A few weeks ago I opened a review I wrote in 2021 about a game I had genuinely liked, and the entire review was the words "really good, kept me up till 2 AM." Four stars. Done. I had no recollection of which scene kept me up. I had no idea which character I had liked, what the controls had felt like, whether I had played it on the couch or in bed, whether I had been in a good mood that month or a bad one. I had successfully logged the game and lost almost everything I had actually wanted to remember about it.
The review wasn't useless. It told me that 2021-me had liked the game enough to give it four stars. That's about all it told me, and it told me with the energy of a receipt. Five years on, the receipt is what's left, and the experience is mostly gone.
This post is a small set of habits I have slowly accumulated, mostly by writing down the wrong thing for a long time, that make a game review worth returning to. They are not rules. They are more like prompts: small ways to push past the verdict and onto something more specific, so that the review you write tonight is still an artifact you can read in 2031 and find yourself in.
The star rating stops being you
The star rating is the most efficient thing you can write about a game and the least durable. Three weeks after you give a game four stars, you can still tell me what you meant by it. Three years later, four stars means almost the same thing as five stars or three: you liked it more than you didn't. The fineness of your judgment, which felt obvious when you closed the game on a Sunday night, has decayed into a coarse "yes."
This is fine for some uses. If all you want is a yes-or-no record of which games you have completed and roughly how they sat with you, a star is enough. But the star is doing different work than the review is, and confusing the two is the source of most of the writing-to-yourself problems I have ever had with a logging app. The star is a sort of bookmark. The review is the thing the bookmark is supposed to point at.
If you find yourself writing a review that is just a longer way of restating the star rating ("really good," "decent," "not for me"), you have written a bookmark with extra steps, and the version of you in 2031 will get nothing from it. The fix is not to write more words. It is to write different ones.
Write the room, not the verdict
The single best habit I have for writing a review I will come back to is to put one concrete sentence about the room in it. Where I was, when I played, what was around me, what I was avoiding, who else was in the apartment.
"Played this on a hot Sunday afternoon, fan on, with the patio door open and the dog asleep on the rug." "Started this on the bus to work, two stops past my stop because I did not want to put it down." "Played in two-hour bursts after dinner for most of November, while the kitchen was being painted."
This sounds incidental. It is the part of the review that ages best, by an unreasonable margin. The combat in the game is a thing that gets discussed in fifty other reviews on the internet; the apartment you played it in is yours, and nobody else has written it down. When you read it five years later, the apartment is the part that comes back. The combat you mostly remember from screenshots.
This is also the piece that mid-period reviewers tend to skip, because it feels self-indulgent. It is not self-indulgent. It is the only part of the writeup that is genuinely about the experience of playing the game, as opposed to the qualities of the game-as-object. The qualities of the game-as-object are public information by the time you write about them. The qualities of your evening with the game are not, and you are the only person ever going to record them.
A small habit that makes the rest stick
The other habit, almost as load-bearing, is to spend ten seconds in the moment you put the controller down. Not ten minutes. Ten seconds. Open the app, the note, the page; write the first specific thing your brain hands you. "The boat scene." "The shopkeeper in the third town." "The weird difficulty spike at hour twelve." It does not have to be the most important thing about the game. It has to be specific.
The reason this matters is that the first specific thing your brain hands you, in the minute after the play session, is almost always more interesting than the considered take you write a week later. The considered take is closer to consensus. It has been polished by the time you spent thinking about it on the way home, the review you skimmed on Saturday, the YouTube video you half-watched while making coffee. The first specific thing is yours, and uncontaminated, and very often the only piece of writing about that game on the internet that contains it.
You can write the long review later, if you want one. The seed sentence in the moment is what makes the long review possible at all. Without it, the long review is a summary of what other people thought.
Length is freedom, not assignment
A surprising amount of bad game writing comes from confusion about how long the review is supposed to be. Star-rating apps imply that one or two sentences is enough. Long-form sites imply that you need three paragraphs and an opener with a thesis. Neither is right, because there is no fixed answer. A four-hour indie that hit you in a particular way might earn three pages. A sixty-hour open-world thing you mostly enjoyed might earn one good sentence.
What seems to make the difference, in my own log, is whether I had something specific to say. When I have one specific thing to say I write a sentence and stop. When I have ten I write four paragraphs. The reviews that are worst, every time, are the ones where I felt I should write some amount and reached for filler to get there.
This is, transparently, why we built Perthro the way we did. Reviews can be a sentence or three pages on purpose. There is no word counter, no minimum, no "make this more thoughtful" prompt. Whatever you have to say is the right length, and you can come back later to add to a review when the game has done its slow work in your head. The TestFlight is open if you want to keep your journal there; iOS 16 and up, free during the beta, with the invite at testflight.apple.com/join/XVxCdRcK. A note app is fine too. The point is that nothing about the tool should be telling you how long to be.
What to do when your taste changes
The other thing nobody warns you about, when you start keeping a real game log, is that your taste will change underneath you, and you will eventually open a five-year-old review and not recognize the person who wrote it. You used to love the long open-world thing. You don't anymore. You used to bounce off short narrative games. You don't anymore. The 2021-you was somebody else, and the review they wrote sounds wrong to you now.
The temptation, when this happens, is to edit the old review to match the current take. Don't. The 2021 review is a true artifact of the 2021 you, and the most interesting thing about a long-running journal is the gap between then and now. You will, in fact, want to read those embarrassing earlier takes someday, because the embarrassment is the data. It is how you will know your taste actually moved, instead of feeling like it always was where it currently is.
If you want to add a current take, add it. Don't overwrite the old one. Add a dated note, append a new paragraph, write a fresh review of the same game and let both stand. The shape you want is not a single review that gets revised, but a small stack of dated takes that show, over time, what the game actually became to you. Most journals can be made to do this if you decide to use them that way.
A few last things
A short list of small things, for anyone who has gotten this far:
- Use names. "The wizard." "The forest stage." "The third boss." Specifics age well. "Combat" does not.
- Write at the wrong time of day on purpose. Mornings are more honest than midnights about whether a game is actually good or just absorbing.
- Don't apologize. "I know everyone hated this but" is a marker that you are writing for an imagined audience instead of for yourself. Cut it.
- It is okay to write a review for a game you haven't finished. "Twenty hours in, here's what I think" is an honest thing to log, and you can update it later.
There is no correct way to keep a game journal. The version that works is the version you will actually keep, and the version you will actually keep is the one that is specific enough to feel like writing and small enough to feel like a habit. Star ratings will fade. Verdicts will age. The room you played in, the first sentence in the minute after the controller went down, the sentence you wrote to yourself instead of to the internet: those keep.
Five years from now, the review you write tonight is going to be the only record of what tonight was actually like. Make sure it sounds like you.