Marvel Rivals counters: who beats who in every matchup

Counter-picking in Marvel Rivals is trickier than in most hero shooters because team-ups, new heroes every season, and role flexibility all warp the matchup picture. This post walks through how to read Counterwatch's MR counter data, five live matchups from the current community data, and when to actually swap.

---

What counts as a counter in Marvel Rivals

A counter is a hero whose community win rate against a specific opponent sits meaningfully above 50%. Not "the internet says so": an actual tracked-match number from opted-in Counterwatch app users.

Every matchup number on the MR counters page is shrunk, not raw. Shrinkage pulls low-sample matchups toward 50%, so a hero showing 58% over 80 matches collapses back toward neutral. By the time a matchup settles at 53% with thousands of tracked games behind it, the edge is real signal.

Rough thresholds after shrinkage:

• 52-53%: a meaningful edge. Prefer this pick if both heroes feel equally comfortable.
• 53-55%: a strong counter. Worth swapping for even if you are less practiced.
• 55%+: rare in MR's aggregate data. When you do see it, the sample is usually small, so check match count before celebrating.

MR edges cluster tighter than Overwatch's because the 6v6 format dilutes any single matchup's impact. You are fighting five other heroes on the enemy team and four on your own. A 53% edge against one opponent still moves the needle, but less dramatically than the same edge in a 5V5 game would.

Team-ups make the number move

This is the thing Overwatch doesn't have and MR players need to internalise: team-ups change counter numbers. When a Duelist is on a lineup with their team-up partner active, their win rate against specific opponents shifts. The counters page aggregates across every team-up state, so the number you see is a weighted average of "team-up active" and "team-up inactive" matches.

If you are hunting a specific counter hero's maximum edge, check the synergies tab for that hero too. The synergy data will show you which partners spike the hero's win rate. If one of those partners is already on your team, that hero's real counter edge is higher than the aggregate suggests.

Five worked examples from live MR competitive

Real matchup numbers from the current community data at All ranks, Standard mode, all game types. Picked for role diversity and for matchups that either have huge samples or teach a lesson that runs against common framing.

1. Black Panther into The Thing: 52.6% over ~14,100 matches

Duelist counters Vanguard. Black Panther's dash-reset kit eats stationary Vanguards alive: every elimination refreshes his mobility, and The Thing's positional commitments give Panther clean dive windows. The Thing's crowd control is potent, but by the time he lands it, Panther has usually already tagged a squishy and reset out. One of the highest-sample counter edges in the current MR data, which makes this a confident read.

2. Iron Man into Hela: 52.4% over ~17,800 matches

Duelist mirror, and the highest-sample matchup in the current top 30. Iron Man's flight lets him dictate sightlines against Hela's precision kit, forcing her to shoot upward at a moving target rather than picking off grounded heroes from her comfortable sightline. Hela out-damages Iron Man in a clean peek duel, but Iron Man rarely gives her clean peeks: the mobility asymmetry is doing the work.

3. Black Panther into Namor: 52.3% over ~15,500 matches

Namor is usually framed as an anti-dive pick because his turrets punish divers who commit to a squishy target. The aggregate data disagrees. A skilled Black Panther sequences the turret kill first (one reset), then the Namor kill (second reset), and the turret economy never catches up. At lower ranks Namor holds better because turret placement is more punishing to unpractised divers; at higher ranks Black Panther's edge widens.

4. Storm into Thor: 52.4% over ~6,500 matches

Another Duelist-counters-Vanguard matchup, this one driven by vertical control. Storm's flight plus her AoE damage punish Thor's forward-commitment playstyle: every time he dashes in, she reposts vertically and he eats damage while recovering. Thor's crowd control can turn a fight if he lands it, but Storm's mobility profile makes "landing it" a coin flip most of the time.

5. Mantis into Human Torch: 52.8% over ~1,900 matches

A Strategist counter. Mantis's sleep is one of the best single-target lockdowns in the game, and Human Torch's flight-based damage dealing means he is often the one Duelist on the enemy team who cannot escape a sleep without cooldown. Sleep timing turns Torch from an uncatchable nuisance into a 40-damage-per-second free elimination. Sample size is lower than the others on this list, so expect the edge to move as more matches roll in.

Common counter-pick mistakes

Reading the matchup number is half the skill. Knowing when not to act on it is the other half.

• Swapping on fantasy, not data. "I always pick this hero against dive" is a habit. Check the number before locking.
• Ignoring your own team-up partners. Swapping to a counter breaks synergy if you were already unlocking a team-up. Sometimes holding your current pick is worth more than the counter nudge.
• Swapping mid-round without a read. Enemy comps swap too. Lock counters after they commit to picks, not during selection.
• Ignoring map context. Some MR maps favor flight heavily (Klyntar for example), others punish it. Matchup numbers can swing by map: check the hero's per-map splits on their detail page.

Mid-match vs pre-queue

Different tools for different moments:

• Pre-queue planning: the MR team builder scores full lineups, combining counters, synergies, and team-up unlocks into a single win-chance prediction. Useful for stacks locking a comp before queuing.
• Mid-match swap calls: the Counterwatch app runs an in-game overlay on top of your MR match. It surfaces counter candidates as the enemy team locks picks, so you do not have to alt-tab and dig through the website in the spawn room.

A 53% edge looks small on its own. Stack three or four of those across a lineup, keep your team-ups intact while you do it, and the aggregate math tilts the fight before anyone presses a button.

For the long-form on how we collect, shrink, and refresh these matchup numbers, see the methodology page.