How we count

Data source

All data comes from Supercell's official Clash Royale API. We use no other source and never read the game in any other way.

Which battles we collect

• Every day, the recent battles of the global Ranked top 1,000 players.
• Players who use Crownscribe (their profiles are refreshed more often while they play).
• Any player whose profile is opened on the site.

So most battles come from top-level play: at the moment about 9 in 10 are from Ranked's top league. Lower trophy ranges have far fewer battles, and pages say so when this applies.

Which modes count

Only 1v1 Ladder and Ranked battles. Clan war, events and challenges, tournaments and friendly battles are left out of the stats (they still appear in a player's battle log).

How we count

• Each battle has two sides. When we collect the same battle from both players, it is still counted once per side.
• Win rate = wins ÷ battles. Draws count as battles that were not won.
• Usage = a deck's or archetype's appearances ÷ all deck appearances. Every battle has two decks, so the total number of decks is twice the number of battles.
• Card usage = the share of decks that contain the card. Evolution and Hero forms are counted separately by how the card was actually played.

Patches and updates

Stats are counted per balance patch, because balance changes make older numbers out of date. Pages show the current patch by default; older patches and all-time numbers are available from the patch switch.

New battles are added to the stats every hour. Pages may be cached for up to 5 minutes.

Decks and archetypes

A deck is its 8 cards: the same 8 cards are the same deck. A different Evolution or Hero choice for those cards is a variant of that deck.

Archetypes are a hand-maintained list. A deck belongs to an archetype when it contains all of that archetype's signature cards; if several match, the most specific one wins. Decks that match none are counted in the totals only.

Sample sizes

• Every win rate is shown with its number of battles. Win rates from fewer than 30 battles are marked low sample.
• Written conclusions (best and worst matchups, counters) only use matchups with at least 100 battles.
• When a deck has too few exact matchups, the matchup table also counts decks sharing 7 or 6 of its cards, and says so.
• Evolution and Hero comparisons need at least 1,000 battles on both sides. The gap partly reflects who plays each form, not only the form itself.

Monthly reports

On the 1st of each month we take a snapshot of the current patch, from its start to the data cutoff shown in the report. A published report never changes, and its data can be downloaded as JSON.

Known limits

• The sample leans towards top-level play, so the numbers may not match what you see at lower trophies.
• The API only shows finished battles, so nothing on the site is live during a battle.
• These are statistics, not explanations: a high win rate shows what happened, not why.

Who runs Crownscribe

Crownscribe is run by analogKit, an independent developer. It is not affiliated with, endorsed, sponsored or approved by Supercell.

Found a wrong number? Email support@analogkit.app with the page link and we will check it.