WinRate

The fraction of strictly-positive returns over a rolling window of returns, in [0, 1].

Quick reference

Field	Value
Family	Risk / Performance
Input type	f64 (per-trade or per-bar return / PnL)
Output type	f64
Output range	[0, 1]
Default parameters	period is required
Warmup period	period
Interpretation	Share of the window that strictly made money.

Formula

WinRate = #(rᵢ > 0) / period

Feed a stream of per-trade or per-bar returns; the indicator reports the rolling hit rate. A return of exactly 0 is treated as a non-win (a flat / scratch), so WinRate is the share of the window that strictly made money — the most basic performance statistic and a building block for Expectancy, Kelly sizing, and confidence filters.

Source: crates/wickra-core/src/indicators/win_rate.rs.

Parameters

Name	Type	Default	Valid range	Source	Description
period	usize	none	>= 1	win_rate.rs:46	Window of returns. 0 errors with Error::PeriodZero.

Inputs / Outputs

From crates/wickra-core/src/indicators/win_rate.rs:

use wickra::{Indicator, WinRate};
// WinRate: Input = f64, Output = f64
const _: fn(&mut WinRate, f64) -> Option<f64> = <WinRate as Indicator>::update;

Python streams as float | None, batches as a 1-D numpy.ndarray. Node streams as number | null, batches as Array<number>.

Warmup

warmup_period() == period. The unit test accessors_and_metadata pins warmup_period() == 20 for period = 20.

Edge cases

• Reference value. +, −, +, + is 3 wins of 4 → 0.75; pinned by reference_value.
• All wins / all losses. 1.0 / 0.0; pinned by all_wins_is_one and all_losses_is_zero.
• Flat returns. Zeros count as non-wins; pinned by flat_returns_are_not_wins.
• Rolling eviction. Old wins slide out of the window; pinned by rolling_window_drops_old_wins.
• Bounds. Output is always in [0, 1]; pinned by output_within_bounds.

Examples

Rust

use wickra::{BatchExt, Indicator, WinRate};

fn main() -> Result<(), Box<dyn std::error::Error>> {
    let mut wr = WinRate::new(4)?;
    println!("{:?}", wr.batch(&[1.0, -1.0, 2.0, 1.0]));
    Ok(())
}

Output:

[None, None, None, Some(0.75)]

Python

import wickra as ta

wr = ta.WinRate(4)
for r in [1.0, -1.0, 2.0, 1.0]:
    print(r, '->', wr.update(r))

Output:

1.0 -> None
-1.0 -> None
2.0 -> None
1.0 -> 0.75

Node

const ta = require('wickra');
const wr = new ta.WinRate(4);
for (const r of [1, -1, 2, 1]) console.log(r, '->', wr.update(r));

Output:

1 -> null
-1 -> null
2 -> null
1 -> 0.75

Interpretation

WinRate is the headline performance statistic for a returns or PnL stream: what fraction of recent trades/bars made money. On its own it is incomplete (a high win rate with large losers can still lose money), so combine it with Expectancy (per-trade edge in risk units) and ProfitFactor / KellyCriterion for a full picture.

Common pitfalls

• Win rate is not edge. A 70% win rate with a 1:5 reward:risk can be a losing system. Always read it alongside the payoff side via Expectancy.
• Flat = non-win. Scratch trades (return exactly 0) lower the rate; they are neither wins nor losses.

References

A standard trade statistic; see Van Tharp, Trade Your Way to Financial Freedom (1998), for win rate in the expectancy framework.

See also

• Indicator-Expectancy — per-trade edge in risk units.
• Indicator-ProfitFactor — gross profit / gross loss.
• Indicators-Overview — the full taxonomy.