Issue 01 · Published 2026
Most reviews in our study recommend the game. Yet a negative review is 3.6 times as likely as a positive one to reach five helpful votes, and hundreds of hours played do not always end in a recommendation. We examined 202,000 Steam reviews across 3,111 games to see where criticism is most common, which reviews get a response, and what an overall score can hide.
Method
An honest warning before you read: we compare reviews from different games at different ages, release stages and playtimes — not the same players over time. The numbers describe how players write, not what caused it. This is a sample built for fair comparison between groups of games, not a portrait of all of Steam.
The strongest findings
Featured finding · Refund window CritField · Steam Reviews 202K
Plotted across a game's playtime, negativity draws a U — and its highest point sits inside the two hours where Steam still allows a refund. Reviews written there are negative twice as often as reviews written after.
Featured evidence
Negativity draws a U: the newest and the oldest players are the unhappiest
% of reviews that are negative, by playtime band
Boundary · refund window ends at 2h
Six playtime bands · categorical intervals · domain 0–40%
Trust runs highest in the middle: newcomers and long-time veterans are both harder to satisfy than everyone in between.
LimitPlaytime is measured at the moment the review was published, not total hours ever played.
In the full studyChapter 02 · First hours
Using this chart?Press
Featured finding · Review language CritField · Steam Reviews 202K
This chart shows where negative game reviews are most common by review language. Traditional Chinese is highest among the 14 groups shown at 31.9%. Russian is 22.7%, English 21.1%, and Latin American Spanish is lowest at 15.7%. The full ranking stays visible in the chart below.
Featured evidence
Share of negative reviews by review language
% of reviews that are negative, by the language the review is written in
1422% ref34%
Traditional Chinese has the highest negative share among the 14 language groups shown. Russian and English sit closer to the study average, while Latin American Spanish is much lower. The first job of this chart is simply to show where negative reviews are more common. After that, a studio can compare the same games across languages and read the complaints themselves.
ScopeReview language is recorded for each review, not the reviewer's nationality or culture — the games represented differ between language groups. The chart shows 14 of the 29 recorded review languages; 22,065 reviews with no language recorded are excluded from this comparison.
In the full studyChapter 05 · Language & scale
Enter the full study · 7 chapters
Explore the findings
Right after launch, players want to believe. The real cross-examination starts once the launch window closes — and only quiets down over years.
Figure 1.1
Negative reviews peak in the 29–180-day group
% of reviews that are negative, by stage of a game's life
Negativity keeps climbing for months after the launch-week calm, then eases only gradually.
"Strict recount" runs the same numbers on a more conservative slice of the sample — the picture holds: 17.7% → 28.2%.
Source · CritField, Steam Reviews 202K · N = 202,000 reviews, 3,111 games · correlational, corpus-specific
Figure 1.2
At launch, one review in three comes from the first two hours
% of reviews written inside the first two hours of play
The audience changes along with the mood — that's the key to reading any comparison between stages.
Source · CritField, Steam Reviews 202K · N = 202,000 reviews, 3,111 games · correlational, corpus-specific
What the numbers say
The important handover comes after launch month. That is when the study's negative share is highest, so review monitoring should continue while the team moves into fixes, updates and ongoing support. Make the first six months part of the release plan from the start.
What to do
Give someone clear responsibility for reviewing new complaints through month six. Keep a record of recurring problems and publish clear updates when they are fixed. For strategy and VR, continue the same work beyond six months: their highest negative shares appear later in the study.
What we can't claim: we compared reviews written in different windows of different games' lives — not the same people over time. So "players grow disillusioned by month two" does not follow from this data; what follows is only that reviews published in that window are markedly more negative.
Each release stage contributes 50,500 reviews. The proportion written within two hours of play becomes smaller in the older game groups, and reviews tend to become shorter. Those changes matter when comparing the conversations around new and older games. Read in full-length reviews only, the same release-stage arc runs from 19.1% up to 30.4%.
The overall peak contains several different stories. MMO and live-service reviews are most negative in months two through six (26.4% → 42.4%), and battle royale follows a similar early climb (29.9% → 39.8% by the mature phase). Narrative indies also peak there, with a particularly sharp rise from their first-month group: 10.0% → 24.9%. Strategy and VR peak later, in the mature phase (strategy reaches 29.6%). Survival and horror stay close to 15–17% at all four stages. Action RPGs and simulations mellow steadily with age, down to 13.9% and 10.6% in their legacy years. Use the genre map to find the release period relevant to the game you are discussing.
