How do you measure reward against risk before entering?
Comparing a $4 stop on a $100 stock to a $0.40 stop on a $9 stock is comparing nothing to nothing — different prices, different sizes, different accounts. Traders solve this with one normalising unit.
R: the unit that makes trades comparable
1R = the money at risk on the trade = (entry − stop) × shares.
That's it. Then every outcome gets restated as a multiple of it. A trade that loses at the stop is −1R. A trade that makes three times what it risked is +3R. A $4 stop on 125 shares and a $0.40 stop on 1,250 shares are both 1R = $500 — identical trades in the only unit that matters.
R does three jobs at once: it makes trades comparable across instruments, it makes a track record comparable as the account grows, and it turns the expectancy formula (next lesson) into something you can do in your head.
Planned risk/reward
Planned R:R = (target − entry) ÷ (entry − stop)
Entry $100.00, stop $95.00, target $115.00 → risk $5.00, reward $15.00 → a 3:1 plan, i.e. a +3R target.
The caveat that makes this honest
The two legs are not the same kind of object. The risk leg is committed — you will take it if price goes there. The reward leg is a hypothesis — it only exists if price reaches the target, and nothing in the ratio says how often that happens.
Which is why "always take trades with 3:1 or better" is arithmetic with a piece missing. Push a distant target far enough away and you improve the ratio while destroying the hit rate. The formula that joins them is the breakeven win rate:
Breakeven win rate = 1 ÷ (1 + R)
- A 5:1 plan needs to work only 16.7% of the time. If it actually fills 10% of the time: 0.10 × 5 − 0.90 × 1 = −0.40R per trade. A losing system with a beautiful ratio.
- A 1.5:1 plan needs 40%. If it fills 50% of the time: 0.50 × 1.5 − 0.50 × 1 = +0.25R per trade. A winning system with a modest ratio.
The "worse" ratio is the profitable one, because the hit rate came attached to it. A ratio without a hit rate is half a statistic.
Where the hit rate comes from
Only from observation: a logged record of comparable setups, or a backtest — with all the caveats Unit 4 is about to make painfully specific. It never comes from the ratio itself, and it never comes from confidence.
The hit rate the ratio needs is a path statistic: the share of comparable setups in which price reached the target before it touched the stop. The count from the patterns course's counting procedure is a different number, the share of instances whose close at a fixed horizon was higher, and the two disagree on every path that crossed the stop or the target before the horizon ended. A setup that ends higher ten sessions later after trading through the stop on day three is a win in the first test and a loss in the second. Feed the horizon count into 1 ÷ (1 + R) and the breakeven comparison is made against a number the trade never sees. Record which of the two your log holds.
The schematic below has everything the ratio needs and nothing the hit rate needs: a band price has turned in, a prior high to aim at, and a level beneath which the idea is wrong.
Try it now
- On the schematic above, mark a hypothetical entry on the last bar, a structural stop below the band, and a target at the earlier high. Compute the planned R:R from those three numbers alone.
- Compute the breakeven win rate for that ratio: 1 ÷ (1 + R).
- Now go looking for the half a ratio can never contain. Scroll back through the five years below and count how many comparable setups reached a similar target versus how many hit a similar stop, then compare with your breakeven rate. If you can't count at least a dozen, the honest answer is "I don't know yet" — which is a legitimate finding, not a failure.