Overwatch 2 Tier List — Season 3
Season 3 — Reign of Talon: Into the Tiger's Den has been live since 16 June 2026, with a midseason content drop on 14 July. Competitive Role Queue is 5v5; 6v6 exists only in Open Queue and the Quick Play Hacked experiments, so if a tier list is ranking heroes for 6v6 ranked, it is describing something that is not the default mode.
The thing that makes most Overwatch tier lists useless is that there are two of them. Ladder win rate says Torbjörn and Symmetra are the two best damage heroes in the game. Expert lists say Tracer and Sojourn. Both are right. Torbjörn punishes players who walk into a turret; Sojourn rewards aim you may not have. The win rates below come from Counterwatch's Season 3 5v5 all-ranks board, read on 31 July 2026, and every place where Mobalytics (17 July) or Dexerto (31 July) disagrees is called out in the row.
Use it this way: if you are climbing below Diamond, weight the win-rate column. If you are already there, weight the expert callouts. Twelve heroes on this page sit at least two tiers apart between the two lenses, and those are the rows worth reading properly.
Tank Tier List
Fourteen tanks. One slot, so this pick shapes the game more than any other.
| Tier | Hero | Ladder win rate | Why it sits there |
|---|---|---|---|
| S | Reinhardt | 53.7% | The highest tank win rate on a huge 27% pick rate. Shield and hammer still beats uncoordinated ladder fights. Mobalytics puts him at B, arguing he cannot find value against a mobile meta. |
| S | Junker Queen | 53.2% | 53.2% after the 14 July shout cooldown buff, and wound pressure carries brawls. The widest split on the page: Dexerto has her at D and Mobalytics at C, both pointing at Kiriko's Suzu. |
| A | Sigma | 51.7% | 51.7% on a 21% pick rate; the 14 July shield health nerf did not dislodge him. All three sources agree he is strong. |
| A | Wrecking Ball | 51.9% | Above average, but a specialist. Rewards a skilled pilot and asks almost nothing of your team. |
| A | Domina | 51.5% | The newer barrier-array tank at 51.5%. Area control that people are still learning, which usually means the number rises. |
| A | Hazard | 51.5% | 51.5% with good eliminations per ten minutes. Not free to pilot, but it pays back. |
| B | Mauga | 50.6% | Oppressive in a tank duel, then folded by anti-heal or a blocked Cardiac Overdrive. |
| B | Doomfist | 50.5% | The 14 July punch collateral nerf did not help, and ladder crowd control still punishes him. Dexerto has him at D. |
| B | Winston | 50.2% | A dive enabler who needs follow-up, which makes him a worse solo-queue pick than his number suggests. |
| C | Zarya | 49.4% | Split: under 50% on ladder, but S on both Mobalytics and Dexerto — bubbling a Tracer or Vendetta is meta-defining when your team can use it. |
| C | Ramattra | 49.2% | Nemesis form is kiteable. The 14 July vortex and block buffs may still lift him. |
| D | Roadhog | 46.8% | Feast or famine on the hook. Mobalytics still rates him a ladder terror after the bruiser buffs; the win rate disagrees. |
| D | Orisa | 45.9% | A Fortify-reliant anchor having a bad patch. |
| D | D.Va | 45.9% | The starkest split in the game: bottom of the ladder board, yet S on both expert lists for Matrix plus the Initiator passive. Do not read either number alone. |
Damage Tier List
Twenty-four damage heroes, and the widest gap between what wins ladder games and what experts recommend.
