Early Access. Whether it's worth it, when to skip, how to avoid traps

Early Access is a model where you buy a game before it's finished. You pay the full (or discounted) price for the right to play a pre-alpha, alpha, or beta version, hoping the game will grow into the final release. Steam introduced this model in 2013, today it's the standard release path for many indie developers.

I have 30+ Early Access purchases under my belt since 2017. 14 of them reached 1.0 status and are excellent games. 10 reached 1.0 but disappointed. 4 are still in EA after many years, 2 were abandoned, 1 the developer disappeared without a trace. From these statistics I draw lessons.

EA model classification

Three totally different types of Early Access worth distinguishing.

Type A: "Polished EA". Game practically finished, EA for feedback and last balance tweaks. Release in 1.0 in 3-6 months. Successful examples: Hades, Vampire Survivors, Slay the Spire, Inscryption.

Type B: "Active development EA". Game in mid-stage. EA for 1-2 years of active development with regular updates. Release in 1.0 in 12-24 months. Examples: Hades II, Valheim, Subnautica, Hollow Knight.

Type C: "Forever EA". Ambitious game, EA for 3-5+ years. Often the developer loses vision or money runs out. Successful examples: Project Zomboid (12 years EA, still developing well). Bad examples: Star Citizen (12 years EA, controversial), DayZ (4 years EA, weaker than promised).

First rule: buy only Type A, carefully Type B, avoid Type C.

How to recognize the type before buying

Four signals I look at.

Signal 1: Developer roadmap. Every EA game on Steam should have an "About This Game" section explaining the development plan. A specific roadmap (Q3 2025: feature X, Q4 2025: feature Y) suggests Type A or B. A "someday we'll add multiplayer and a new world" roadmap without dates = Type C.

Signal 2: Update frequency. Steam shows "Update history" under every title. An active developer is 1-2 big updates monthly. No updates for 6+ months = red flag.

Signal 3: "Recent" reviews. Steam splits reviews into "Overall" and "Recent (30 days)". If Overall is 90 percent positive but Recent is 65 percent, it means recent updates disappointed the community.

Signal 4: Developer reaction to feedback. Check the game's subreddit or Discord. If the developer responds regularly, discusses feedback, is engaged, good signal. If the subreddit is full of "developer ignores us" frustration, bad omen.

Three of my successful EA purchases

Subnautica (bought 2017, released 1.0 in 2018). EA price: 22 USD. 1.0 price: 30 USD. Savings 8 USD plus access to the game a year earlier. The game was already Type A at purchase, you could feel it. 80 hours of solid entertainment.

Hades (bought 2019, released 1.0 in 2020). EA price: 20 USD. 1.0 price: 25 USD. Plus a year of earlier play. The game was Type A with a great update tempo. The best EA decision I made.

Slay the Spire (bought 2017, released 1.0 in 2019). EA price: 14 USD. 1.0 price: 25 USD. Savings 11 USD plus 2 years of earlier play. Today the game has over 100k reviews at 96 percent positive. Classic.

Two failed purchases

For balance.

Wildermyth (bought 2019). Tactical fantasy RPG. Released 1.0 in 2021. EA price: 25 USD. I expected something on the level of Banner Saga. I got something plain. 30 hours of play, then put it down. The pre-1.0 decision was a mistake because today the game on sale is 12 USD, I could have waited.

Mount and Blade II: Bannerlord (bought 2020). I waited for this game 10 years (M&B Warband is a classic). EA price: 65 USD. Released 1.0 only in 2022, after 2 years of EA. During that time I had to patiently wait for patches. Today the game is great, but I paid 65 USD and went to play after 2 years. I should have bought in 2022 for 50 USD.

Total failure

Cyberpunk 2077 (technically not EA, but similar effect). Launch December 2020, price 70 USD. After the first week it required 1-2 years of fixing to be enjoyable. Classic example of "buying too early". Today the game is great, but 70 USD for a 1.0 product was an unusually bad buying decision. Lesson cost me 30 USD of savings.

Star Citizen (bought 2014 for 70 USD). Game still hasn't released. Star Citizen showed me the limits of Early Access. The developer actively collects money but the game doesn't finish. Today after 11 years I have 70 dollars "invested" in something I probably won't play.

Five rules I follow today

After 8 years of EA I developed rules.

Rule 1: Only if the EA price is at least 25 percent below the expected 1.0 price. If EA costs 25 USD and 1.0 will be 28 USD, there's no point buying EA. I wait for 1.0.

Rule 2: Only if the developer has a proven track record. Studio with a previous successful game (Supergiant Games, Larian Studios, Motion Twin) = safe. New studio with no history = risk.

Rule 3: Only if the game is playable in its current state. EA should be playable, not "I promise in a year it'll be fun". Hades in EA was already playable. Star Citizen in 2014 wasn't.

Rule 4: I check roadmap and recent updates. If the developer is active and has a plan, OK. If the last update was half a year ago, I wait.

Rule 5: Maximum 2 EA purchases a year. I keep myself in check. This keeps the wallet healthy and forces selection.

What to buy and what to avoid in 2026

Specific EA I recommend and warn against today.

Recommend:

• Hades II (bought in EA, released 1.0 fall 2025). Classic Type A. Safe.
• Manor Lords (medieval city builder, EA since April 2024). Type B with active developer.
• Project Zomboid (Type C but with 12 years of consistent update tempo). Special game, for those who love survival.

Avoid:

• Star Citizen. After 13 years of EA still not released. Donate, not purchase.
• Dwarf Fortress (Steam version). Works but huge learning curve. Not EA per se, but weird situation.
• Any new MMO EA. MMO in Early Access is practically always a flop. Statistics merciless.

EA refunds

It's good that Steam allows refund within 14 days from purchase / 2 hours of gameplay. This works for EA exactly like for 1.0 games. If you play EA for 30 minutes and see it doesn't fit, you have the right to a refund.

This is important insurance. If you buy EA and it turns out terrible, you can get your money back if you notice quickly.

Final thought

Early Access is a model that offers an exchange: savings of time and money in return for the risk that the game might never release or release in a weak state. With proper selection (Type A, proven developer, clear roadmap) the risk is minimal and returns high. Without selection it's burning money on promises, many of which won't be kept.

My statistics over 8 years: 14 out of 30 EA purchases were great decisions, 10 OK, 6 poor. That's 47 percent "great" decisions, better than 1.0 purchase stats (where I have about 40 percent of bought games never played). EA with thought is safer than "buying 1.0 mindlessly".

Just one final piece of advice: never buy an EA game you plan to "someday" play. Buy EA only if you start playing the first weekend after purchase. Otherwise you're buying a promise, not a product.