Every trade measured against what you risked, where a full stop is always minus 1R.
An R multiple sizes every trade against the risk you put on it. Risk $100, make $250, and that trade is +2.5R. Get stopped at your full stop and it's -1R, every single time, whatever the dollar amount was. R turns a messy column of dollar results into one clean scale, so a $50 trade and a $5000 trade can finally be compared on the only thing that matters... how much you put at risk to make them.
Divide the result of the trade by what you risked. Risk 200 and make 600, that's +3R. Lose the whole 200, that's -1R. Your starting risk is the distance from your entry to your stop, times your position size. Lock that in as your 1R and every result reads as a multiple of it. A clean stop out always lands at -1R.
Because dollars hide your edge behind your position size. A trader who made 1,000 dollars risking 2,000 did worse than one who made 800 risking 400, and only R shows it at a glance. R also travels across accounts. The same strategy on a $5000 account and a $500000 account draws the same R curve. Dollars tell you how you did. R tells you how good the strategy is, which is why your Edge Score runs on the R scale.
Never annualise them. There's no square root of 252 scaling for R, and slapping it on turns a real +40R year into some fantasy number in the thousands. R multiples aren't percentage returns and don't compound like them, so report Total R or expectancy per trade, never a made up annual figure.