| Tier | Hero | Ladder win rate | Why it sits there |
|---|---|---|---|
| S | Torbjörn | 55.7% | The best win rate of any hero. Turret and Overload punish careless flankers, which is most of them below Diamond. Experts rate him mid — both readings are correct at different ranks. |
| S | Symmetra | 54.7% | Second best on ladder. Teleporter utility plus a beam that shreds tanks; map and choke dependent, brutal when the map suits her. |
| A | Mei | 53.0% | Walls isolate tanks and win scrappy rush fights. |
| A | Venture | 52.2% | 52.2% at a low pick rate. Assassination plus Burrow stalling rewards aggression. |
| A | Anran | 51.9% | 51.9% on a 20% pick rate, still climbing as the combo kit gets learned. |
| A | Pharah | 51.8% | Vertical maps and a Mercy pocket still farm the lower half of the ladder. |
| A | Ashe | 51.6% | High pick rate, consistent midrange poke, Dynamite for chip. One of the safest picks on this list. |
| A | Junkrat | 51.5% | Spam and traps punish bad spacing, and bad spacing is what solo queue is made of. |
| B | Genji | 51.3% | High ceiling, needs peel and a composition. Feast or famine. |
| B | Vendetta | 51.3% | Still lethal on the flank after nerfs, at only a 7% pick rate. Dexerto has him at D. |
| B | Tracer | 50.6% | Split: middling on ladder, S on both expert lists after the projectile size buff. A skill check, not a free climb. |
| B | Shion | 50.5% | The new Season 3 damage hero at a 35% pick rate — everyone is trying him and nobody has solved him yet. Mobalytics already rates him S. |
| B | Reaper | 50.5% | The 14 July lifesteal buff helped; closing range into armour is still the problem. |
| B | Sierra | 50.0% | A 31% pick rate on a flexible hitscan that is good at everything and best at nothing. |
| C | Hanzo | 49.8% | Projectile feast or famine against hitscan consistency. |
| C | Soldier: 76 | 49.4% | Easy to pilot with a high pick rate, and a lower ceiling than Ashe or Sojourn. Mobalytics and Dexerto both rate him A for exactly that ease. |
| C | Echo | 49.4% | Strong alongside dive, badly exposed to hitscan. |
| D | Bastion | 48.4% | A niche barrier and Domina shredder. Dexerto rates him S, which overweights that niche — treat with care. |
| D | Emre | 48.3% | Entry-level hitscan consistency; Mobalytics notes Override Protocol stomps the lowest ranks even though the overall number is poor. |
| D | Sombra | 48.1% | A counter-pick to Ball and Doomfist, rare above that purpose. |
| D | Cassidy | 48.0% | A 33% pick rate on an anti-dive comfort pick, and the Silver Bullet perk was nerfed on 14 July. |
| D | Freja | 47.7% | Versatile ranger; the 14 July buffs may not have landed in the numbers yet. |
| D | Widowmaker | 47.6% | Map-locked. Oppressive on long sightlines and irrelevant everywhere else. |
| D | Sojourn | 45.2% | Split: near the bottom on ladder, yet experts call her the most versatile hitscan in the game. This is an aim check in its purest form. |
Support Tier List
Fourteen supports, including the most-picked hero in the game.
| Tier | Hero | Ladder win rate | Why it sits there |
|---|---|---|---|
| S | Zenyatta | 53.8% | The best support win rate at a 23% pick rate. Discord Orb melts tanks whenever nobody dives him. Mobalytics has him at B, Dexerto at C. |
| A | Brigitte | 52.4% | Anti-dive peel, improved again by the 14 July instant-heal change to Inspire. Strong on all three lists. |
| A | Mizuki | 52.2% | 52.2% on a 36% pick rate, and Mobalytics rates her S. One of the season's defining supports. |
| B | Wuyang | 51.2% | A solid middle pick. |
| B | Illari | 51.0% | Pylon plus poke carries when your aim holds up. |
| B | Lúcio | 50.8% | Speed enables every rush composition, which is why experts often rate him S next to Kiriko. |
| B | Jetpack Cat | 50.5% | Permanent flight, half utility and half assassin. Dexerto rates it S. |
| B | Juno | 50.5% | A 35% pick rate on an orbital support that experts rate above its ladder number. |
| C | Moira | 49.6% | A 47% pick rate — the easiest support to survive on, and the hardest to win a fight with. The classic comfort pick. |
| C | Mercy | 49.4% | A 40% pick rate on a pocket enabler whose area healing struggles against the damage passive. |
| C | Lifeweaver | 49.3% | Grip and Platform are real tools attached to thin all-round utility. |
| C | Kiriko | 48.5% | The pick rate paradox: most-picked support in the game at 67%, a sub-50% win rate, and S on both expert lists. Suzu still decides fights; the average Kiriko does not use it well. |
| D | Ana | 48.0% | A 51% pick rate on the highest skill ceiling in the role, punished by dive. Aspirational rather than a climb tool. |
| D | Baptiste | 47.5% | Immortality Field is worth a lot; the long cooldowns and dive vulnerability cost more. |
Unlock Competitive mode →That is the gate, and it has nothing to do with the heroes above. A PRO clears it so your season starts in placements instead of in Quick Play.
The Twelve Heroes the Sources Fight Over
These are not errors in either source. They are two honest answers to two different questions.
| Hero | Ladder | Experts | What it actually means |
|---|---|---|---|
| D.Va | 45.9%, bottom | S | Defence Matrix is the highest-value ability in the game in the hands of someone who tracks cooldowns, and nearly worthless held down at random. No hero on this page has a wider skill spread. |
| Sojourn | 45.2%, bottom | A to S | Pure aim check. If Widowmaker is not comfortable for you, neither is she. |
| Zarya | 49.4% | S | Her value is in enabling somebody else. In solo queue there is often nobody to enable. |
| Kiriko | 48.5% at a 67% pick rate | S | The clearest case of pick rate not being skill. Suzu wins fights when timed; most of those 67% are not timing it. |
| Tracer | 50.6% | S | The projectile size buff made her genuinely strong at the top and no easier below it. |
| Junker Queen | 53.2%, S | C to D | The reverse case. She wins ladder games and experts think Suzu answers her, which matters far more in coordinated play than in yours. |
| Torbjörn and Symmetra | 55.7% and 54.7%, the top two | Mid | They beat players, not teams. Which is exactly what you are facing while climbing. |
One source to treat carefully: Dexerto rates Bastion S, which overweights a narrow barrier-shredding niche, and its roster is incomplete — Anran is missing entirely.
