Overwatch team composition guide: synergy, balance, and win rate

An Overwatch team comp that works is not a list of five meta heroes. It is a lineup where the roles cover each other, the archetypes match the map, and the hero picks synergise instead of competing for the same space. This post walks through the three classic archetypes, the synergy vs counter trade-off, and how Counterwatch's team builder actually scores a lineup.

Role balance is the floor, not the ceiling

Overwatch 5V5 is built around one tank, two damage heroes, and two supports. Every rules-of-thumb about team composition starts from that: if your tank dies, your frontline collapses. If neither of your damage heroes can hit the enemy supports, fights feel hopeless. If your supports can't keep the tank alive, you lose the engagement before it starts.

Role balance is just the floor. Every standard queue composition already satisfies it. The ceiling: the thing that separates a comp that works from a comp that fights itself: is the archetype layer on top.

The three archetypes: Dive, Poke, and Brawl

Every 5V5 composition in Overwatch sits somewhere on a spectrum between three archetypes:

Dive: fast heroes who collapse on a single target before the enemy team can react. Tank picks: Winston, D.Va, Doomfist, Wrecking Ball. Damage: Genji, Tracer, Sombra, Soldier. Supports: Lucio, Kiriko, Mercy. Dive wins when the map has flanks, high ground, and vertical options.
Poke: long-range damage dealers who farm ultimates from mid-range and force the enemy into bad positions. Tank picks: Sigma, Orisa, Ramattra. Damage: Widowmaker, Hanzo, Ashe, Pharah. Supports: Ana, Baptiste. Poke wins on open maps with clear sightlines.
Brawl: high-health, close-range fighters who win 50/50 engagements through sustain and area-denial. Tank picks: Reinhardt, Junker Queen, Roadhog, Mauga. Damage: Mei, Reaper, Torbjörn. Supports: Brigitte, Moira, Lifeweaver. Brawl wins on chokes and tight indoor maps.

A comp that mixes archetypes too hard fights itself. A Winston tank wants to dive and leave; a Widowmaker DPS wants to hold a sightline and stay; Ana wants to stand back and scope. Three different agendas, no shared engagement window. The team loses fights because they are never actually on the same page.

The fix is archetype coherence: pick heroes that all want to be in the same type of fight, then let the map decide which archetype to run.

Synergy vs counter: the trade-off you can't escape

Here is where most players mis-pick: you cannot max synergy and counter at the same time.

Synergy is about your team's internal cohesion. A Junker Queen plus Mercy pair means Mercy's damage boost is amplifying Queen's axe swings while the axe pulls pull enemies into Queen's shotgun range. The pair wins fights partly because each hero makes the other hero more dangerous. The current community win rate on that pair is 56.9% across roughly 6,000 tracked matches in 5V5: one of the cleanest synergy numbers in the game right now.

Counter is about beating what the enemy picked. A Reaper against a Winston dive is a classic "I saw your tank and I'm swapping to shut you down" move.

The problem: high-synergy picks and high-counter picks often point at different heroes. If you swap to counter, you may break your team-up. If you stay synergised, you may walk into a bad matchup. The team builder's scoring acknowledges this: no hero wins on both dimensions at once, so the ranking always reflects a trade-off.

How the team builder scores a lineup

Counterwatch's team builder combines three inputs into a single win-chance prediction:

Counter score (50%): how each friendly hero performs against the enemy lineup, aggregated across all five matchups.
Synergy score (30%): how each friendly hero performs alongside the other friendly heroes, aggregated across the four same-team pairings.
Map performance (20%): how each friendly hero performs on the selected map, weighted by their sample size on that map.

Those weights are tuned against thousands of tracked matches. The counter term dominates because counter-picking is empirically the highest-leverage decision in a match. Synergy contributes meaningfully but less: strong synergies matter, but a comp that perfectly synergises against the wrong enemy still loses. Map performance is the smallest weight because map effects are weaker signals than hero matchups, and many players don't optimise by map anyway.

Three synergy examples from live community data

Real numbers from the current 5V5 data at all ranks, all game types:

Roadhog + Kiriko: 56.0% over ~14,500 matches

The highest-sample synergy pair in the current data. Roadhog's problem is that he trades too much health on any engagement and dies the moment a second cooldown is used against him. Kiriko's Suzu cleanses the one thing Hog can't escape on his own (crowd control, burst damage, anti-heal), and Teleport lets her get into Suzu range when he over-extends. The synergy is "one hero papers over the other's single hardest weakness." Huge sample means you can trust this number.

Doomfist + Sojourn: 56.3% over ~4,400 matches

A dive archetype comp. Doomfist's kit is built around pinning a single target long enough to kill them; Sojourn's right-click rail erases low-health targets Doom has already chipped. Doom closes and locks down a priority pick; Sojourn finishes. The synergy is temporal: Doom creates the opening, Sojourn converts it in the same two-second window. Comps built around this pair want dive-friendly maps with flank routes.

Junker Queen + Mercy: 56.9% over ~6,000 matches

The top synergy in the current data. Mercy's damage boost pushes Queen's axe swings and shotgun breakpoints into one-shot territory against squishies, while Queen's team-pull ultimate creates clustered targets Mercy's boost can capitalise on. Queen's self-sustain from bleed damage also reduces Mercy's healing load, letting Mercy stay in damage-boost mode longer. Brawl archetype: works on chokes and close-range maps.

When to break your comp

The team builder gives you a current best lineup. But locking in a "mathematically optimal" comp and never swapping is worse than a thoughtful mid-match adjustment. Two moments where breaking your comp pays off:

Enemy picks a hard counter to one of your damage heroes. Taking a 1% comp-score hit to swap into a better matchup is usually a win.
Enemy locks a full brawl comp on a poke map. Your synergy number drops slightly if you adjust, but the map-weighted factor gives you more than you lost.

The rule is: the team builder's score is the average of a lot of real matches. If a specific adjustment clearly helps against what's in front of you, the specific beats the average.

Mid-queue vs mid-match

The team builder is the right tool before you load in or during role queue. It shows you full-lineup predictions and swap suggestions. The Counterwatch app takes over from there: the in-game overlay runs in real time as enemy picks lock, so synergy and counter updates happen on the fly without alt-tabbing.

For the long-form on the scoring formula, shrinkage, and how the individual terms are weighted, see the methodology page.