How ELO Ratings Work

Understanding the rating system behind Grappling Stats

The Basics

We use an ELO rating system adapted for competitive no-gi grappling. Every athlete starts at a base rating (around 1500-1760 depending on weight class) and gains or loses points based on match results.

The core idea: beating a higher-rated opponent earns more points than beating a lower-rated one. Losing to a lower-rated opponent costs more than losing to a higher-rated one.

Expected Score: E = 1 / (1 + 10^((R_opponent - R_player) / 400))
Rating Change: ΔR = K × (Actual - Expected)

K Factor

The K factor determines how much a single match affects your rating. It's not a fixed number — it varies based on several factors:

FactorEffect
Base K60
Provisional (new athlete)Up to 3x multiplier for first 15 matches, declining linearly
Event tierADCC Worlds: 1.9x, Trials: ~1.4x, Opens: ~1.15x
Round (winner only)Finals: +40%, Semifinals: +12%, Quarterfinals/3rd Place: +6%
Rating gap dampeningKicks in at 120+ ELO gap. Floor of 0.40x — elite players always learn from matches
Opponent provisionalReduced K when beating unproven opponents

Margin of Victory

Not all wins are equal:

ResultScore
Submission1.0 / 0.0
Points / DQ0.9 / 0.1
Decision / Overtime0.8 / 0.2
Draw0.5 / 0.5

Event Tiers

TierEvents
10 (highest)ADCC Worlds, CJI #1
5ADCC Trials, CJI #2, WNO Championship
4WNO cards, BJJ Stars, UFC FPI
3EBI, Polaris, Kasai, IBJJF No-Gi majors
2.5ADCC Opens, SUG, ACBJJ
2F2W, GrappleFest
1PGF, regional events

Tournament Batching

Within a tournament, established players' ratings are frozen at the event start. All their match deltas are accumulated and applied after the event. Provisional players update live during the tournament.

This prevents early-round results from distorting late-round match expectations.

Hover Over Any Match

On athlete and event pages, hover over any match to see the ELO breakdown: rating before, K factor, expected win probability, and points gained/lost. This transparency lets you understand exactly how each match was scored.

Studies

Detailed analysis of our rating system and match data:

Accuracy & Calibration — predicted vs actual win rates by ELO gap

Data & Event Tiers — what events we cover and how they're weighted

The Submission Advantage — submission rates, top finishers, and upsets

Full methodology details coming soon. For questions or feedback, see the project repository.