Can You Make a Steam Game in Two Months? Yes — But the Genre Chooses You

Tools

Short answer: yes — and one just did it. But it worked because the developer picked a genre that had already solved its hardest design problems in public.

This week a game called Three Kingdoms: Yi entered the Steam 250 seven-day chart at #18, up four places, with a 7.54 score from 119 reviews at 95% positive. The first line of its Steam store page is the claim: solo-developed in two months. That number is worth taking seriously, and worth taking apart.

What the two-month game actually is

Three Kingdoms: Yi is a 2D turn-based roguelite deckbuilder. You command a sovereign across 8 provinces and 49 cities, form a battle line, sway enemy generals to your side and awaken an avatar. It sells for $8.37, supports Steam Workshop, and is written in Simplified Chinese with English subtitles.

Claim	Verifiable result
Solo-developed in two months	Developer statement on the Steam page; team size not independently confirmed
Released 22 October 2026	Confirmed via Steam store API
Quality held up	95% positive from 119 reviews, #18 on the weekly chart
Priced like a two-month game	$8.37, discounted from $9.30

Why a deckbuilder is the genre that allows this

Two months is not enough time to invent a genre. It is enough time to execute one well, and deckbuilders are unusually generous to a short cycle for three structural reasons.

The conventions are already public

A developer starting a roguelike deckbuilder in 2026 does not have to discover that draw-discard-shuffle works, that relics should change rules rather than numbers, or that a run needs an escalating map. Those answers were worked out by games the audience has already played. The design risk that normally eats the first six months of a project has been pre-paid by the genre, so the developer starts at the implementation stage rather than the prototype stage.

Content is numbers and text, not art and animation

This is the real cost saving. A card is a name, an illustration slot, a number and a rule. A platformer level is geometry, art, collision, scripting and testing. One developer can author a hundred cards in a sprint; nobody authors a hundred levels in a sprint. When your content is data rather than assets, content volume scales with typing speed instead of production pipeline, which is why 49 cities is a plausible claim for a solo two-month project and 49 hand-built dungeons would not be.

Runs replace levels, so replay covers for scope

A roguelite asks the player to see the same content repeatedly and rewards them for it. That inverts the usual scope problem: instead of needing twenty hours of distinct content for twenty hours of play, you need a system whose recombination stays interesting. Three Kingdoms: Yi ties deck changes to territory, so the map is the upgrade system and a single mechanic has to carry the campaign. That is a cheap design in asset terms and an expensive one in balance terms — and balance is the thing a solo developer can actually iterate on alone.

What two months cannot buy

This is the part the success story leaves out, and it matters more than the headline.

A verified claim. The two-month figure comes from the developer’s own store page. There is no changelog, no devlog and no third-party reporting confirming it, and “two months” is excellent marketing copy. Treat it as a claim, not a measurement.

A durable review score. 95% from 119 reviews is a real signal, but a small one. Early reviews skew toward people who sought the game out deliberately, which is the friendliest possible audience. Scores on games this size routinely move several points once a wider audience arrives.

Localisation depth. Full audio exists in Simplified Chinese only. English players get subtitles. That is a sensible cut for a short cycle, and it is a cut.

An audience. Nothing about a two-month build shortens the marketing timeline. Wishlists accumulate on calendar time, not development time, so a fast build mostly means arriving at launch with less of a runway than a slower project would have had.

A second game. Short cycles work once per genre convention. The savings here came from conventions the developer did not have to invent; a follow-up that wants to differentiate pays that bill back.

How to tell whether your idea fits a short cycle

The question is not how fast you work. It is whether your content is data or assets, and whether your genre has settled answers you are allowed to borrow.

Trait	Short cycle is realistic	Short cycle will fail
Content type	Cards, items, numbers, text, procedural layouts	Hand-built levels, cutscenes, voice acting, bespoke animation
Design risk	Genre conventions are public and proven	Core loop is novel and needs discovery
Playtime source	Replay and recombination	Linear authored progression
Art demand	2D, stylised, or UI-forward	3D environments, character rigs, lighting
Failure mode	Needs balancing, which one person can iterate	Needs content volume, which one person cannot

Deckbuilders, incrementals, puzzle games and score-attack arcade games sit on the left. Narrative adventures, metroidvanias and open-world survival games sit on the right, and the chart reflects it: the fastest-built games that chart well are almost never the ones with the most authored content.

Frequently asked questions

How long does it take to make an indie game on Steam?

There is no single answer, but the spread is wide and genre-driven. Deckbuilders and incrementals have shipped to good reviews in two to six months. Narrative adventures and metroidvanias more often take two to five years, because their content has to be authored by hand rather than generated from rules.

Can one person really make a Steam game in two months?

The developer of Three Kingdoms: Yi says they did, and the result holds a 95% rating from 119 reviews. The claim is unverified beyond the store page, but the genre makes it plausible.

Which genres are fastest to develop?

Genres where content is data rather than hand-built assets: roguelike deckbuilders, incremental games, puzzle games and arcade score-attack games. They reuse public design conventions and get playtime from replay instead of authored levels.

Does a short development time hurt review scores?

Not inherently. Players review what they played, not how long it took to build. What does hurt scores is shipping a genre that needed more content volume than the schedule allowed.

Is it better to ship fast or polish for years?

It depends which resource is scarce. Shipping fast buys you real market feedback and revenue earlier, at the cost of a shorter wishlist runway. Polishing longer buys authored content and audience build-up, at the cost of runway of a different kind. Neither is universally right, and the genre usually decides for you.

How much should a two-month game cost?

Three Kingdoms: Yi launched at $9.30 and discounted to $8.37. That band is typical for short-cycle deckbuilders on Steam, where sub-$10 pricing works with the two-hour refund window rather than against it.

Sources and citation

Three Kingdoms: Yi store page, including the solo-developed-in-two-months claim, price and language support
https://store.steampowered.com/app/4771880/
Steam 250 seven-day chart, for chart position, score and review count
https://steam250.com/7day

Can You Make a Steam Game in Two Months? Yes — But the Genre Chooses You

What the two-month game actually is