Imp4ctScore

Transparent by design

How your scores work

Personal, Impact, and Team Score look at the same completed match from different perspectives. Here is what each one means—and what it does not.

01
Completed match

Scores describe evidence from a finished Competitive match.

02
Relevant context

You are compared only with the group defined for that score.

03
Clear result

Every available score uses the same readable 0–1000 scale.

Prototype, not Riot rankImp4ctScore is not MMR, ELO, a skill guarantee, or a prediction. Candidate inputs, weights, and score bands still require validation before production use.

Choose a perspective

One match. Three different questions.

Select a score to see what it answers, what it compares, and which performance signals can contribute to it.

Personal Score answers

How well did I perform in this match?

Personal Score brings your individual Competitive performance signals into one result, then places that result in the right player context.

Compared with
Players in the same broad rank, region, queue, and Act.
Shown on
Profile, match history, and Match Detail
Result
0–1000, with its calculation state clearly labelled
Performance signals

What can contribute

Signals are used only when the required source data passes the current quality rules.

  • Average Combat Score
  • KAST percentage
  • Clutch success percentage
  • Multikill rate
  • Damage delta per round
Reading the result

One scale, five provisional tiers.

A higher score means stronger performance inside the score's comparison group. It does not replace your VALORANT rank or MMR.

  1. D · Developing0–299
  2. C · Consistent300–474
  3. B · Strong475–649
  4. A · Excellent650–824
  5. S · Exceptional825+

These equal-width ranges are a synthetic prototype. Production boundaries remain disabled until the scoring model is validated and jointly approved.

Fair comparison

Context comes before the number.

Personal Score compares like with like. These four checks keep unrelated matches out of the comparison group.

01Broad rank tier

Your rank context when the match was played.

02Riot region

Players from different Riot regions stay separate.

03Competitive

Unrated and other modes do not enter the comparison.

04Same Act

Matches belong to the same Act whenever enough evidence is available.

Versioned calculation

From match evidence to a score.

The calculation follows the same visible sequence every time. If a required step fails, the interface explains why instead of inventing a value.

1

Validate

Check that the required match evidence is present and reliable.

2

Calculate

Turn each available performance signal into a comparable value.

3

Compare

Place the result in the correct rank, region, queue, and Act context.

4

Combine

Bring only validated signals together under the documented definition.

5

Label

Show the 0–1000 result together with its state and calculation context.

Score availability

No number is better than a misleading number.

A short status always tells you whether the score is ready, provisional, or cannot be calculated from the available match evidence.

Ready
The required evidence passed the current quality rules.
Provisional
A clearly labelled temporary benchmark or fallback is being used.
Missing data
Required evidence is incomplete, so no score is created.
Unavailable
The source or calculation is not available for this match.