Overwatch Season 2 "Summit" halfway meta snapshot: what's changed in 5 weeks

Five weeks into Overwatch Season 2 "Summit" and the tier list has done something rare: barely moved. The same names that led each role at the four-week snapshot are still leading now, with a few percentage-point nudges and one notable flip at the top of Support. Here's the full picture from today's live community data.

---

Five weeks in: the promised snapshot

Season 2 "Summit" launched 2026-04-14 and runs through 2026-06-16. We're now at the halfway point. The first-month snapshot published 2026-05-12 flagged three things to watch: whether Torbjörn's 56% edge would survive a patch, whether the compressed Tank role would sort out, and where the Sierra/Venture/Mei Damage cluster would land. This article has the answers.

Short version: the top-five composition per role is almost identical to what it was a week ago. The order shifted slightly in two roles, Mei dropped out of the Damage top 5, and win rates drifted by fractions of a point. No hero stormed into the top 5 from outside it. Summit has been a stable season, which is either a sign of unusually good balance or a sign that no mid-season adjustment has landed hard enough to change the shape of the leaderboard.

All numbers below are shrunk win rates (5V5, all game types, all ranks, aggregated across all maps) from the current daily-refresh dataset.

Tank: Sigma climbs past Hazard, compression holds

The Tank top 5 at five weeks:

• Reinhardt: 52.47% across ~94,700 matches
• Sigma: 52.03% across ~91,700 matches
• Hazard: 52.00% across ~39,100 matches
• Domina: 51.85% across ~62,000 matches
• Junker Queen: 51.81% across ~35,000 matches

The composition is unchanged from the first-month snapshot. The movement: Hazard and Sigma swapped. A month ago, Hazard was the surprise #2 at 52.1%, sitting level with Sigma and just behind Reinhardt. Now Sigma has pulled ahead at 52.03% on ~91,700 matches vs. Hazard's 52.00% on 39,100. The larger sample behind Sigma's number makes that edge stickier than the decimal suggests.

As for whether the compressed Tank role would sort out: not yet. All five are still within 0.66pp of each other. Reinhardt holds at 52.47% on nearly 95,000 matches, the steadiest placement in the whole tier list, but the gap between him and #2 Sigma is 0.44pp, smaller than most seasons' #1/#2 spread. Domina drifted down from 51.9% to 51.85% and JQ from 51.8% to 51.81%, both minor moves as the sample grew and shrinkage did less lifting. JQ at top-5 across five weeks remains notable for a 5V5 season where solo-tank pressure has historically punished her engage-and-recover playstyle.

Damage: Torbjörn holds, Mei drops out

The Damage top 5 at five weeks:

• Torbjörn: 55.58% across ~39,900 matches
• Symmetra: 53.29% across ~27,500 matches
• Pharah: 52.79% across ~43,600 matches
• Sierra: 52.04% across ~141,400 matches
• Venture: 52.03% across ~25,700 matches

Torbjörn's 56% edge did not fully survive: his win rate dropped from 56.1% to 55.58%, about half a point. That is not a collapse. A shrunk win rate above 55% on ~40,000 matches is still the most dominant single-hero number in the game right now, and the gap over #2 has barely narrowed. If mid-season balance adjustments touched him, they did not register meaningfully in the aggregate.

Symmetra slipped from 53.6% to 53.29%. She's still clearly #2 with more than 3pp separating her from the heroes below, and the passive chip-damage story that put her and Torbjörn at the top of the role early in the season has not changed.

The most interesting move is Pharah. She went from 52.7% to 52.79%, the only top-5 Damage hero to gain ground. With ~43,600 matches behind that number it is not noise, and it opens a real gap between her and the Sierra/Venture cluster below.

As for the Sierra/Venture/Mei question from the first-month snapshot: Sierra and Venture held their #4 and #5 spots. Mei dropped out of the Damage top 5 entirely. A month ago she was essentially tied with Sierra and Venture at 52.1% on ~66,000 matches. Now Venture holds #5 at 52.03% and Mei is out. The cluster did not consolidate: Mei slipped while the other two held.

Support: Zenyatta takes the top spot

The Support top 5 at five weeks:

• Zenyatta: 53.56% across ~101,900 matches
• Illari: 53.55% across ~91,600 matches
• Brigitte: 52.32% across ~43,700 matches
• Lúcio: 51.66% across ~61,700 matches
• Mizuki: 51.38% across ~123,500 matches

The headline swap: Zenyatta is now #1 and Illari at #2. A month ago the order was reversed, Illari at 53.7% and Zenyatta at 53.6%. Now Zenyatta leads 53.56% to 53.55%, a single basis point. With over 100,000 matches behind Zenyatta's number and 91,600 behind Illari's, this is not sample noise, but the gap is not meaningful in practice. Treat them as functionally tied: either is the right top Support pick for this season.

Brigitte rose from 52.2% to 52.32% while the heroes below her drifted down. She now sits a full percentage point above Lúcio, which is a real tier boundary. Lúcio at 51.66% and Mizuki at 51.38% each dipped a fraction from the first-month numbers (51.7% and 51.6% respectively). Mizuki's 123,500-match sample makes his #5 placement very confident even if the absolute number is modest.

Who rose, who fell, who held

Net risers: Pharah (52.7% to 52.79%), Sigma (climbed from #3 to #2 in Tank), Brigitte (52.2% to 52.32%), Zenyatta (flipped to #1 Support from #2).

Net fallers: Torbjörn (56.1% to 55.58%, still dominant but off peak), Hazard (dropped from #2 to #3 Tank), Illari (slipped from #1 to #2 Support), Symmetra (53.6% to 53.29%), Mizuki (51.6% to 51.38%), Mei (fell out of the Damage top 5 entirely).

Held their spots: Reinhardt (#1 Tank throughout), Domina (#4 Tank), Junker Queen (#5 Tank), Sierra (#4 Damage), Venture (#5 Damage), Lúcio (#4 Support).

The biggest finding: the top-5 composition is identical in Tank and Support, and in Damage only Mei changed. Summit has not had a mid-season swing.

What to watch before Season 3

Season 3 starts 2026-06-16. Four weeks of data still to come. The final snapshot will publish in early June before the season closes.

Torbjörn's lead. The dip from 56.1% to 55.58% is mild, but the trajectory is slightly downward. If mid-season balance adjustments catch him in the final stretch, the closing snapshot will look very different. If his edge holds through June, this is the strongest "learn this hero now" signal Counterwatch has tracked for the Damage role in recent memory.

The Zenyatta/Illari gap. One basis point is noise. The final snapshot will settle which Support technically owned Summit, though for practical pick decisions both are the correct answer at the top of the role.

Tank compression. All five Tanks are within 0.66pp. One balance adjustment to any of the top three could reshuffle the whole list before the season ends. If nothing changes, this might go down as the most balanced Tank pool of any 5V5 season in recent Overwatch history.

For the live data at your specific rank, filter the full tier list to your division. The All-rank aggregate above smooths across skill levels and can hide rank-specific metas. The best one tricks page adds consistency, matchup breadth, and confidence signals on top of win rate, and often produces a different ordering than raw WR alone. The Counterwatch app runs these numbers live in your match: when the enemy team locks in, the in-game overlay scores your current lineup against the live enemy comp using the same community data you just read.