Beast of Reincarnation Review Scores: Why It Is Mixed

Three scores, read the same day

They do not agree, and the gap is not an error. Metacritic and OpenCritic score overlapping but different pools of outlets and weight them differently; the Steam figure is not critics at all. Reading all three is more informative than picking one.

The critic spread is narrow, and it has no bottom

This is the part a single number hides. Metacritic’s own breakdown records no negative reviews at all — on PlayStation 5, 31 positive and 26 mixed; on PC, 13 positive and 17 mixed. A 73 built that way is not a divisive game. It is a game almost nobody disliked and slightly under half of the press was enthusiastic about, which is a different thing from a 73 assembled out of nines and fours.

• PS5: 31 positive, 26 mixed, 0 negative.
• PC: 13 positive, 17 mixed, 0 negative.
• The high end is real — Vice scored it 95 and GamesRadar+ 90 — but the mass sits in the mid-80s and below.
• Nearly every article about this launch calls the reception divisive. The critic pool is the opposite of divisive; the division is in the section below, among the people who bought it.

Players are harder on it than critics, and that did not soften

Steam’s own summary is "Mixed", and this page has now read it twice. On 5 August it was 1,267 positive against 978 negative from 2,245 reviews. On 7 August it was 1,883 against 1,418 from 3,301 — the pool grew by 47% and the share positive moved from 56.4% to 57.0%. The earlier version of this section argued that a two-day-old pool over-represents launch-week technical complaints and would therefore move as it filled. It filled, and it did not move. That is one extra reading rather than a trend, and it is the first evidence either way; it points against the optimistic reading rather than for it. Both figures are the all-language totals from Valve’s public endpoint, queried the same way both times — the English-only pool sits at a different ratio, and swapping between the two would manufacture a movement out of a change in method.

The "Mixed" is a language split, and almost nobody is at the headline number

Read on 19 August 2026, Steam’s all-language score sat at 57.9% positive — and not one of the game’s eleven largest review audiences sat near it. The spread ran from Italian in the low eighties down to Korean and Simplified Chinese in the high thirties, more than forty points apart, with English and Japanese together in the middle. What makes that more than a curiosity is the size of the audience at the bottom: Simplified Chinese was the second-largest review language on the game, ahead of Japanese, German, French and Russian combined. So the band Steam prints is a weighted average across audiences that disagree with each other violently, and the weight pulling it into "Mixed" is Chinese and Korean. The chart below re-reads all eleven every day rather than quoting this paragraph back at you, so if the gap closes after a patch it will close here first.

• This is the same endpoint as the score above, filtered by `language`. Valve publishes no breakdown of its own, so the split does not exist anywhere until something asks once per language.
• It is NOT comparable with the all-language readings in the previous section, and this page never mixes them: swapping an all-language figure for a filtered one manufactures a movement out of a change in method.
• What it cannot tell you is why. A review count knows the language a review was written in and nothing about the person — localisation quality, regional pricing, platform mix and genre expectation are all live candidates and none of them is measured here.

The developer agreed with the criticism, in public, within two days

This is the part that makes a launch score a moving object rather than a verdict. On 5 August GAME FREAK posted a statement naming five things it intends to change in a first patch, and the third of them is "story pacing improvements, with further adjustment planned after this patch". Story pacing is precisely the criticism the mixed reviews shared — the combat is the thing almost every positive review singles out, and the pacing is the thing almost every mixed one does. A studio conceding the specific complaint behind its own mixed column, four days after release, is unusual enough to be worth stating plainly. What it is not is a fix: nothing had shipped when this page was last read, no date was attached beyond "within a week", and the developer says the contents are subject to change. Which of the five items actually arrives, and what was left off the list entirely, is tracked item by item on this site rather than summarised here.

The video reviews the aggregates do not contain

9 video reviews, ordered by how many people watched them, captured 2026-08-07. This table is deliberately only what a machine can read — channel, title, runtime, views, upload date. We have not watched or transcribed any of them, so nothing here says what a reviewer argued; where a verdict is visible it is because the creator put it in their own title, and the titles are reproduced verbatim. 4 of the 9 do that, which means 5 of these are legible as "a review exists" and nothing more. The counts in this paragraph are computed from the table rather than typed, because the table grows.

The review most people are now watching arrived two days late

This table was captured twice — 2026-08-06 and 2026-08-07 — and the second reading is the more interesting one. Between the two, every launch-day review moved by one to three per cent, which is what a four-day-old video does. One row did not. Skill Up's, uploaded 2026-08-06 rather than on launch day, went from 32,985 views to 131,327 in a single day — very nearly quadrupling, overtaking IGN's, and taking the top of this table from behind. Its title is "I do not recommend: Beast of Reincarnation (Review)". What this does not establish, and what no amount of view counting could: that the game changed, or that the critics who published on launch day were wrong. Two readings of a public counter cannot reach either. What they do establish is that the review the largest number of people have now seen is one whose title is a refusal to recommend, and that it got there after every aggregate score on this page had been read. If you are deciding whether to buy, that is a fact about the conversation you are walking into.

Why they are worth a table of their own

Between them these 9 have been watched 640,079 times, against the 64 written reviews OpenCritic averages into a 76. That is not an argument that the aggregate is wrong — it is the observation that "did the press like it" and "what did the audience for this game actually hear" are different questions, and only the first one has a number. Two entries are worth naming: Fextralife, the incumbent wiki in this niche, published its own review and titled it "Polished but Generic"; and Skill Up's, uploaded 2026-08-06, is the only one from outside launch week — it is titled "I do not recommend: Beast of Reincarnation (Review)", it is the fastest-moving number on this page, and it landed after every aggregate figure above was read.

What the number is not telling you

An aggregate answers "did the press like it" and nothing else. It does not tell you whether the combat clicks for you, which is the thing this game actually stands or falls on, and it does not tell you whether it runs well on your hardware — a separate question with a separate answer on this site.

How to read these two aggregates against each other

A three-point gap between Metacritic and OpenCritic is normal and mostly mechanical. OpenCritic publishes how many critics recommend the game as a distinct figure — 57% here — which is often more useful than either average, because it converts a pile of decimals back into a decision. Metacritic splits by platform, which is what surfaces the PC build scoring a point lower on a pool half the size.