Figure 1.3
Critics typically have fewer hours played than reviewers who recommend the game
median hours played when the review was written
Negative reviewers have a lower median playtime at every release stage. The memorable thousand-hour complaint is only one kind of criticism. A large part of the feedback also comes from people who have spent much less time with the game.
Source · CritField, Steam Reviews 202K · N = 202,000 reviews, 3,111 games · correlational, corpus-specific
How this was checked: the criticism arc survived a recount without the five biggest games, a recount on full-length reviews only, and the strict recount — the peak stays in months two through six in all of them.
A player can leave a negative review after 30 minutes or after 300 hours. Those reviews point to different questions. The first can be about getting into the game at all; the second comes from someone who stayed for a long time and still ended up saying they would not recommend it. The study finds high negative shares at both ends of the playtime range.
Figure 2.1
More than half of early MMO and live-service reviews are negative
% of refund-window reviews that are negative, by type of game
The first-hours problem is especially pronounced in MMOs and live-service games: 57.8% of reviews written within two hours are negative. For narrative indies, the figure is 20.1%. The opening experience is a far sharper pressure point in some categories than others.
Source · CritField, Steam Reviews 202K · N = 202,000 reviews, 3,111 games · correlational, corpus-specific
What the numbers say
All 15 game categories show a kind of playtime “honeymoon”: reviews written in the first two hours are more negative than reviews in the middle playtime groups. Across the full 202,000-review sample, the low point comes at 10–50 hours, with 16.7% negative reviews, versus 37.6% in the first two hours. After that, criticism rises again overall — to 29.0% at 200–1,000 hours and 29.6% at 1,000+ hours. The late rise differs by game type, so veteran reviews should still be read separately.
What to do
Review the opening experience separately: installation, first launch, settings, tutorial and the first meaningful stretch of play. Use recurring complaints to decide what to test. Keep a second reading list for reviews posted after 200 hours so feedback from experienced players remains easy to find.
What we can't claim: the curve describes who writes a review and in what mood — not "stages every player goes through." And it names no causes: the data shows the language of players, not the mechanism of their discontent.
Older simulation games look comfortable at first glance: just 10.6% of their reviews are negative. Look only at reviews written within two hours, however, and the figure is 31.0%. Among reviewers with 50–200 hours, it is 4.9%. An established audience's approval and a newcomer's first impression can be very different things. A healthy overall score is a reason to keep reading, not a reason to stop checking the opening experience.
Older VR games show the same pattern more sharply: 13.1% negative overall, but 44.4% among reviews written within two hours, versus 7.9% at 10–50 hours.
In all 15 game categories, reviews written within two hours are more negative than reviews from the middle playtime groups. That cross-category result is the evidence behind the “playtime honeymoon” shorthand. Its severity varies widely: in MMOs and live-service games, every second refund-window review is negative (57.8%); in MOBAs it's 45.8%, battle royales 43.6%, VR 43.3%. Narrative indies (20.1%) and puzzle games (21.4%) pass the same filter almost unscathed. The late-playtime story is less uniform: simulations, action RPGs and Early Access have relatively positive veteran groups, while several service-heavy categories become more negative again — veteran negativity rises in 11 of the 14 groups with enough data.
Supporting comparison
First impressions stay a problem long after launch
% of reviews that are negative, MMO and live-service games aged six months to two years
The first two hours remain a pressure point well beyond release week. For MMO and live-service games between six months and two years old, 63.2% of reviews written within two hours are negative. Among reviews written after 10–50 hours, the figure is 28.2%. For a new arrival, the opening experience can look very different from the game an established player describes.
Source · CritField, Steam Reviews 202K · 793 of 1,254 reviews negative in the first-two-hours group; 492 of 1,742 at 10–50 hours · correlational, corpus-specific
A game's storefront is shaped not by the average player but by the loudest one. And the loudest players are the unhappy ones.
What this means
Negative reviews are only 22% of the full sample, but they make up roughly half of the reviews that reach at least five helpful votes. That is the easiest way to see what the 3.6× result changes in practice: the helpful-vote group is far more critical than the full recommendation count.
