Features
Every game on this week’s Steam 250 weekly chart that launched in the last seven days is above 76% positive, and eleven of them are at 95% or higher. That is not a sign that games suddenly got better. It is what happens when a review score is calculated from the people who chose the game.
Look at the top fifty for the week ending 12 September 2026 and the pattern is almost comic. Garden Ink: 100%. Olga – Episode 1: 100%. Frontier: Path of Shadows: 100%. Push Push: 98%. Dark Bone Prison, Sporting Goods Shop, King of Jade: Stone Tycoon, Monster Alchemy: 97%.
Meanwhile Wardogs, the one title on the chart with a mass-market marketing budget, sits at 76% from 36,409 reviews.
The argument: small review counts measure self-selection, not quality
A game with 36 reviews has been bought almost exclusively by people who went looking for it. They read the store page, watched the trailer, recognised the genre and decided in advance that this was for them. The purchase is already a filter, and by the time someone writes a review they are reporting on whether a game they pre-approved delivered what it promised.
A game with 36,000 reviews has been bought by people who saw a thumbnail, people whose friends were playing, people who wanted something else and settled. That audience contains everyone the first audience filtered out — and every one of them can leave a review.
So a 100% score at 36 reviews and a 76% score at 36,409 are not measuring the same thing at all. The first measures whether the game found its audience. The second measures what happens after it runs out of them.
How the numbers move as a game gets bigger
| Game | Reviews | Positive | Audience |
|---|---|---|---|
| Frontier: Path of Shadows | 41 | 100% | People who searched for XCOM-likes |
| King of Jade: Stone Tycoon | 60 | 97% | Tycoon sim regulars |
| The Loopler | 101 | 95% | Deckbuilder players |
| The Legend of Fancy Realm Remake | 447 | 88% | SRPG fans plus nostalgia buyers |
| Wardogs | 36,409 | 76% | Everyone a $39.99 FPS reaches |
Read top to bottom, the score falls as the review count rises. That is the shape of self-selection dissolving, and it is close to a law rather than a coincidence.
Three things this does not mean
It does not mean small games are secretly bad
Nothing here says a 97% score is fake. It says the score answers a narrower question: did this game satisfy the people most likely to want it? For a genre title you are already sure about, that is exactly the question you were asking.
It does not mean big games are underrated
Wardogs’ 8,478 negative reviews are real people with real complaints. Broad reach exposes genuine problems that a small audience never encounters — server load, matchmaking, hardware variety. A wide audience does not just dilute a score, it stress-tests the game.
It does not mean the scores are useless
They are useful once you read them as two numbers rather than one. A percentage without its review count is not information.
How to actually read a launch-week score
Under 100 reviews: read the reviews, not the number
At this sample size the percentage can swing ten points on two people. Open the negative reviews first — there will be few enough to read all of them, and they are usually specific about a bug or a missing feature.
100 to 1,000 reviews: the number starts to mean something
The score has enough sample to be stable and the audience is still mostly people who sought the game out. This is the range where a high score is genuinely encouraging.
Over 10,000 reviews: compare against the genre, not against 100%
At this scale 76% and 85% are meaningfully different, and 95% would be remarkable. Judge against other games with similar reach.
Always check the release date
Every game discussed here launched within seven days. Launch-week scores are the most self-selected scores a game will ever have, and they typically drift down as reach grows. A score six months old has survived contact with a general audience.
The honest caveat
This is a pattern drawn from one week of one chart. It is consistent with how Steam review scores behave generally, but the table above is eleven games in September 2026, not a study. Individual games break the pattern in both directions — a genuinely broken launch will crater at 300 reviews, and a genuinely exceptional game can hold above 95% into six figures. Use the reasoning, not the numbers.
Valve does not publish a weighting formula for the percentages shown on store pages, so everything here is inferred from the public review counts, not from documentation.
Frequently asked questions
Why are new Steam games always rated so highly?
Because launch-week buyers are people who deliberately sought the game out. They already knew the genre, the art style and the price before buying, so their reviews mostly report whether the game delivered what it advertised.
How many reviews does a Steam game need for the score to be reliable?
Around 100 is where the percentage stops swinging wildly on individual reviews. Below that, read the negative reviews directly instead of trusting the number.
Is a 100% positive Steam game better than a 90% one?
Not necessarily. A 100% score usually comes with a very small review count and a narrow, self-selected audience. A 90% score across thousands of reviews has survived a far broader range of players and hardware.
Do Steam review scores go down over time?
They commonly drift downward as a game reaches beyond its core audience, because later buyers are less pre-filtered. A score that stays high as the review count grows is a much stronger signal than a high launch score.
Why does Wardogs have a lower score than smaller games?
Wardogs has 36,409 reviews against 27,931 positive and 8,478 negative. At that scale the audience includes everyone a 39.99 dollars shooter reaches, not only the people who went looking for it.
What is the difference between review score and review count on Steam?
The score is the percentage of reviews that are positive. The count is how many reviews it is based on. A percentage without its count tells you very little, because the same percentage means different things at 30 reviews and 30,000.
Sources and citation
- Steam 250 weekly chart, week ending 12 September 2026 — positions, scores and review counts
https://steam250.com/7day - Steam review summary API, Wardogs — 27,931 positive and 8,478 negative of 36,409 total
https://store.steampowered.com/appreviews/1867240?json=1 - Steam store pages for each game cited, for review counts and release dates
https://store.steampowered.com/
Recommended
- Every Recent Steam Game Rated Over 90% (And How to Read the Score)
- Is That Game Actually Dead? The Palworld Dev Says You Are Reading Steam Numbers Wrong
- Wardogs Player Count: 175,000 Playing Right Now — But the Number That Matters Comes in Six Weeks
- What Is SteamDB? The Unofficial Steam Database — And Who Owns It Now
- How to Actually Find Good Games on Steam (The Front Page Is Not It)
- The Best Steam Tools and Game Discovery Sites (Complete 2026 Directory)
- How I Find Good Steam Games Every Week (And What It Found This Time)
- The 6 Highest-Rated Indie Games on Steam Right Now (One Is at 100%)
- The Steam Refund Policy, Explained (Including the Cases Everyone Gets Wrong)
- 12 New Indie Games Released This Week on Steam (And Which Are Worth It)
