short side did, long only against long and short.
None on its own: the win rate a strategy needs depends on how big its winning trades are against its losing ones. With wins the same size as the losses it needs more than 50%; with wins twice the size, more than a third; with wins half the size, more than two thirds. In these 117 backtests the runs that made money won a median 50.4% of their trades and the runs that lost won 36.0%, yet 6 of the 37 runs that won at least half their trades lost money and 11 of the 56 that won fewer than 40% made money.
Above 1, the winning trades made more than the losing trades lost; at 1 or below they did not, and all 51 runs here at 1 or below lost money. Above 1 is not enough on its own: 8 of the 66 runs above 1 still lost money, all with a profit factor of 1.20 or less, because every trade used the whole account and a loss after a gain takes more than the gain added. The runs that survived their held-out third had a median profit factor of 1.45. Read it with the number of trades: on a handful of trades any profit factor is luck.
Profit factor, read with the number of trades. It weighs how much the winners made against how much the losers lost, which is what decides the result; the win rate only counts how often a trade closed with a gain. In these runs a profit factor at or below 1 lost money every time, while runs that made money and runs that lost money sat side by side at every win rate from 25.9% to 54.7%.
Yes, when its losses are bigger than its wins. The Bollinger Bands rule (20, 2.0) on SPY, long and short, won 54.4% of 138 trades, but its average win was 1.6% and its average loss 2.4%, so it needed to win 59.5% just to come out flat. It lost 28.2% after costs.
Yes, when its wins are much bigger than its losses. The Moving average rule (200-day SMA) on SPY, long only, won only 31.0% of 29 trades, but its average win was 12.7% against an average loss of 1.8%, so 12.7% would have been enough to come out flat. It made 91.0% after costs. Trend rules work this way: many small losses while they wait, a few large wins when a trend runs.
Add up the returns of all the winning trades, add up the losses of all the losing trades, and divide the first sum by the second. A profit factor of 1.5 means the winners made one and a half times what the losers lost. Alpha Charts adds up each trade's return in percent after costs, so a run that puts the whole account in every trade can show a profit factor just above 1 and still lose money, as the Moving average rule (50-day EMA) on SPY did: 1.01 over 191 trades, -19.5%.
They are the 117 runs on the strategy result pages, from the free Alpha Charts tester's own engine: Pine v6 scripts with their textbook settings on Alpaca daily data from 4 January 2016 and 5-minute data from the last 600 sessions, one position at a time with the whole account, fills at the next candle's open, 0.05% commission and 0.025% slippage on every fill. A trade counts as a win when it closed with a gain after costs. 13 runs had fewer than 20 trades, 4 of them with a win rate of 75% or more; the tester calls those too few to judge.
Every number here is the output of the Alpha Charts backtest engine on 25 October 2026, on Alpaca market data, for the scripts on the strategy result pages. Past results are not a forecast and nothing here is investment advice. TradingView and Pine Script belong to TradingView, Inc.; Alpha Charts is not affiliated with it.