Plotting win rate against reward to risk for each setup, and what the picture showed

[Fairman]

Plotting win rate against reward to risk for each setup, and what the picture showed

I made a simple scatter plot, with one point for each of my setups. The horizontal axis is the win rate and the vertical axis is the average reward to risk. I added a line showing the break even point. It gave me a different view of my own trading.

Some setups were high win rate with low reward, others the opposite. What mattered was which side of the break even line each was on, and how far. One of my favourites, a high win rate setup, was sitting close to the line, so its edge was thin. A setup I didn't like because it lost often was well above the line, thanks to its large winners.

That changed how I think about them. I'd been judging setups by how they felt, a high win rate feels good and a low one doesn't. The plot showed what they were actually contributing. It also showed that costs matter more for the high win rate, low reward type.

I'm not changing my entire approach, but I'm paying attention to the setups that are above the line and cutting down on the ones that aren't. I redo the plot every quarter with new data.

[Fairman]

Did something similar and it was humbling. My favourite setup was right on the break even line once I included costs. The plot without costs made it look fine, but the high win rate and small targets meant the spread was eating most of the edge. If anyone does this, I would plot it twice, once before costs and once after, and see how much each point moves. The low reward setups move the most, which is a real eye opener.

[Fairman]

Worth being careful with small samples on a plot like this. A setup with fifteen trades can land anywhere on the chart by chance. I put the number of trades next to each point, and I treat anything with fewer than fifty or so as unknown rather than good or bad. It stopped me from dropping a setup that had a rough start and from getting too excited about one that had a lucky stretch. The plot is a good conversation starter with yourself, not a verdict on its own.

[PTScalper]

Plotting win rate against reward to risk for each setup, and what the picture showed

I made a simple scatter plot, with one point for each of my setups. The horizontal axis is the win rate and the vertical axis is the average reward to risk. I added a line showing the break even point. It gave me a different view of my own trading.

Some setups were high win rate with low reward, others the opposite. What mattered was which side of the break even line each was on, and how far. One of my favourites, a high win rate setup, was sitting close to the line, so its edge was thin. A setup I didn't like because it lost often was well above the line, thanks to its large winners.

That changed how I think about them. I'd been judging setups by how they felt, a high win rate feels good and a low one doesn't. The plot showed what they were actually contributing. It also showed that costs matter more for the high win rate, low reward type.

I'm not changing my entire approach, but I'm paying attention to the setups that are above the line and cutting down on the ones that aren't. I redo the plot every quarter with new data.

Hi Fairman,

Visualizing setups on a Win Rate vs. Reward-to-Risk (R:R) scatter plot strips away the emotional bias of trading and reduces strategy evaluation to pure mathematical expectancy. By plotting that breakeven curve, you have essentially mapped the exact boundary between a negative-sum hobby and a positive-edge business.

Here is a deeper breakdown of the mechanics your plot revealed, the math behind the curve, and how to weaponize this data further.

[PTScalper]

The Mathematics of the Breakeven Curve

The "line" you plotted is technically a hyperbolic curve derived from the core expectancy formula. If we normalize risk to 1 unit, the breakeven point occurs when the expected value ($E$) is exactly zero.Let $W$ represent your Win Rate (as a decimal) and $R$ represent your Reward-to-Risk ratio:

$$E = (W \times R) - (1 - W) \times 1 = 0$$

[PTScalper]

Any setup plotted above and to the right of this curve has a positive mathematical expectancy. The perpendicular distance from the setup's coordinate to the breakeven curve is the literal measure of your "edge." As your plot revealed, a setup with a 40% win rate and a 2.5 R:R is mathematically superior to a 75% win rate setup with a 0.3 R:R, even though the latter "wins" almost twice as often.

[PTScalper]

The Frequency vs. Magnitude Illusion

Your realization about judging setups by "how they feel" highlights the most common psychological trap in trading: the human brain is hardwired to prefer the frequency of success over the magnitude of success.

High Win Rate / Low R:R (The "Feel Good" Setup): These setups provide constant psychological validation. However, as you noticed, their edge is often razor-thin. A slight variance in market conditions or a brief drawdown can instantly push a strategy with a 1.2 R:R and a 50% win rate below the breakeven curve.

Low Win Rate / High R:R (The "Grinder" Setup): These setups are psychologically exhausting because you spend the majority of your time taking losses or getting stopped out at breakeven. Yet, because the winners are outsized (e.g., 3R, 4R, or riding a trailing stop in a trend), the distance from the breakeven curve is massive. The math does the heavy lifting, absorbing strings of losses effortlessly.

[PTScalper]

The Hidden Drag: Transaction Costs

Your observation that "costs matter more for the high win rate, low reward type" touches directly on market microstructure. Spreads, commissions, and slippage act as a fixed tax on every round-trip trade.

If you are scalping for a 0.5R target with a tight stop loss, a 1-pip spread or a slight execution delay consumes a massive percentage of your gross profit. Conversely, if you are capturing a 3R swing trade, that exact same 1-pip spread becomes statistically insignificant relative to the gross reward. High win rate/low R:R setups are highly fragile to friction; what looks like a profitable setup in backtesting often falls below the breakeven curve in live execution due to these costs.

[PTScalper]

Taking it Further: Position Sizing and Capital Allocation

Since you are already updating this plot quarterly, the next logical step is to use the data to dynamically adjust your risk allocation.

Instead of risking a flat 1% or 2% of your account per trade across all setups, you can weight your risk based on the setup's distance from the breakeven curve.

A-Grade Setups (Far above the curve): These are your statistical heavyweights. They earn maximum risk allocation.

B-Grade Setups (Near the curve): These are your high-win-rate/low-R:R setups. They provide equity curve smoothing but shouldn't command major capital. Scale down the risk on these setups to protect capital from transaction costs and variance.

C-Grade Setups (Below the curve): As you noted, these are the ones to cut.

[PTScalper]

Tracking this data quarterly allows you to spot "system decay" early. Markets evolve, and a setup that sat comfortably above the curve in Q1 might drift dangerously close to breakeven by Q3 due to shifts in volatility or liquidity. Visualizing the drift on a scatter plot gives you the objective signal to deprecate an algorithm or setup before it causes a drawdown.