Dawnwalker: Russian Steam reviews gained ground, Japanese did not

The Blood of Dawnwalker News 10 min read

Short answer: the Russian score climbed from 57.8% to 73.8% as the review count more than tripled, cutting the gap to English from 20 points to 12. Japanese moved the other way relative to the field — still last, still the language nobody expected to be there. The numbers below are from Steam’s own review-summary endpoint, queried twice on the same account: 2026-09-04 (day two) and 2026-09-13 (day eleven), for The Blood of Dawnwalker (appid 3751260).

On 31 August, before any review existed, we flagged that the game shipped without a Russian localisation and predicted it would show up as a depressed score in the review data. Two days after launch it did — Russian sat at 57.8%, the second-lowest language on the board. Eleven days in, the picture has moved enough that the original snapshot is now the less interesting number.

What changed between day two and day eleven

The whole review pool grew from 5,292 to 19,922 — nearly four times larger — and the overall rating climbed from “Mostly Positive” (78%) to “Very Positive” (85.7%). Individual languages did not move uniformly with it.

LanguageDay 2 (Sep 4)Day 11 (Sep 13)Reviews (day 11)Localised?
Latam Spanish92.1%93.5%107text
Czech90.9%93.4%227text
Spanish (Spain)87.4%93.0%543full audio
Turkish86.2%92.6%391text
Brazilian Portuguese84.2%91.8%476text
Italian91.9%89.3%196full audio
German87.6%88.6%1,177full audio
Hungarian73.7%87.8%74text
English78.2%86.4%11,903full audio
French83.9%86.3%717full audio
Polish84.0%84.4%1,251full audio
Traditional Chinese88.5%80.1%156text
Simplified Chinese69.3%79.4%1,249text
Korean70.0%76.7%163text
Russian57.8%73.8%852none
Japanese33.3%45.0%80text
Forty-eight points still separate the top and bottom language The Blood of Dawnwalker, positive review rate by Steam review language, 19,922 reviews, read 2026-09-13 Latam Spanish 93.5% (107) Czech 93.4% (227) Spanish 93.0% (543) Turkish 92.6% (391) Brazilian PT 91.8% (476) Italian 89.3% (196) German 88.6% (1,177) Hungarian 87.8% (74) English 86.4% (11,903) French 86.3% (717) Polish 84.4% (1,251) Trad. Chinese 80.1% (156) Simp. Chinese 79.4% (1,249) Korean 76.7% (163) Russian (no loc.) 73.8% (852) Japanese 45.0% (80) Russian gained 601 reviews and 16 points between day two and day eleven. Japanese gained 41 reviews and 12 points and is still the lowest score in the table by 29 points.

Russian: the gap is closing, not gone

The prediction from launch week landed, but the eleven-day data is the more useful number now. A Russian reviewer base with no localisation at all went from 251 reviews at 57.8% to 852 reviews at 73.8% — more than triple the volume and a 16-point jump, while English gained only 8 points over the same window. The gap between them shrank from 20.4 points to 12.6.

That is not what a permanent penalty looks like. It reads more like a rough launch score correcting as more of the audience that stuck around long enough to form a settled opinion started leaving reviews, on top of whatever goodwill patches and community translation mods bought back in the first eleven days. We have not checked for a fan translation mod; if one exists, it would be the more direct explanation than sentiment alone.

Japanese: still the outlier, and now further behind

Japanese sits at 45.0% on 80 reviews — the lowest score in the table, 28.8 points under Russian and 41.4 under the top language — despite shipping with text localisation. It improved from 33.3% at day two, but every other language improved too, and Japanese improved the least in relative terms among the bottom four. A supported locale is still the worst-scoring one in the dataset.

We are not going to assign a cause with 80 reviews to work from. Text-quality complaints, regional hardware performance, and ordinary small-sample noise remain the live candidates, in that order of how often we have seen each explain a similar gap on other Steam pages we track.

What the eleven-day table says that the two-day table didn’t

English moved from eleventh of sixteen to ninth of sixteen — up from 78.2% to 86.4% — but it is still short of the top half of the pack by score, on by far the largest sample (11,903 reviews, five times the next-largest language). Volume and satisfaction are not the same axis here.

The full-audio languages (Italian, German, Spanish, French, Polish, English) now span 84.4% to 93.0%, a tighter and higher band than the 78.2%–91.9% spread they occupied at day two. Whatever launch-week friction depressed some of those scores has mostly washed out in a week.

Playtime, on day eleven

The refund-window effect flagged in this game’s first three days has faded rather than closed. A fresh pull of the 1,000 most recent reviews (all languages, 12–13 September, day ten and eleven) puts the under-two-hour band at just 25 reviews — 2.5% of the set, against 23% of the first-three- day sample — and it still reads as the weakest band by far, at 44.0% positive. Every band past two hours clears 79%, rising to 94.6% at 30–100 hours (n=538, now more than half the sample).

Hours at review (day 11)ReviewsPositive
Under 2 h2544.0%
2–10 h14079.3%
10–30 h29385.3%
30–100 h53894.6%

The purchase-route gap the first-three-day sample found has also narrowed to the point of reversing: key activations now read 87.5% (n=376) against 89.1% for store purchases (n=624), a 1.6-point gap the other way from the 5.3-point key advantage the launch-week set showed. Whichever early-access or reseller channel supplied those first keys is a shrinking share of the review pool eleven days in.

Whether Russian keeps closing on English

  • Whether Russian keeps closing the gap or plateaus here. Two readings is a trend line with two points on it. A third reading in a month would tell us which.
  • Why Japanese trails a locale with no localisation at all. The sample stayed small (80 reviews); we report the number and are not fitting a story to it.
  • Whether a community translation exists for Russian. We checked Steam’s own review data, not Steam Workshop or third-party mod trackers.
  • Review language versus player country. Steam records the language the review was written in, not where the reviewer lives.

Sources