A score is a summary, not an argument. This page defines each band precisely so that a 7.4 on this site cannot be mistaken for a failure and a 9.0 cannot be mistaken for perfection.
Every review lands on a single number between one and ten, given to one decimal place. There is no weighted formula behind it and no category averaging, because averaging produces scores that describe nothing. The number is a judgement, and the review is the reasoning.
| Band | Meaning | Buying advice |
|---|---|---|
| 9.0 – 10 | Genre-defining. Does something no competitor does, and does it well enough that its flaws are footnotes. | Buy at full price. |
| 8.0 – 8.9 | Strong, confident design with clearly nameable weaknesses. The kind of game we recommend without hedging. | Buy at full price if the genre appeals. |
| 7.0 – 7.9 | Excellent in one dimension, uneven in others. Usually a great twenty hours inside a forty-hour structure. | Worth it on sale, or at full price for the specific strength named in the review. |
| 6.0 – 6.9 | Interesting ideas held back by execution, scope or pacing. Not a bad game; a compromised one. | Wait for a discount and read the flaws list first. |
| 5.0 – 5.9 | Functional but thin, derivative, or noticeably unfinished at the point we reviewed it. | Only if the concept specifically appeals to you. |
| Below 5.0 | Broken, hostile to the player, or hollow enough that we cannot recommend spending the hours. | Skip it. The review explains which of the three applies. |
Price is not part of the number. A twenty-hour game at a low price and a sixty-hour game at full price are judged on what they do, not on value per hour, because value depends entirely on how much you play. Where price genuinely changes the recommendation, the review says so in plain words instead of bending the score.
Nor does the score account for potential. A promised roadmap, an announced expansion or a developer's stated intention to fix something has no effect on a verdict. We review the game that exists on the build we finished. If it improves, we come back.
Published numbers are not permanent. A review is reopened when one of the following happens:
When a score moves, the old number stays visible alongside the new one with the date of the change and a short paragraph explaining what shifted. We do not silently edit verdicts.
Three categories are covered without a number. Early access and open beta builds receive impressions only, because scoring an unfinished product misleads readers and pressures developers. Games we abandoned before the end are published as play logs with an explanation of why we stopped. Live-service titles are reviewed in seasons, with each write-up naming the season it covers rather than pretending to judge a moving target.
If a writer has any personal or professional connection to a game or its developer, they do not review it, and the connection is noted on the review that somebody else writes. Review keys are logged and disclosed at the top of the play log. No writer on this desk accepts paid work from a studio whose games they cover.