Figure 3.1
Official replies are rare, even under negative reviews
Share of reviews carrying an official developer reply · one mark = 1% of reviews
No reply Received a reply Each dot represents 1% of reviews.
Developers reply to 4.9% of negative reviews and 1.2% of positive reviews. Criticism gets a response more often, but most negative reviews still have no official answer attached. Across the whole corpus, only 2% of reviews receive any official reply at all.
This measures who receives a reply, not what a reply changes: nothing here shows whether a developer response moves anyone's opinion.
Source · CritField, Steam Reviews 202K · N = 202,000 reviews, 3,111 games · correlational, corpus-specific
What the numbers say
Steam surfaces the reviews the community marked helpful, and criticism clears that bar far more often than praise. Five or more "helpful" votes go to 19.6% of negative reviews and only 5.4% of positive ones. The first page of reviews is the shop window, and negative reviews reach that page's most-visible slot at more than three times the rate of positive ones. Meanwhile the response field is nearly empty: only 2% of reviews get an official developer reply, and it goes to criticism 4.3 times more often than to praise.
What to do
Start with negative reviews that describe a specific problem and already have helpful votes. Give a clear answer where the team has one. When a fix ships, add the update beneath the relevant review so a future reader can see the complaint and the resolution together.
What we can't claim: how many votes a review collects also depends on Steam's visibility algorithms — we see the combined outcome of "text plus storefront," not the pure effect of the text. And this data doesn't measure whether a developer's reply changes anyone's mind: it only shows who gets the attention.
Critical reviews tend to offer more text as well as attracting more helpful votes. Their median length is 183 characters, compared with 106 for positive reviews. That helps explain why criticism can occupy so much of a reader's attention even when recommendations are positive overall.
Reply habits differ between categories. Free-to-play reviews receive official responses most often at 5.1%, followed by VR at 3.9%; puzzle games are lowest at 0.7%. Use these as context for community practice, with the overall reply rates kept separate from negative-only rates.
Negative reviews receive their strongest helpful-vote response during the first half-year. About one in four reaches five helpful votes, and the average is 5.6–5.8 votes, compared with 1.9 for games older than two years. This gives the post-launch support recommendation a second reason: criticism is common and other players are responding to it.
A 200-hour review tells you that someone has spent time with the game. It does not tell you whether they recommend it. In MMOs and live-service games, 38.0% of those reviews are negative; in simulations, 10.0% are. The kind of game changes how the same playtime should be read.
Figure 4.1
Experienced MMO players are far more critical than experienced simulation players
% of 200+ hour reviews that are negative, by type of game · 14 groups against the 29.1% veteran reference
200+ hours played · 14 groups
Higher share negative
Lower share negative
A single benchmark would flatten an important difference. A 30% negative share among experienced reviewers sits below the MMO and shooter rates in this study, but well above the simulation rate. Compare a game's veteran reviews with similar games before deciding how unusual its reception looks.
Groups shown have at least 100 veteran reviews. Position on the axis is the measured share and nothing else — these are separate groups of reviews observed past 200 hours, not a path any player travelled.
Source · CritField, Steam Reviews 202K · N = 202,000 reviews, 3,111 games · correlational, corpus-specific
Using this chart?Press
Hours played are one part of the picture. The next comparison asks when in a game's life the toughest reviews appear.
