Overwatch tier list explained: how Bayesian shrinkage ranks every hero

A tier list is only as useful as you understand what it measures. Every row on the Counterwatch Overwatch tier list is sorted by "shrunk" win rate, not raw win rate. This post walks through why that distinction matters, how the math keeps the list honest when some heroes barely get picked, and how to actually use the list when you queue up for ranked.

---

What the tiers actually measure

The Overwatch tier list sorts every hero S through F based on their shrunk win rate, computed across community-tracked matches at your chosen mode, rank, and game type. Not raw win rate. Shrunk.

S-tier means a hero is beating the 50% baseline by a comfortable margin with enough matches behind them to trust the number. F-tier means the opposite: enough matches to be confident the hero is underperforming at that rank.

Middle tiers (A, B, C, D) are where most of the roster lives. That is the reality of a game balanced around "everyone is playable": the gap between the median damage hero and the S-tier damage hero is not huge at most ranks. Your individual mastery on a hero usually matters more than a one-tier delta.

What Bayesian shrinkage actually does

Here is the small-sample problem. Imagine a hero with five matches this week at a 100% raw win rate. Naive rankings would put them at the very top of the list. Obviously that is not real: regression to the mean will eat that number the second the sample grows.

Bayesian shrinkage fixes it. Every hero's raw win rate gets blended with a neutral 50% baseline, weighted by how many matches feed the sample. Low-sample heroes get pulled almost all the way to 50%. High-sample heroes barely move.

Concretely, we add 400 imaginary 50/50 matches to every hero's record before computing the win rate:

• 5 games at 100% raw → shrunk to (5 × 1.0 + 400 × 0.5) / (5 + 400) = 50.6%. Basically the baseline. That 100% fluke does not land S-tier.
• 10,000 games at 55% raw → shrunk to (10,000 × 0.55 + 400 × 0.5) / 10,400 = 54.8%. Barely moves. Real signal survives.

The 400-match prior weight is tuned so that hero slots with fewer than ~100 tracked matches flatten to near-baseline, while hero slots with thousands of matches land within a fraction of a percent of their raw win rate. A hero sitting at 60% across 15 matches reads identically to a hero at 50.7% once shrinkage runs.

If you want to rank S-tier, you need both a strong raw win rate and enough community data behind you to trust it. That is the whole point.

Sample sizes and freshness

The shrunk win rate on any tier list row is computed across a rolling recent window of community-tracked matches. The website aggregates refresh once a day in the early UTC morning, so today's tier list reflects yesterday's matches.

The minimum threshold for display is low: we show every hero even when the sample is thin. But thin-sample heroes get shrunk aggressively, so a three-match fluke never floats to the top of S-tier.

Every hero page shows the underlying match count. Before you take a placement at face value, click through and look: tens of thousands of matches = confident, a few hundred = provisional. A provisional placement can flip on any given day once the sample grows.

Per-rank filtering: the biggest lever most players miss

The Overwatch meta at Bronze is not the same as the meta at Grandmaster. Execution skill, game sense, and coordination all scale with rank, so heroes that demand precise aim or tight team coordination shift tiers dramatically as you climb.

The rank filter on the full tier list goes from "All" through every individual tier: Bronze, Silver, Gold, Platinum, Diamond, Masters, Grandmaster, Champion, plus a composite "Grandmaster+" that combines the top two.

Most players default to "All" and never realise how much their own rank's tier list differs from the aggregate. If you are in Silver, the Silver tier list is the one that matters for your climb. "All" smooths across every skill level and often tells you what works on average, not what works for you.

The same page also filters by game type (Ranked vs Unranked, plus Stadium variants) and by mode (5V5, 6V6, Stadium). Set all three to match what you actually queue for. Otherwise you are reading a tier list that includes matches you will never play.

How to actually use a tier list while climbing

A tier list is a starting point, not gospel.

Every row is a snapshot of aggregate win rates across every map, every match, every lineup. It does not know which map you just loaded into, which comp the enemy locked, or which hero you have 800 hours on. Those three factors dominate your actual win rate on a hero-by-hero basis more than any one-tier delta does.

Concrete recommendations:

• Don't abandon mastery for a tier bump. If you have deep hours on a B-tier hero and you are comfortable on them, the individual skill gap beats a one-tier generic upgrade.
• Click through to the hero page. Every tier list row links to that hero's counters, synergies, and per-map win rates. That is where mid-queue decisions actually happen.
• Cross-reference with best one tricks. The tier list answers "what is strong in aggregate." The best one tricks page answers "which hero rewards mastery." For a climbing main, the second question usually matters more.
• Use the Counterwatch app mid-match. The in-game overlay shows live hero scores against the enemy lineup as picks lock in, so you don't have to alt-tab and dig through filters while you are in spawn.

For the full write-up on how every number on the site is computed: shrinkage, rolling windows, per-rank filters, data source: see our methodology page.