Skip to content
OneKitly

Two Kills and Twenty Assists Outranks Ten Kills and Three Deaths

Published 11/5/2026 · 3 min read · Everyday calculators

Daniel Okonkwo

Daniel Okonkwo — Front-end developer and tech writer at OneKitly

Web performance · File formats

Checked against 2 sources

View profile →
In short

The formula adds kills and assists, then divides by deaths. Ten kills, three deaths and five assists gives (10 + 5) ÷ 3 = 5.0. Two kills, four deaths and twenty assists gives (2 + 20) ÷ 4 = 5.5, a better score from a player who eliminated almost nobody and died more often. That is not a bug: the statistic was designed to credit contribution rather than finishing, and in team games the player who sets up a kill often did more than the one who landed the last hit. What it cannot do is distinguish a player creating those opportunities from one standing at the edge of every fight collecting participation.

Advertisement

The first line scores 5.5, the second 5.0. Assists count the same as kills in the numerator, which is the whole design and the whole problem.

Dividing by zero is handled by pretending

A player who never dies would produce an undefined score, so implementations substitute one death and report the result as though it were real. Ten kills and five assists with no deaths therefore shows 15.0 rather than an error, which is convenient and quietly dishonest: the same line with a single death would also show 15.0, and the two performances are not equivalent. Some systems mark the figure as 'perfect' instead, which at least signals that the arithmetic was bypassed. The general lesson is that any ratio deployed at scale eventually meets a zero denominator, and how it is patched determines what the leaderboard rewards.

What a leaderboard rewards is what players will do

Because deaths sit in the denominator, the cheapest way to improve the score is to avoid dying, and avoiding dying often means declining fights the team needed someone to take. A player who holds back at the critical moment posts a better line than one who traded a death for a won objective, and objectives are how matches are actually decided. This is the standard failure of any single-number rating: it becomes a target, and behaviour reshapes itself around the target rather than the goal. Games that care about this either weight the components differently or publish several numbers so that no one of them can be optimised alone.

Score
Four lines and their scores
Kills / deaths / assistsScoreNote
10 / 3 / 55.0Strong all round
2 / 4 / 205.5Scores higher
5 / 5 / 52.0Perfectly ordinary
10 / 0 / 515.0Zero deaths treated as one

Worked with our own calculator

KDA ratio calculator

Given

Kills
10
Deaths
3
Assists
5

Result

KDA ratio
5

These figures are produced by the calculator below, not typed in by hand — they are recomputed whenever the tool changes.

Run it on your own figures →
Advertisement

Frequently asked questions

Should assists really count as much as kills?
Most analysts think not, and the common alternative gives an assist half the weight of a kill. That change alone reorders the two lines in the example: ten kills, three deaths and five assists becomes 4.2, while two kills, four deaths and twenty assists becomes 3.0, reversing the verdict. Which weighting is right depends on the game — in some, an assist requires nearly as much of the work as a kill, and in others it is awarded for standing nearby. No universal answer exists, which is a decent argument for reading the raw three numbers instead of any ratio built from them.
Why is a 5 / 5 / 5 line worth 2.0 and not 1.0?
Because assists join kills above the line. Five kills plus five assists is ten, divided by five deaths gives two — the score reads as though the player achieved twice as much as they lost, when kills and deaths were in fact even. Anyone using the number as a rough sense of whether a match went well should therefore know that the neutral point is not 1.0 but somewhere near 2.0 to 3.0 depending on how many assists the game format generates. Comparing scores across different games or modes without knowing that baseline is comparing numbers that were never on the same scale.

Articles you may find interesting

All guides →

Related tools

Sources

Spotted a mistake in this article?