How is this Marvel Rivals tier list calculated?

It is computed from the official Hero Hot List on marvelrivals.com, which publishes each hero's win rate and pick rate. Heroes are ranked by win rate relative to the roster average, with rarely-played heroes pulled toward that average because their win rates are less reliable. Tier boundaries are fixed distances from the average, so a balanced patch genuinely produces no S tier instead of inventing one.

Can I make my own Marvel Rivals tier list?

Yes. Just drag any hero into a different tier — the list becomes yours as soon as you move anything, and picking a platform or rank button rebuilds the calculated list. The role buttons only filter what is shown, so they leave your list alone. On touch, tap a hero and then tap the tier you want it in. Your list is saved in your browser, and you can copy a share link or download it as an image.

Does the tier list change by rank?

Yes. The official stats are published per rank bracket, so you can view the tier list for Bronze, Silver, Gold, Platinum, Diamond, Master or Celestial and above, on PC or console. Hero strength differs meaningfully between low and high rank.

How the Marvel Rivals Tier List Is Calculated

Q: How often is the tier list updated?

It follows the official Hero Hot List, which the developers refresh each half-season rather than daily. The current figures are as of 7 July 2026, and that date is always shown on the page so you know how current the numbers are.

← Back to the tier list

The source

The official Hero Hot List publishes two numbers per hero: win rate and pick rate. There is no ban rate and no games-played count, so those two are all any honest calculation has to work with. The developers refresh it each half-season rather than daily; the current figures are as of 7 July 2026, and that date is shown at the top of the tier list so you always know how current the numbers are.

Win rate does the ranking

Each hero is measured against the roster's average win rate for the selected view — currently 50.35% for PC competitive. The number on each card is how far above or below that average the hero sits.

Pick rate is confidence, not power

A hero played in 3.0% of matches or more is taken at face value. Below that, their win rate is pulled toward the roster average, because a small sample swings wildly. This is why a rarely-played hero with a spectacular win rate doesn't automatically top the list — those cards are marked, and their tooltip shows both the raw and the adjusted figure. Adding pick rate to the score instead would just rank popularity, which is the most common flaw in stats-based tier lists.

Tier boundaries are fixed, not percentiles

S is +4.0 percentage points above average or better, A +1.5, B within -1.5, C -4.0, and D below that. Because the cuts are absolute, a balanced patch can genuinely produce an empty S tier — a percentile system would manufacture one every time.

Roles filter, they don't re-score

Picking a role narrows the same global ranking rather than re-ranking within that role, so a weak role doesn't get a flattering S tier of its own. It filters the board in front of you rather than building a different one, so it applies to a list you've arranged yourself just as it does to the calculated one — only platform and rank rebuild the list. Heroes too new to appear in the official stats are absent until the developers' next refresh.

The whole thing, in five steps

Everything above, as arithmetic. The two published numbers are a win rate $w_{i}$ and a pick rate $p_{i}$ , both percentages, so a hero's placement is these five steps over the $n$ heroes in whichever rank and platform you're viewing, and nothing else.

μ = \frac{1}{n} \sum_{i = 1}^{n} w_{i}

1. The roster average. The unweighted mean win rate across every hero in the view — currently 50.35% on PC competitive. Everything below is measured against it, so a hero is never scored in isolation.

c_{i} = \min (1, \frac{p_{i}}{3.0})

2. Confidence, from pick rate. A hero picked in 3.0% of matches or more gets

c_{i} = 1

and is taken at face value. Below that, confidence falls off linearly — pick rate is the only proxy for sample size the Hot List offers, and a 0.5%-pick win rate is mostly noise.

{\hat{w}}_{i} = μ + (w_{i} - μ) c_{i}

3. Shrink toward the average. Confidence scales the hero's distance from the mean rather than their win rate, so a low-pick hero is pulled back toward average and has to be far from it before the list will call them strong. Note that

p_{i}

is used only here — pick rate never adds to the score, because a list that rewards popularity is just a popularity chart.

δ_{i} = {\hat{w}}_{i} - μ = (w_{i} - μ) c_{i}

4. The delta. This is the +/− number printed under each hero's win rate on the board. Substituting step 3 collapses the whole pipeline to a single product: how far from average, times how much we trust it.

$T (δ_{i}) =$

S	$δ_{i} \geq + 4.0$
A	$+ 1.5 \leq δ_{i} < + 4.0$
B	$- 1.5 \leq δ_{i} < + 1.5$
C	$- 4.0 \leq δ_{i} < - 1.5$
D	$δ_{i} < - 4.0$

5. The cut. The boundaries are absolute distances in percentage points, not percentiles — the same point made above, and the reason tier sizes are an output rather than a quota.

Worked example, PC competitive: Storm wins 55.69% of matches, a full 5.34 points clear of the 50.35% average, which on raw win rate alone would be a comfortable S. But Storm appears in only 0.91% of matches:

δ_{Storm} = (55.69 - 50.35) \min (1, \frac{0.91}{3.0}) = 5.34 \times 0.303 = + 1.62 \Rightarrow A

A tier, not S, on a sample too thin to support the claim — and the card says so, by showing that win rate in gold.

How is this Marvel Rivals tier list calculated?: From the official Hero Hot List win and pick rates — see the method above. Heroes are ranked by win rate against the roster average, with rarely-played heroes pulled toward that average.
Can I make my own Marvel Rivals tier list?: Yes — just drag a hero into a different tier. The list becomes yours as soon as you move anything; picking a platform or rank button rebuilds the calculated list, while the role buttons only filter what’s shown and leave your list intact. On touch, tap a hero then tap a tier. Your list saves in your browser; “Copy link” shares it and “Download image” saves it as a PNG.
Does the tier list change by rank?: Yes — the official stats are published per bracket, so you can see Bronze through Celestial+ separately, on PC or console. Hero strength differs a lot between low and high rank.
How often is the tier list updated?: It follows the official Hero Hot List, which the developers refresh each half-season rather than daily. The current figures are as of 7 July 2026, always shown at the top of the tier list.