On May 14 at 08:00 PDT, the strange, slow-burn miracle of the original Subnautica finally gets a sequel. Subnautica 2 enters Early Access after a development saga messy enough to deserve its own documentary, and players will start logging hours into a build that, by the developer's own estimate, sits two to three years away from finished. Which raises a question every gaming journal eventually has to answer: when, exactly, is an Early Access game finished enough to review?
What May 14 actually delivers
Here is what lands on launch day. The new submersible is called the Tadpole, and you will use it to dip into a fresh batch of alien biomes. There is optional four-player co-op, which is itself a quiet revolution for a series whose first game treated solitude as a feature rather than a bug. Four pre-designed characters ship at launch, with more customization promised later in development. The game lands day-one on Steam, the Epic Games Store, the Microsoft Store, Xbox Series X|S, and Xbox Game Pass via the Game Preview program. Price is $29.99 USD with regional adjustments.
The developer, Unknown Worlds, is still owned by Krafton, but Krafton is conspicuously no longer publishing the title. That last detail has its own longer story involving a leadership shake-up, lawsuits, and the kind of corporate ugliness that tends to leak into release-week sentiment whether the game deserves it or not. We are choosing to set that aside and look at the build that is actually coming out.
Early Access keeps eating the calendar
Subnautica 2 is hardly alone in this. The original game itself spent more than three years in Early Access before a 1.0 in 2018 that everyone, including us, would call the version that "counted." Baldur's Gate 3 lived in Early Access from 2020 to 2023, and a sizable share of its eventual fanbase started playing it in Act 1 well before Acts 2 and 3 existed. Hades was in Early Access for almost two years and walked away with a Game of the Year nomination on the way out. Valheim has been in Early Access since 2021 and is still, technically, in Early Access. Vampire Survivors did the entire arc in roughly a year and changed the shape of an entire indie subgenre on its way to 1.0.
The point is not that Early Access is a marketing label anymore. It used to be a warning. Now it's a release model. Big-budget projects do it, AA studios do it, indie darlings do it, and a startling fraction of the games that meaningfully shaped the conversation in the last five years went through some version of it on the way to whatever 1.0 ended up looking like.
If you keep any kind of record of what you play, this changes the work. The simple, satisfying schema of "release date" and "playthrough" and "rating" assumes a finished thing being finished by you. Early Access shoves a third party into that equation: a game that is also still in flux, sometimes for years, that may look meaningfully different by the time you put it down than the version you first sat down with.
The journaling problem
This is where personal tracking gets interesting, in the way "interesting" is sometimes a polite word for "irritating."
Say you start Subnautica 2 at 8:01 Pacific on May 14. You poke around the first biome, build a small base, run out of oxygen in a way that feels meaningful, and keep going on weekends. Eventually you hit the end of the content that exists in this build — credits, or just the point where the world runs out of new things to do. Are you done?
By any honest definition of how you spend your evenings, you finished what was in front of you. But "finished" is a word that wants the game to also be finished, and Subnautica 2 won't be for another two or three years. If you mark it complete now and write the review, you are reviewing something that, by the developer's own description, is missing whole acts. If you wait, you might wait until 2028, and the version you actually fell in love with — the rough early build with the hostile creatures the team would later rebalance — won't be the one you're rating.
Then in October, the developers ship a content patch that adds a whole second act and rebalances the early game. Have you played that? You played a version of it. Does your earlier completion still hold, or does it need an asterisk now? When you do finally rate the game, are you rating the build you finished, or the game in aggregate, or some weighted average of every version you spent meaningful time in?
Most journals just punt on this. Steam tells you hours played and doesn't care whether the game is finished. Achievement systems pretend the game is fixed at the moment you earned each one. Star ratings pretend every game gets one definitive verdict, once. None of this is wrong, exactly. It is just thin.
How we think about it in Perthro
The reason we ended up building Perthro in the first place is that the existing places to keep a record of the games you play weren't quite making space for the way games actually arrive in our lives anymore. Early Access is one piece of that. So is the four-platform-deep library, the long-shelved JRPG you might pick back up in five years, and the indie someone recommended that you might never get to.
Our take on Early Access is the boring one and, we think, the honest one: a completion is whatever counted as one for you. If you played to the end of what existed when you played, mark it finished and write the review you have. The five-star rating is editable for a reason, and reviews are open-ended on purpose. When 1.0 lands in 2028, revise the rating, append to the review, or add a fresh entry alongside the original — the journal will still remember what you thought in May 2026, in the build with the strange oversized creatures, before the rebalance.
We deliberately did not build a "this game is in Early Access, you can't review it yet" rule. Forcing the schema onto people who play a lot of indie games would just push them somewhere else. The most useful gaming journal is the one that respects the player's actual relationship to a game over the publisher's marketing window — and a review you wrote at the time is more honest than one you couldn't write at all.
Why you should probably start it on May 14 anyway
Early Access is not just a logistical question. It is, at its best, the most interesting time to be playing a game. The early build of the first Subnautica in late 2014 was raw and a little broken and full of the strange, oversize creatures that the developers would eventually trim or rebalance, and a particular kind of player will tell you, with feeling, that the game was never quite as good after the polish. You don't have to agree with that to recognize the impulse.
A 1.0 build is a finished sentence. An Early Access build is a sentence being written in front of you. Knowing how the original got made (a small studio, a real ocean to draw from, a willingness to leave creatures hostile and unexplained), there is real reason to want to be in the room while the second one is still being written, even with the publisher chaos and the long road ahead. Especially with all that. The conditions around this game are unusual; the game will reflect those conditions, for better and for worse, and the version that ships at 1.0 won't.
So pick it up on May 14. Or don't. Either is honest. The point of keeping the record is not to file it correctly. It is to remember what you actually did, when you actually did it, and what it meant at the time. Subnautica 2 is going to be one of those games that means different things at different points along its arc. That is worth a journal entry, not a backlog ticket.
See you under the water.