Clash of Clans · Live data
Clash of Clans Army Meta
35 attack archetypes ranked by real usage and three-star rate — 116,238 classified battles shown, from a larger deduplicated corpus over the last 7 days (snapshot: 2026-08-10). Decoded straight from attacker army compositions in official ranked battle logs — not polls, not guesses.
How do your own attacks compare to this meta?
The Battle Report reads your recent attacks from the official API and grades them against these same archetypes — which armies you actually three-star with, and where you are leaking stars.
This snapshot covers a live event
- Mass Elephant Riders and Super Valkyries (2026-08-09 → 2026-08-10) — 50 Elephant Riders and 20 Super Valkyries added to your regular army for 24 hours. Attacks in this window can carry granted troops that no Town Hall could train, so unit-based usage shares are inflated for those armies. Armies over their camp capacity are labelled as event armies rather than counted as a real strategy.
Ranked battles are the only ones counted here, which excludes limited-time event modes — but these events add troops to a player's regular army, and that army still gets taken into ranked. The filter does not remove them.
Falling
- Mass Thrower-0.6pp usage
| # | Army | Usage | Δ Usage | 3★ rate | Avg ★ | Battles |
|---|---|---|---|---|---|---|
| 1 | Mass Dragons | 34.8% | -0.4pp | 46.7% | 2.32 | 40,652 |
| 2 | E-Drag Spam | 16.8% | +0.5pp | 29.7% | 2.07 | 19,585 |
| 3 | Root Rider | 12.8% | +0.3pp | 51.1% | 2.39 | 14,986 |
| 4 | Mass Thrower | 11.9% | -0.4pp | 55.4% | 2.46 | 13,947 |
| 5 | Super Bowler Smash | 5.7% | -0.1pp | 59.2% | 2.52 | 6,703 |
| 6 | Mass Super Yeti | 2.4% | +0.1pp | 55.5% | 2.46 | 2,780 |
| 7 | Mass Super Dragon | 2.1% | +0.0pp | 42.5% | 2.24 | 2,415 |
| 8 | Mass Meteor Golem | 1.8% | +0.1pp | 47.1% | 2.34 | 2,142 |
| 9 | Yeti Smash | 1.8% | +0.1pp | 40.5% | 2.25 | 2,081 |
| 10 | Mass Dragon Rider | 1.6% | -0.1pp | 53.9% | 2.41 | 1,869 |
| 11 | Mixed | 1.5% | +0.0pp | 39.8% | 2.14 | 1,778 |
| 12 | Mass Valkyrie | 1.4% | +0.0pp | 15.7% | 1.45 | 1,608 |
| 13 | Mass Witches | 1.3% | 0.0pp | 44.0% | 2.24 | 1,556 |
| 14 | Mass Loons | 0.4% | -0.0pp | 43.0% | 2.18 | 456 |
| 15 | Electro Titan | 0.4% | -0.0pp | 30.8% | 2.06 | 435 |
| 16 | Mass Miners | 0.3% | +0.1pp | 34.1% | 2.12 | 399 |
| 17 | Mass P.E.K.K.A | 0.3% | +0.0pp | 17.7% | 1.59 | 328 |
| 18 | Hybrid (Miner-Hog) | 0.3% | +0.0pp | 52.0% | 2.32 | 327 |
| 19 | Hog Rider | 0.3% | +0.0pp | 28.1% | 1.79 | 310 |
| 20 | Sneaky Goblin Farm | 0.2% | -0.0pp | 5.2% | 1.59 | 272 |
| 21 | LavaLoon | 0.2% | +0.0pp | 31.7% | 1.98 | 208 |
| 22 | Pekka Bowler | 0.2% | -0.0pp | 19.4% | 1.91 | 201 |
| 23 | Mass Super Barbarian | 0.2% | -0.0pp | 1.1% | 0.64 | 183 |
| 24 | Mass Super Archer | 0.1% | -0.0pp | 73.1% | 2.69 | 175 |
| 25 | Mass Rocket Balloon | 0.1% | -0.0pp | 42.0% | 2.11 | 169 |
| 26 | Mass Furnace | 0.1% | -0.0pp | 40.3% | 2.18 | 149 |
| 27 | Mass Super Minion | 0.1% | 0.0pp | 4.4% | 1.20 | 135 |
| 28 | Mass Healer | 0.1% | -0.0pp | 20.2% | 1.04 | 89 |
| 29 | Golem Smash | 0.1% | 0.0pp | 19.1% | 1.79 | 68 |
| 30 | Mass Giant | 0.0% | +0.0pp | 26.2% | 1.52 | 42 |
| 31 | Mass Wizard | 0.0% | +0.0pp | 21.9% | 1.46 | 41 |
| 32 | Mass Bowler | 0.0% | 0.0pp | 25.5% | 1.90 | 51 |
| 33 | Mass Super Hog Rider | 0.0% | 0.0pp | 39.4% | 2.03 | 33 |
| 34 | Mass Baby Dragon | 0.0% | 0.0pp | 22.6% | 1.81 | 31 |
| 35 | Mass Dragon | 0.0% | 0.0pp | 29.4% | 1.62 | 34 |
How is this computed?
TrophyCoach collects multiplayer battle logs from the official Clash of Clans API, deduplicates every battle by fingerprint, and decodes each attacker's army composition into a named archetype. Usage share is measured against all deduplicated battles in the window — including unclassified armies — so shares reflect the real meta, not just what we can label. Archetypes below a 30-battle sample are hidden rather than shown with noisy rates.
Frequently Asked Questions
- What is the best army in Clash of Clans right now?
- The table above ranks every tracked attack archetype by real usage share and three-star rate, recomputed daily from ranked ladder battles. The highest-usage army is not always the highest three-star army — check both columns before committing your camps.
- Where does this army meta data come from?
- TrophyCoach continuously collects battle logs from the official Clash of Clans API and decodes the attacker army in each battle. Only ranked battles count toward this table: regular village attacks mix in farming armies and limited-time event modes, which measure something different. Every battle is deduplicated by fingerprint, and an archetype only appears once it clears a 30-battle minimum sample.
- How is an army archetype like "Root Rider" or "Hybrid" detected?
- Each attack log entry carries the attacker's army composition. TrophyCoach decodes it and classifies the composition into a named archetype — specific comps first (Root Rider, LavaLoon, Hybrid Miner-Hog, E-Drag Spam), then mass-unit families, then Mixed.
- How often does the Clash of Clans army meta update?
- The aggregates recompute every 24 hours over a rolling 7-day battle window, and this page refreshes hourly. The snapshot date and sample size are shown above the table.
- Do limited-time events skew the Clash of Clans army meta?
- They can, and TrophyCoach says so on the page when they do. August 2026 runs a mass-troop grant on August 9-10 (50 Elephant Riders and 20 Super Valkyries) and keeps the Elephant Rider available through August 31. Any army whose troops exceed its Town Hall camp capacity — 352 housing at Town Hall 17 and 18, 340 at Town Hall 16 — is physically impossible to train, so it is labelled an event army instead of being counted as a real strategy. Whenever the snapshot window overlaps one of these events, a dated notice appears above the table.
- Does this include farming, event and war attacks?
- No. This table is built from ranked battles only. Regular village attacks are excluded because they blend farming armies — which are not trying to three-star — with limited-time event modes that hand out armies far larger than any camp can hold. Clan-war attacks are not exposed in battle logs by the official API, so war-specific performance is not included either.