About MLCounters & Our Methodology

What is MLCounters.app?

MLCounters.app is a free counter-pick and meta-statistics site for Mobile Legends: Bang Bang (MLBB). It tracks all 133 heroes with current win rates, pick rates, ban rates, counter matchups, and teammate synergies, plus a live draft counter tool that recommends picks in real time during your draft. Data was last refreshed on 2026-08-27.

Where the data comes from

• Win, pick, and ban rates are collected from the official Mobile Legends ranked statistics published at mobilelegends.com, covering current ranked-play data.
• Counter and synergy scores come from the official per-hero matchup data: a positive score means the hero performs measurably better against (or alongside) the other hero than average; a negative score means the matchup is unfavourable. Counter data is read from five ranked tiers rather than one, which is why a matchup can appear here that the in-game screen does not list.
• Hero attributes (durability, offense, control ratings) are sourced from the Mobile Legends wiki via its public API.
• Win rate by game length comes from the same official matchup data, which reports each matchup split into game-length brackets. It is what tells us whether a hero gets stronger or weaker as a match runs long.

We do not hand-edit stats or accept payment to change rankings. Every number on the site is derived from the sources above by the same automated pipeline.

How the meta score is calculated

Each hero's meta score (0 to 100) blends three components:

• Ranked performance (70%): win rate weighted by pick and ban rate, so a high win rate on a frequently picked/banned hero counts for more than one on a niche pick.
• Base strength (15%): the hero's attribute ratings.
• Scaling (15%): how much the hero's win rate actually rises or falls as a match runs long, measured from ranked results by game length.

Heroes are then ranked #1 to #133 to produce the meta tier list. The live draft counter combines these meta scores with matchup data, lane assignments, team-composition needs, and blind-pick safety, weighted by draft phase.

Scaling used to be estimated from how much a hero's stats grow between level 1 and level 15. We replaced that in August 2026: stat growth turned out not to be a reliable proxy for late-game strength, so scaling is now measured directly from win rate over game length rather than inferred. The change was applied to the stored history as well, so trends on meta movers stay comparable across that date.

How counter scores are read

A counter score is the difference in win rate when two heroes meet, in win rate points. A score of +3.20 means the hero wins that matchup by 3.2 points more than they win in general.

Those figures are not all measured equally well. A matchup against a hero almost nobody picks rests on very few games, and at face value a loud number from a rare hero would outrank a solid one from a common opponent. So each score is scaled by how much evidence sits behind it, and matchups thin enough that the exact figure should not be leaned on are labelled thin data where they appear. The direction of a thin matchup is usually right; the precise number is not.

Lane pages, such as a hero's EXP Lane or Jungle counters, use the same scores filtered to heroes who play that lane. Lane assignments are the ones Mobile Legends publishes for each hero, so a lane matchup here means both heroes are commonly played in that lane, not that the game recorded them facing each other there.

Update frequency & accuracy

The scraper re-collects all hero data regularly and the whole site (hero pages, tier list, sitemap, and llms.txt) re-renders from the fresh database, so pages always reflect the most recent scrape (dated on every page). Because stats shift with each patch and emblem/item changes, treat scores as a current-meta snapshot, not a permanent verdict on any hero.

Found something that looks wrong? The fastest fix is re-checking after the next data refresh; persistent issues are usually corrected upstream within a patch cycle.