Event[0]

First Impressions

Event[0] sits at Mostly Positive on Steam: 79% of 1,439 reviews (snapshot 2026-09-12). Counting every language and purchase type, it is 1,838 of 2,287, or 80.4%. Either way it rests just above the eighty-percent line.

The breakdown is the interesting part. Read the pool in helpful order and both camps describe the same scene. A retrofuturist ship. A terminal keyboard. You type a sentence, and Kaizen-85, the ship's AI, types back.

The top positive reviews call that reply a dynamic personality. The top negative reviews call the same reply ELIZA from decades ago. One writer says they cried. Another says it is not worth twenty dollars.

So the split is not about quality. It is about what to name one mechanism. As an experiment in naming, this 2016 game still repays reading.

Event[0] in play — Steam store screenshot, first in the store gallery

The World

The vocabulary the pool uses for this ship is shared across both camps: stylish, retrofuturist, eerie. The setting is a 1980s idea of space that simply kept running.

The same ship also draws blurry and low-resolution textures, and not only from the negative side. Several positive reviews concede the roughness and recommend it anyway, which says a lot about how this game is received.

The frequency of atmosphere words sits close to the pools around Tacoma and Observation. Quiet space stations are an established value here. What is unusual is that this quiet has someone in it to talk to.

Event[0] in play — Steam store screenshot, second in the store gallery

Putting the Mechanics into Words

Count the verbs the player is given and there is effectively one: type. Walking and looking are accessories. You cannot run, jump, or pick anything up.

The most reproducible complaint on the negative side lands exactly there. No strafing, no jumping, you feel like a ghost. That is the cost of subtraction: to narrow the verbs to one, the rest were dropped.

So how fine is the one remaining verb? The store copy promises over two million procedurally generated lines. The negative side says it matches a limited set of keywords. What matters is that the positive side does not dispute this — top positive reviews reach for chatbot style and rickety scaffolding on their own.

The argument is not about how thick the machinery is. It is about whether a personality stands on thin machinery. The player learns too, learning how to phrase things so Kaizen understands. The negative side calls that dumbing yourself down. The positive side calls it building a relationship. One learning curve, two names.

Event[0] in play — Steam store screenshot, third in the store gallery

Pacing

On length, both camps report the same figure. A first ending takes about three hours, and recorded playtimes cluster between three and five.

Only the adjective changes. Positive reviews write short and sweet. Negative reviews write an intro, or a first episode. A twenty-dollar list price amplifies that gap, and even positive reviewers often say to wait for a sale.

One complaint belongs to the negative side alone. There are three endings, each requiring a full replay, and door codes are randomized between runs. Someone wrote repeat, repeat. The attempt to cover a short runtime with structure is where most of the friction comes from.

Event[0] in play — Steam store screenshot, fourth in the store gallery

Place in the Lineage

The game shipped in September 2016 and was nominated at IGF 2017 for the Grand Prize, Excellence in Design, and Excellence in Narrative. Critic averages sit near 75 of 100, and the praise concentrated on treating an AI as a partner rather than a villain.

The interesting movement is in the recent sort. Reviews written in 2026 measure the game against today's large language models almost without exception: good enough for 2016, disappointing next to what we have now. What astonished people ten years ago has become a household utility.

But the recent pool also runs the other way. Some writers now list it as a virtue that these lines were written by people rather than generated. The same technology becoming common subtracts points for one reader and adds them for another.

Among games that make conversation the mechanism, Eliza and Duskers sit on either side. Duskers removes ambiguity by letting you address only machines; Eliza removes typing altogether. Event[0] keeps the ambiguity and makes only the listener finite. The friction comes from there.

Event[0] in play — Steam store screenshot, fifth in the store gallery

Reviews Consulted

This piece was written from the Steam review pool as it stood on 2026-09-12. No review text is quoted directly; recurring claims are reconstructed.

・Steam: Event[0] (Mostly Positive, 79% of 1,439 reviews; across all languages and purchase types, 1,838 of 2,287 = 80.4%)

・Read in full: the top 10 helpful positive reviews, the top 10 helpful negative reviews, and the top 6 sorted by recency.

・Awards and the critic average are taken from Wikipedia: Event 0 (Metacritic 75/100, three IGF 2017 nominations).

Closing

Steam reads 79% positive. I give it 7.2, which lands in nearly the same place, because the weakness and the strength come from one decision.

A free-text box hands the player an unbounded set of moves. What receives them can only be finite. Handling that gap is the central design problem for any game that makes conversation its mechanism. Ocelot Society left the gap visible and asked atmosphere to carry it. The reviewer who cried and the reviewer who felt let down are both results of that choice.

The audience is clear: people interested in the relationship itself, unwilling to pay list price for three hours, and unbothered by soft textures. Buy it for the puzzling and you will miss. Ten years on, that has not changed.

Event[0] in play — Steam store screenshot, sixth in the store gallery