Teamfight Manager 2 tier list

Updated 2026-05-27 · based on v0.4.3 · current patch v0.4.12 · 9 min read

Sur cette page

Quick Answer

TFM2 launched into Early Access on 2026-05-25. Per-champion stat numbers are visible in the game but not in any published source we'd trust to cite, and the in-game Auto Patch System rebalances champions every simulated season. So this is a creator-sourced commentary list, not a math-driven ranking — pinned to v0.4.3 and first-week creator coverage. Patch 0.4.5 made tier classifications editable from Champion Analysis, so use this as a starting frame, then keep your own Personal Tier List current in-game once you've drafted a few series.

Position passives (read this before the rankings)

Every lane carries a passive that shapes what kind of champion earns its keep there. The tier picks below read with these in mind.

RolePassive
Top+1% max HP regen per second
Jungle+20% MS out of combat; execute epic monsters at ≤700 HP
MidBonus experience
Bot+20% gold; −30% experience
Support−15% gold; on last hit, the nearest ally receives the gold instead

Standout

6 champions

Creators called these out by name as best-in-class on the EA build.

  • JiangshiADDED · V0.4.3

    LettucePlate: “still the best tank, same as TFM1.” Returning from TFM1 but skill-reworked — don’t import old builds.

  • BerserkerADDED · V0.4.3

    LettucePlate: “best 1v1 champ.” Resetting ult on takedown — slots into split-push and skirmish drafts.

  • CavalryADDED · V0.4.3

    Charging dive class. Creator-recommended blind pick — works before you know the enemy comp.

  • VampireADDED · V0.4.3

    Aatrox/Vladimir-style lifesteal bruiser with strong top-lane sustain. Pairs with the Top regen passive.

  • SniperADDED · V0.4.3

    Long-range carry that scales with attack range — Caitlyn analogue. Punishes drafts without an answer to range.

  • Whip MasterADDED · V0.4.3

    Ranged whip with cone knockback. LettucePlate noted a high win rate in the stream’s meta panel.

Defined

11 champions

Mechanically clear identity with a usable LoL parallel — no creator hype, but no red flags. Reliable in-draft.

  • Wind MageADDED · V0.4.3

    Bard/Janna enchanter — AS buff, CDR zone, channeled team-buff ult.

  • AndroidADDED · V0.4.3

    Shen analogue: shield on demand + global teleport-to-ally ult. Clean savior identity.

  • LancerADDED · V0.4.3

    Anti-structure top: bonus damage to towers, self attack-speed ult. The siege pick.

  • BomberADDED · V0.4.3

    Ziggs analogue: AOE bombs + trap mine + a giant ult bomb (physical damage).

  • Lightning MageADDED · V0.4.3

    Brand/Malzahar analogue: chain lightning + AoE stun + channeled DOT ult.

  • Plague DoctorADDED · V0.4.3

    Renata + Milio support: AS buff, range/dash buff, ult drops ally to 1 HP but immortal w/ AD/AS/MS for 5s.

  • DruidADDED · V0.4.3

    Annie with Tibbers — summons a bear that can be healed.

  • NecromancerADDED · V0.4.3

    Yorick/Sion analogue: summons a ghoul; the ult revives a fallen ally as a revenant.

  • ExorcistADDED · V0.4.3

    Cleanse-on-cooldown support: 2s nullify-CC + 300 HP heal, only triggers on CC.

  • ChefADDED · V0.4.3

    Mundo: knife slow + food heal + ult gives ally +50% max HP and 30% damage redistribution. Do not run in a solo lane — see Caveats.

  • Shield BearerADDED · V0.4.3

    Thornmail support: shields ally, 60% damage redirect, slows attackers in radius. Level-3 spell scales with armor — defensive and offensive stack.

Caveat

2 champions

Kit is fine on paper, but creators flagged AI issues or post-demo nerfs. Use with awareness.

  • PythonessADDED · V0.4.3

    Damage mid mage with a shrinking-airborne redirect ult. LettucePlate: “AI uses this ult badly” — fires into empty space. Strong on paper, AI-limited in practice.

  • NinjaADDED · V0.4.3

    Mobility assassin. LettucePlate: “strong in the demo, weaker in EA.” Tuned down since playtest.

Why a hard tier list is dishonest right now

  • Team Samoyed designs champions to be flexible across roles and only commits to strengthening role identity if too many become universally viable. A static ranking ages out of date fast.
  • The in-game Auto Patch System collects per-season win-rate data and auto-buffs or nerfs champions on its own schedule, distinct from developer patches. The meta moves under you.
  • The roster ships at 60 champions with a roadmap to 100 during EA — new picks arrive mid-save (default two per developer patch, configurable).
  • The only numerical balance data Team Samoyed has published is about stat development: an equal-build 100-stat team beats a 1-stat team about 75% of the time. Who you draft matters less than how well you develop the roster you have — another reason a static ranking misses the picture.
  • The game already ships a Personal Tier List overlay you fill out yourself during draft. Building yours is the real answer.

Not yet rated

The rest of the launch pool. No public creator commentary yet — lack of a placement here is not a weakness rating. We would rather leave a champion unranked than fabricate one.

Where to go from here

The strongest next step isn't memorising a static ranking — it's seeing how these champions actually slot into a team. The best comps page breaks down the draft archetypes that win games, and the full champion database has every champion's class, abilities, and per-level stats. In game, the Personal Tier List overlay lets you build your own frame as your runs accumulate — treat the meta as a moving target, not a fixed ranking.

FAQ

Sources

Tier list - Teamfight Manager 2