Figure 4.2 · The negativity map
Different games face their toughest reviews at different ages
% of reviews that are negative · 15 types of game read across four stages of a game's age · separate groups of reviews at each stage, not games followed through time
| Type of game | 0–28 days | 29–180 days | 181–730 days | 2+ years |
|---|---|---|---|---|
| MMO & live-service | 26.4% 2,381 | 42.4% 8,619 | 37% 6,876 | 28.7% 4,612 |
| Action RPG | 17.6% 9,586 | 19.2% 2,240 | 16.5% 3,082 | 13.9% 5,208 |
| Early Access (cross-genre) | 17.6% 4,684 | 16.5% 2,216 | 20.3% 7,280 | 23.7% 3,852 |
| Narrative indie | 10% 7,222 | 24.9% 975 | 11.7% 2,180 | 12.4% 4,428 |
| Shooters | 18.5% 3,095 | 17.1% 4,577 | 21.5% 3,767 | 18% 3,218 |
| Sports & racing | 18.2% 2,111 | 18.6% 4,944 | 24.5% 5,142 | 17.4% 2,661 |
| Simulation | 17.7% 3,046 | 20.1% 1,917 | 14.9% 3,103 | 10.6% 4,588 |
| Strategy | 18.1% 1,672 | 23.8% 3,421 | 29.6% 4,830 | 16.7% 4,636 |
| Survival & horror | 16.8% 3,074 | 16.5% 1,609 | 15.2% 3,582 | 16.7% 2,544 |
| Free-to-play | 22.7% 2,785 | 67.7% 362 | 34.6% 677 | 23.8% 4,688 |
| VR games | 21.4% 594 | 26.2% 2,232 | 31.5% 1,932 | 13.1% 3,251 |
| Battle royale | 29.9% 1,296 | 38.9% 2,094 | 39.8% 2,259 | 25.1% 1,738 |
| Open world | 19.7% 6,038 | 21.2% 1,209 | no data | 47.3% · 91 |
| Puzzle & casual | 11.3% 2,239 | 5.9% 699 | 12.9% 1,400 | 13.9% 3,568 |
| MOBA | 26% 677 | 34.2% 1,530 | 36% 1,883 | 31.1% 1,417 |
Narrative indies have a sharp spike in months two through six. Strategy and VR peak later, between six months and two years. Early Access has its highest negative share after two years, while survival and horror remain relatively steady. The useful question is when to watch most closely for your kind of game. Each cell shows the negative share and number of reviews.
A hatched cell has no data for that group at that stage. Cells with few reviews carry a dot — read them with care. The fiery free-to-play cell in the early months is one game's story, not a genre trend: a single troubled launch (Stormgate) accounts for more than 99% of the reviews in that cell.
Source · CritField, Steam Reviews 202K · N = 202,000 reviews, 3,111 games · correlational, corpus-specific
What the numbers say
Veteran criticism is worth reading in the context of the game. Hundreds of hours commonly coexist with a negative recommendation in service games, while simulation and action RPG veteran reviews are much more positive. The playtime badge is the beginning of the interpretation, not the end of it.
What to do
Keep a separate view of reviews written after 200 hours. Compare their negative share with similar games and read the recurring complaints. A result that looks ordinary in one category can deserve closer attention in another.
What we can't claim: genre groups are groups of games with different audiences and different business models; the data shows where each tone is more common, not whether "the genre is to blame." A few cells of the map are empty or thin — we don't hide that.
The MMO-versus-simulation comparison is spread across multiple titles. The largest game contributes 6% of the MMO veteran group and 19.5% of the simulation veteran group. The broad difference remains visible beyond a single game's reviews. It is also not simply "angry genres": strategy veterans are calmer than their mid-cycle players (24.5%), while sports and racing veterans are angrier (32.8%).
These groups show why "they played for hundreds of hours" is not a complete answer to a negative recommendation.
A review score becomes more useful when you know what you are comparing. The language groups above show different balances of praise and criticism. Games with very different numbers of reviews do too. For a developer judging reception, both comparisons can help decide where to look more closely.
Figure 5.1
Games with more reviews have a lower negative share
% of reviews that are negative, by lifetime review volume
Negative reviews are more than twice as common in the group of games with fewer than 1,000 lifetime reviews (25.4%) as in the group with over a million (11.9%). A smaller game's reception is more informative when compared with closer peers than with the most reviewed titles in the store.
Review-volume tier describes how many reviews a game has accumulated — not its budget and not “AAA status.” 26,722 reviews from games of unknown tier are excluded from this cut.
Source · CritField, Steam Reviews 202K · N = 202,000 reviews, 3,111 games · correlational, corpus-specific
What the numbers say
The choice of comparison changes the question. Is this game receiving unusually critical reviews for a title like it, or are we comparing a small review page with an established giant? Start with similar review counts, then look at genre and release stage.
What we can't claim: a language segment is not a country and not an audience temperament — the mix of games differs between segments, so the spread describes the reviews, not the people who wrote them. Review volume is a measure of how much a game has been reviewed, not of its budget or its quality.
