RIGGEDROYALE
AnalyzePlayersDecksCardsData LabMethodology

This material is unofficial and is not endorsed by Supercell. For more information see Supercell's Fan Content Policy. Despite our name, Clash Royale is not actually rigged: Supercell itself explains that matchmaking is based on Trophies and King Level, not on your cards or your wallet. "Rigged" here is satire — we just measure how unlucky you got.

Public methodology·Public player board·Top decks·Public correlations·Fan content under the Supercell Fan Content Policy.

Privacy·Terms·Legal notice·Cookies

THE DECK LAB

SCORE ANY DECK AGAINST THE LIVE META

Build a deck and the model scores it against the live meta, or find single-card swaps that raise its win rate. Model Rankings shows the model's current top decks.

1 · Build the deck

Cards
8/8
Avg elixir
4.0
Special
0/3
Evo
0/2
Hero
0/2

Tap a card's EVO or HERO pill for a special form — 1 Evo, 1 Hero, plus 1 Wild slot either can take (3 max). The model scores evolutions; hero forms are shown for the build.

Card library

113 cards

2 · Score it

Score the current deck against the live meta.

500
50 · faster6,000 · full panel

Time

~3s est.

Actual coverage

48.3%

More decks increase meta coverage and runtime. Coverage comes from current live deck frequencies; times remain rough estimates.

UPGRADE PRIORITY ESTIMATE

Uses recent card usage, a modeled one-level effect, and relative upgrade-cost weights. It does not know your gold, wild cards, or exact inventory.

Pro

Included with Pro — sign in to get started. See what Pro includes.

THE MODEL'S OWN RANKING

TOP DECKS ACCORDING TO THE WIN-CHANCE MODEL

RankedLadder

Scoring the meta with the win-chance model — first build can take a minute…

ANALYST NOTESHow to read this data
  • The model board scores each popular deck against the frequency-weighted meta panel. It reads matchups, not piloting difficulty.
  • The deck lab and the experimental ML upgrade run on the same model and the same meta panel, and are also exposed as commercial API endpoints under /api/v1.
  • Descriptive data. Small samples, mode mix, patches, card levels, and selection bias can all move a deck before the wider player base confirms it.

594 popular decks screened; 10 distinct picks each scored against 100 frequency-weighted meta decks / League 7 / model v5-e2aa7bfc0c76 / built Jul 14, 03:33 PM

01

EVO
EVO
0.2% of scored meta playbest 80%worst 27%volatility 27

58.8%

model WR vs meta

02

EVO
EVO
0.1% of scored meta playbest 84%worst 22%volatility 33

58.6%

model WR vs meta

03

EVO
EVO
1.3% of scored meta playbest 83%worst 30%volatility 25

57.8%

model WR vs meta

04

EVO
EVO
0.0% of scored meta playbest 78%worst 29%volatility 24

57.7%

model WR vs meta

05

EVO
EVO
0.0% of scored meta playbest 83%worst 30%volatility 25

57.6%

model WR vs meta

06

EVO
EVO
0.0% of scored meta playbest 82%worst 35%volatility 25

57.4%

model WR vs meta

07

EVO
EVO
0.1% of scored meta playbest 78%worst 26%volatility 27

57.3%

model WR vs meta

08

EVO
EVO
0.1% of scored meta playbest 80%worst 22%volatility 23

57.1%

model WR vs meta

09

EVO
EVO
0.0% of scored meta playbest 78%worst 36%volatility 19

56.9%

model WR vs meta

10

EVO
EVO
0.7% of scored meta playbest 87%worst 19%volatility 27

56.8%

model WR vs meta