Best Heroes for Climbing Below Diamond
Below Diamond, the heroes that climb are the ones with individual impact that does not require your team to cooperate.
- Tank: Reinhardt, Sigma, Domina. High win rate and legible. Leave D.Va and Zarya until you have a duo who follows up.
- Damage: Torbjörn, Ashe, Junkrat, Soldier: 76, Reaper, Emre. Every one of them punishes positioning mistakes rather than demanding perfect aim.
- Support: Moira, Brigitte, Mizuki, Lúcio. Moira in particular is the safest way to stop dying; understand that she wins fewer fights than she survives.
Heroes to postpone, not avoid: Sojourn, Widowmaker and Ana all have higher ceilings than anything above. They are the reward for having climbed, not the method.
Placement matches boost →Placements set the rank you spend the season climbing out of. Played on the heroes that actually win at your bracket, not the ones the pros are on.
Ranks, Divisions and How to Unlock Competitive
Nine tiers in Competitive Role Queue: Bronze, Silver, Gold, Platinum, Diamond, Master, Grandmaster, Champion, then the regional Top 500 leaderboard.
Division numbering trips people up, including publishers. Within every tier, Division 5 is the lowest and Division 1 is the highest — so Gold 5 is where you arrive and Gold 1 is where you leave. At least one widely-read guide states the opposite; it is wrong.
Tank, Damage and Support ranks are tracked separately in Role Queue. Open Queue carries its own single rank.
Getting in
Win 50 Quick Play games to unlock Competitive. Then 10 placement matches per role, with a predicted rank shown along the way and your real rank set after the tenth.
How the rank moves
Progress is per match now, with modifiers for win streaks, uphill battles, consolation and loss streaks, plus demotion protection the first time you would fall out of a tier. Sit out a full season and your rank hides until you play five games. The older “recalculates every 5 wins or 15 losses” cadence is still widely quoted — it may well still be accurate, but no 2026 Blizzard source confirms it, so check in client rather than planning around it. Rank decay was removed back in April 2023, though a soft reset at season start can still drop your visible rank a division or two while your matchmaking rating is kept.
Stadium Mode
Stadium is still live and still getting seasonal balance, rank resets and rewards. What stopped is content expansion — no new heroes or maps. Reports that support has ended overstate what Blizzard actually said.
Its ladder is separate: Rookie, Novice, Contender, Elite, Pro, All-Star, Legend, each with divisions counting down from 5 to 1, tracked per role. It does not require a standard-mode rank to enter, whatever you may read elsewhere.
Stadium uses a curated hero pool, not the full roster — several newer standard heroes are absent. Strongest picks there as of 17 July: Hazard and Junker Queen among tanks; Ashe, Junkrat, Freja and Cassidy for damage; Jetpack Cat, Zenyatta and Moira for support. The theme is early cash generation, burst damage, and utility that cuts through Stadium's inflated health pools.
Sources and Methodology
Counterwatch — Season 3, 5v5, all ranks, read 31 July 2026. Source of every win rate quoted above. Ladder win rate is not high-rank or professional meta, and it flatters heroes that beat inexperienced opponents.
Mobalytics — Role Queue tier list, updated 17 July 2026, plus the separate Stadium board of the same date. Expert lens.
Dexerto — editorial list, updated 31 July 2026. Same-day fresh but the roster is incomplete and a few placements look like niche overweighting, so it is treated as a third opinion.
Blizzard official — Season 3 launch post (16 June 2026), midseason post (14 July), the Director's take of 15 July for the 5v5 share of play and Stadium's status, and the retail patch notes of 14, 28 and 30 July.
Methodology and limits. Placement follows the win-rate board; every case where the expert lists differ by two tiers or more is named in the row rather than averaged. Season 3's end date is not officially announced — around 11 August is widely reported from the battle pass timer, and this page does not treat it as confirmed. The 30 July patch changed Quick Play Hacked and Open Queue 6v6 values rather than the ranked ladder, so it does not move these tiers. Published rank distribution figures still trace to a 2025 Blizzard post and may have drifted. This page describes Season 3 and will be revisited when Season 4 begins.
FAQ
Overwatch rank boost →Tank, Damage and Support rank separately, so pick the role you actually want to play and have the other two handled. Piloted or duo.