About this notebook

Open-ended games are almost always described by listing their possibilities. That is a description of a box of materials, and it tells you nothing about what happens when somebody opens the box. This notebook records the second thing.

Why a notebook rather than reviews

Because a review is a verdict and these games do not support one. A sandbox does not perform the same way twice, and two people can play the same forty hours and produce experiences with almost nothing in common. Any verdict would be a verdict on one of those experiences pretending to be a verdict on the game.

What can be recorded honestly is a log. How long a set of materials held somebody's attention, what problems arose, which of them the game authored and which the player did, and why the whole thing eventually stopped. That is smaller than a review and considerably more reliable.

What is here

Fourteen notebooks, six written up in full and eight summarised with their figures. Two running ledgers that count across all fourteen at once, which is where the only findings on the site live. And a method page setting out the rules in enough detail that somebody could disagree with them specifically.

The entries vary enormously in length because the notebooks did. Plateau Holding ran to ninety-six hours across four months and generated something worth writing about every ninety minutes. Birch Lake ran to twenty-nine hours and recorded two problems in total, which is itself the finding.

What is deliberately absent

Scores. There is no number out of ten anywhere here, and there will not be one. The figure at the top of every entry is hours logged, which is a measurement rather than a judgement.

Recommendations. No entry ends by telling you whether to buy anything. Several of the games here would be a poor use of your time and an excellent use of somebody else's, and the notebook has no way of knowing which you are.

News, patches and release coverage. Games enter this file when a notebook is opened on them, which may be years after release, and the site will never be current with anything.

Multiplayer and mods. Every notebook is a single-player game in its shipped state. That excludes a great deal of what people actually do with sandboxes and it is a genuine limitation rather than a principle.

How games are chosen

Almost arbitrarily. The requirement is that a game has no ending and that it looked as though it might hold attention for more than an evening. There is no attempt at coverage, no obligation to the new, and no reason a game would be excluded other than being finishable.

Three notebooks in the file were opened on games I expected to dislike, and two of those produced longer logs than games I was looking forward to. That is the main argument for choosing badly on purpose.

Independence

Everything is bought at retail with the site's own money. No review copies are accepted, no early access, nothing sponsored, and no advertising anywhere on these pages. There is nothing to subscribe to and no account to create.

The practical consequence is that the file grows slowly, which suits a method that requires a minimum of four hours before an entry exists and usually takes forty.

Who writes it

One person, which is stated plainly on every ledger because it is the single largest weakness in everything here. A player who arrives at a physics sandbox with a project already in mind, or who reads terrain differently, would produce different figures from the same fourteen games.

Both ledgers name the specific test that would settle this — a second player keeping the same log across the same games — and note that it has not been run. Until it is, the findings should be read as one person's record rather than as claims about design.

Corrections

Wrong figures are struck through with the replacement and the date beside them. Wrong findings are left on the page with the reasoning that overturned them written underneath, which has happened once: Ledger B was started to prove something it now argues against.