Review count shows how much feedback a game has accumulated. The five groups range from fewer than 1,000 lifetime reviews to more than a million. A game's budget is a different measure. Part of the scale effect likely reflects survivorship: a game with a million reviews has already passed a filter that small games are still standing in.
Detailed criticism is easy to notice because it gives you something to read. Brief approval can be easier to pass over. In this study, selecting only longer reviews raises the negative share from 22.0% to 25.8%. The way you choose reviews changes the picture you get.
Figure 6.1
The longer the review, the likelier it is negative
% of reviews that are negative, by length of text
Longer reviews give you more to analyze, but they also contain a higher share of criticism. A summary built only from those texts will sound more negative than the complete recommendation count. Use every rating to measure the balance of approval, then use detailed comments to understand what people are responding to.
Figure 6.2
Most reviews contain more than a short reaction
% of the sample, by length of review
There is plenty of text to work with: 62.8% of reviews meet the study's longer-text rule. But the shorter recommendations still belong in the overall count. Removing them changes whose opinions the analysis represents.
Source · CritField, Steam Reviews 202K · N = 202,000 reviews, 3,111 games · correlational, corpus-specific
What to do
Keep two outputs side by side: the positive-versus-negative split across all reviews, and a summary of issues described in the longer comments. Label the second clearly so it is not mistaken for the opinion of the whole sample.
What we can't claim: length doesn't make an opinion more correct — it only shows who was motivated to write. The effort–negativity link is correlational: we don't know whether people sit down to write because they're angry, or get angrier as they write.
Nearly three quarters of negative reviews meet the longer-text rule (73.7%), compared with about three fifths of positive reviews (59.9%). That difference gives critical feedback more room to explain itself. Short recommendations still matter even when they offer little detail about why the reviewer liked the game — the ultra-short form is mostly praise: "+" and "gg" come from the satisfied (9.8% of positive reviews) and almost never from the dissatisfied (4.3%).
Before summarizing the most helpful reviews, check how their mix of praise and criticism differs from the full review set.
Reviews get shorter in the older game groups. The typical text falls from 171 characters in the first month to 68 after two years, and the share meeting the longer-text rule falls from 76% to 47%. An older game's review page can contain plenty of recommendations with relatively little explanation attached.
Use the review groups to decide where to look, then read the comments to decide what to investigate. These six actions connect the study to QA, player experience, support and communication.
Collect recurring complaints from reviews written within two hours. Turn them into checks for installation, first launch, settings, the tutorial and the first meaningful stretch of play. Give the opening experience its own review instead of assuming a complete-game test will cover it.
An established game still receives first-time players. Compare their reviews with feedback from people who have played longer. If the opening group is much more critical, read those complaints together and investigate where the experience is breaking down.
Assign time and ownership for reviewing feedback after launch month. Track recurring problems and make fixes visible to players. Keep watching beyond six months for categories such as strategy and VR, whose most negative groups appear later.
Find negative reviews that describe a concrete problem and have helpful votes. Answer clearly when you have useful information. If the problem is fixed later, return to the review with an update that future readers can understand.
Keep reviews written after 200 hours easy to find. Read their recurring concerns and compare the group's reception with similar games. Large playtime totals do not tell you whether those reviewers recommend the game.
Check what the trailer and store page promise against what a player actually reaches in the first two hours. Use early reviews to find recurring mismatches worth investigating. Keep the full recommendation count beside any summary built from longer comments.
Choose a release stage, game category and playtime group to see the number of reviews and the share that are negative. Try comparing early impressions with reviews written after hundreds of hours.
You are welcome to quote findings and share charts in articles, videos, newsletters and presentations. Credit CritField and link back to this study. For republication of substantial material — the analysis, the tables, the visual design or the written text — or other permission requests, contact [email protected].
"Source: CritField, Steam Reviews 202K." For a specific number, add the chapter — for example: "CritField, Steam Reviews 202K study, chapter 01: Lifecycle."
Questions about the data and method can go to [email protected]. The methodology chapter above answers most questions about how the numbers were made, and the press room has a fact sheet and downloadable charts.
© 2026 CritField. All findings are correlational and come with the caveats stated in each chapter — if you quote a number, please carry its caveat with it. Steam is a trademark of Valve Corporation; this is an independent research project, not affiliated with or endorsed by Valve.