Perthro

Industry··9 min read

How Metacritic and OpenCritic scores work

Two numbers out of 100 that look alike but are built differently, and what each one hides.

Metacritic and OpenCritic scores look like the same thing: one number out of 100 that sums up what critics thought of a game. They aren't calculated the same way. Metacritic's Metascore is a weighted average, where some publications count more than others and the weights are secret. OpenCritic's Top Critic Average is a plain average of its "top critics," shown next to a separate percentage of critics who recommend the game.

Both numbers are useful. Both also hide things. Here's how Metacritic and OpenCritic scores work, using games from this year, and how to read past the headline number.

How Metacritic calculates the Metascore

Metacritic's own help pages describe the Metascore as "a weighted average of reviews from top critics and publications." Staff collect reviews from an approved list of outlets, convert each score to a 0-100 scale, and average them. Publications that Metacritic considers higher quality or more prestigious get more weight. Asked whether it will explain how each critic is weighted, Metacritic's FAQ answers, in full: "Absolutely not."

A few rules matter for games specifically:

OpenCritic splits its numbers in two. The Top Critic Average is, in its FAQ's words, "the simple average score of all numeric reviews written by top critics after normalizing the score on a 0-100 scale." No secret weights. A game needs at least three numeric reviews from top critics before it gets one.

The second number, Critics Recommend, covers a wider pool. It includes top critics and regular critics. For a numeric review from a top critic, the cutoff for "recommended" is that publication's own median score, unless the outlet sets its own threshold. Other critics tick a recommend box when they submit a review, and some unscored outlets with clear verdicts, like Eurogamer, are counted too.

That second number is often the more honest one. Hot Wheels Infinite Rush has a respectable 74 Top Critic Average, but only 57% of critics recommend it. A 74 tells you it's fine. The 57% tells you nearly half the critics wouldn't point you toward it.

A person playing a video game on a laptop
A person playing a video game on a laptop

Metacritic vs OpenCritic: the differences side by side

MetacriticOpenCritic
Headline numberMetascore: weighted averageTop Critic Average: simple average of top critics
Weights published?NoNo weights; every top critic counts equally
Minimum reviews4 (shows "tbd" until then)3 numeric top-critic reviews
Second numberUser score (0-10)% of critics who recommend
Score bands for gamesGreen 75-100, yellow 50-74, red 0-49Mighty 84+, Strong 75-83, Fair 66-74, Weak 65 and below
Top badgeMust-Play: 90+ from at least 15 reviewsMighty: top 10% of games
Outlet changes its scoreOriginal score staysUpdated review replaces the old one

How 5-star and letter-grade reviews become a score out of 100

Every outlet grades differently, so both sites convert everything to 0-100. A score out of 10 is multiplied by ten. A score out of five is multiplied by twenty, which is why Eurogamer's 3/5 for Marvel's Wolverine shows up on Metacritic as a 60.

Metacritic publishes its conversion tables. On a four-star scale, 3.5 stars becomes 88 and 3 stars becomes 75. Letter grades run from 100 for an A down to 0 for an F: an A- is 91, a B+ is 83, a B is 75 and a C is 50.

The conversion is where a lot of meaning gets lost. A 3/5 from an outlet that uses the whole scale is a respectable, recommendable score. Turned into a 60, it lands in the lower half of Metacritic's yellow band and reads like a failing grade.

What the Metacritic colours mean, and why scores cluster between 70 and 90

For games, Metacritic's bands are stricter than for movies. A game needs 75 to be green ("generally favorable") and 90 to hit "universal acclaim." Anything from 50 to 74 is yellow, and under 50 is red. Metacritic explains why: game publications almost all treat scores below 50 as negative, and "it usually takes a score in the upper 70s or higher" to mean a game is unequivocally good.

That's why most game scores bunch up in the 70s and 80s. Outlets use the top half of their scale for almost everything they review. Of the 115 scored Crimson Desert reviews on its Metacritic PC page, 83 fall between 70 and 90. Saros has 139 scored PS5 reviews, and 52 of them are exactly 90.

It also means small gaps matter. Hot Wheels Infinite Rush sits at 73 on Metacritic. Two points higher and it would be green. OpenCritic's tiers split things the same way: a 74 is "Fair," a 75 is "Strong."

Embargoes, early scores and user reviews

Most big games get their reviews on a single embargo day, often before launch. Marvel's Wolverine reviews went live on September 10, five days before it releases on September 15. On embargo day the number moves fast. Embargo-day roundups caught Wolverine at 78 from 112 reviews, then 77 from 121, while OpenCritic had it at 79 with 75% recommending.

Metacritic says reviews for games can arrive anywhere from a month before release to three to six months after. Every late review can nudge the average, as Wolverine's first day showed. Check the review count before you trust the score: a Metascore from 20 reviews can still shift, one from 120 mostly won't.

User scores are separate. Metacritic's user score is out of 10, it never feeds into the Metascore, and players can only post one 36 hours after a game's release. Metacritic added that wait in July 2020, weeks after The Last of Us Part II was hit by a wave of low user scores in the days after launch, a practice known as review bombing. So on September 14, Wolverine has a Metascore but no user score at all.

A player holding a controller in front of a monitor, the moment a Metacritic score stops mattering
A player holding a controller in front of a monitor, the moment a Metacritic score stops mattering

How to read a split review score

An average can't tell you whether everyone agreed. A game made of 100s and 40s and a game made of straight 80s can land on the same number, and they're very different bets.

Crimson Desert is the clearest example this year. It sits at 77 on PC from 119 critics, a green score. But its reviews run from 45 to 100: five perfect scores, five 50s and a 45. Our Crimson Desert review deep dive covers why critics fought over it. Compare Saros, where almost everyone landed between 80 and 100. Even a score in the mid-80s can hide a real argument, as our Onimusha: Way of the Sword deep dive found.

Metacritic makes the spread easy to see. Every game page lists how many reviews were positive, mixed and negative. Crimson Desert's PC page shows 84 positive, 30 mixed and one negative. Saros shows 131 positive, 8 mixed and none negative.

What aggregate scores can't tell you

An aggregate score tells you how critics as a group reacted. It can't tell you whether you'll like the game. The most useful thing you can do with a split score is find one or two critics whose taste matches yours and read them.

Your own ratings help here too. If you rate games as you finish them, in Perthro or anywhere else, you'll soon notice where you part ways with the Metascore, and which kinds of 75s you tend to love.

Sources

Keep reading.

Every post →
Industry7 min read

Why so many big games release in the fall

Industry7 min read

Why video game music belongs in the concert hall

Industry6 min read

How Steam Deck is changing PC game design