Every figure shows its source. Where a source is missing, the number is a placeholder — and it is never written as fact.

By GamGem ResearchLast updated 15 Sept 2026

Rank bracket

Win rate, its sample and the standing letter follow this choice. Pick rate, lane, matchup, game-length and seven-day figures are counted across every bracket, and the panels that print them say so.

All 127 heroes

…and what your draft is missing

Patch 7.41f46,100,631 hero picksPublic, all brackets2026-09-16

0:30

Your own pick clock. Nothing here reads a running match.

Your side0/5

  • 1Safe lane
  • 2Mid lane
  • 3Off lane
  • 4Support
  • 5Hard support

What your side is missing

CarrySupportInitiatorDisablerFrontlineBurstTowerEscape
Bright: your side. Faint: the whole roster.
Phase
MeepoNo counted lane · Melee
ATier
Win
52.0%
Pick
0.2%
Ancient+53.7%
Sample79,815
Difficulty
7 days

Mid52.1%

Opens with

Optimal

Dragon KnightMid lane · Melee
BTier
Win
51.3%
Pick
0.6%
Ancient+51.6%
Sample298,893
Difficulty
7 days

Mid49.1%

Opens with

Learnable

KunkkaOff lane · Melee
CTier
Win
49.3%
Pick
0.4%
Ancient+49.7%
Sample196,895
Difficulty
7 days

Mid50.8%

Opens with

Naga SirenSafe lane · Melee
CTier
Win
48.8%
Pick
0.2%
Ancient+48.6%
Sample77,752
Difficulty
7 days

Mid52.6%

Opens with

AlchemistSafe lane · Melee
DTier
Win
47.4%
Pick
0.3%
Ancient+46.1%
Sample142,270
Difficulty
7 days

Mid52.9%

Opens with

Meepo struggles against

40%60%
  • Ember Spirit32.3%31 games
  • Queen of Pain35.5%31 games
  • Pudge47.2%36 games
  • Jakiro48.4%31 games
  • Disruptor48.5%33 games
  • Rubick55.0%40 games

Wins most against

40%60%
  • Mars58.3%36 games
  • Hoodwink56.1%41 games
  • Shadow Fiend56.7%30 games
  • Rubick55.0%40 games
  • Disruptor48.5%33 games
  • Jakiro48.4%31 games

Lane versus game

    • 0–20 min49.4%
    • 20–30 min53.6%
    • 30–40 min50.4%
    • 40–50 min48.5%
    • 50 min +45.4%

    Sources

    Why Meepo

    Won 52.0% of 79,815 counted public matches.

    win 52.0%sample 79,815

    Above Ancient it holds up, at 53.7%.

    above Ancient 53.7%sample 14,415

    Where the data disagrees

    Nothing in the counted figures for this hero points two ways. Manufacturing a tension here would be inventing one.

    Explain simply

    Out of every 100 public games with Meepo, about 52 were won. Valve calls it hard to play.

    win 52.0%Valve difficulty 3/3

    All 127 heroes

    127/127
    HeroLaneDifficultyStandingWin ratePick
    AbaddonSafe 43.3%B51.1%0.3%
    AlchemistSafe 53.4%D47.4%0.3%
    Ancient ApparitionSafe 51.4%A51.6%0.7%
    Anti-MagebuffedSafe 52.2%C49.8%1.1%
    Arc WardenMid 43.7%A51.7%0.4%
    AxeOff 48.6%B50.4%1.7%
    BanenerfedSafe 52%C49.8%0.4%
    BatriderbuffedOff 48.1%D44.4%0.1%
    BeastmasterOff 52.4%D45.0%0.2%
    BloodseekerSafe 55.5%B51.3%0.4%

    1–10 of 127 heroes

    Win rate, pick rate and sample are counted over 4,610,063 public matches. What is NOT here: a per-position breakdown, item timings, and the reader's own history — none of which this site collects. The tier letter is derived from the win rate, not published by Valve. What we do read.

    Official images

    • Dota 2, from the publisher's own gallery
    • Dota 2, from the publisher's own gallery
    • Dota 2, from the publisher's own gallery

    Game data and media via IGDB · igdb.com

    Sources

    • Roster · 127 heroesread 15 Sept 2026

      Name, primary attribute, attack type and role list, read from Valve’s published hero data.

    • Win rate, pick rate and sampleread 15 Sept 2026

      Counted over 4,610,063 public matches, per hero, with the same figures again above Ancient. The tier letter is derived from the win rate and is labelled so wherever it appears.

    • Served from Valve’s own CDN and never copied into this site. Steam’s API terms permit commercial use; GamGem is not endorsed by Valve.

    • What players buildread 15 Sept 2026

      The five most common items on each hero in each of Valve’s four game stages, counted over public matches. Each row carries its own pick count.

    • Matchups and game lengthread 16 Sept 2026

      Who each hero loses to and beats, counted over public matches with a floor of 30 games per pair, plus the same hero's win rate in each of five game-length bands. ⚠️ Observational: a matchup win rate is not a claim that one hero counters another by design. Synergy — how a hero does WITH another — is not published by this channel and is not shown.

    • Patch movementread 16 Sept 2026

      Which heroes the live patch names, and which way each line moved, counted off Valve’s own structured notes. A line the rule cannot read leaves the hero unmarked rather than guessed.

    The roster is the one Valve publishes for patch 7.41f. Draft picks are kept in this browser and never sent anywhere.