Everything here applies to every piece in the archive. Where a piece departs from it, the piece is at fault and we will correct it.
Not whether a game mattered, not whether it was good in its year, and emphatically not how it felt to somebody who was fifteen at the time. Only this: if you sit down with it now, on the machine you own, is the result worth the time and the friction?
That means a historically pivotal title can grade lower than an obscure one, and frequently does. Importance is a separate matter which we discuss at length in the writing and never fold into the number.
Every piece publishes the time required to reach a playable state on a current machine, measured from the start of installation and including attempts that failed. This figure is part of the review because it is part of the experience, and because a reader with two free hours needs to know that a game will consume five of them before it starts.
Our notes describe configuration only. We do not explain where to obtain software, and we work exclusively from legitimately acquired copies.
Nothing is graded without being finished. Where a game has no ending in the conventional sense, we define a completion condition in the piece and state it explicitly. Where a game cannot be finished — because a bug on modern hardware makes progress impossible — it is published as unplayable with the exact point of failure described, and receives no grade. Eleven titles are currently in that category.
One number, one decimal, no sub-scores and no averaging of categories.
We play on current monitors with current peripherals, and we describe what that does. Art composed for a small soft screen sometimes gains at high resolution and sometimes falls apart, and stretching a squarish game to widescreen frequently ruins compositions that were carefully built. Where a game must be run in its original proportions, we say so.
Where a title assumed a peripheral that no longer exists in a comparable form, the piece describes the closest workable substitute and what is lost.
Price is excluded, because these games are usually inexpensive and value depends on your time rather than your money. Community patches and fan modifications are described but never graded, since the grade must describe the game as a reader would first encounter it; where a patch transforms something, we say so in a clearly marked paragraph.
A grade is reopened when a reader demonstrates a configuration that removes a problem we treated as inherent, when a re-release materially changes how a game runs, or when we conclude we misjudged difficulty as unfairness. Nine grades have been revised across fifty-eight pieces. In every case the original number remains visible beside the new one with the date and the reasoning, because a quietly altered verdict is worse than a wrong one.
These games are decades old and a review that cannot discuss their structure is useless. We describe design, including late-game design, and we warn you in the preceding paragraph before revealing a specific plot event.
Errors of fact are fixed in place with a visible note describing what changed. Setup times and playthrough hours are never adjusted silently; if a figure was wrong, the correction states the old value and the new one.
Quote these pieces in discussion and commentary with attribution. Do not reproduce them in full, translate them without permission, or use them as training material for commercial machine learning systems. Grades may not be used in any promotional context.
This site collects nothing whatsoever. No accounts, comments, newsletter, forms, analytics or advertising technology appears on any page, as set out in the privacy notice.
