The year was 2007. I still remember the feeling of finishing Episode 2. That gut-wrenching cliffhanger—Eli Vance dead on the floor, Alyx sobbing, and the Borealis waiting somewhere in the frozen wasteland. It felt like a promise. We all thought Half-Life 3 (or Episode 3, as it was then known) was right around the corner. Maybe a year? Two at most?
Nobody expected a decades-long silence.
But honestly, the story of why we never got that game is way more complicated than "Valve got lazy." It’s a mix of engine trouble, massive creative shifts, and a company culture that basically lets developers work on whatever they want. When you don't have a boss telling you to finish a project, and the project becomes the most anticipated sequel in history, people get paralyzed. They get scared of failing.
The Half-Life 3 Script That Actually Leaked
Let’s talk about 2017. That was the year Marc Laidlaw, the lead writer for the series, posted a "fan fiction" story on his blog titled Epistle 3. He changed the names—calling the protagonist "Gertie Fremont"—but everyone knew what it was. It was the outline for the lost game.
In that version, Gordon and Alyx head to the Arctic to find the Borealis, that mysterious Aperture Science ship. They find it phase-shifting in and out of existence. There’s this massive battle with the Combine, and eventually, they realize the ship is a suicide weapon. They try to crash it into the heart of the Combine’s command center. But right at the end, the G-Man shows up, ignores Gordon, and takes Alyx instead. Gordon gets dumped back on a beach, years later, alone.
It was dark. It was weird. It was very Valve.
Why Epistle 3 Wasn't the End
Some people think that script leak was the final nail in the coffin, but internally at Valve, things were already messy. According to Geoff Keighley’s The Final Hours of Half-Life: Alyx, there were at least five or six different attempts at a third game between 2007 and 2020.
💡 You might also like: All Barn Locations Forza Horizon 5: What Most People Get Wrong
One version was going to be a procedurally generated VR game. Another was a traditional shooter. But every time they started, the Source 2 engine wasn't ready. You can't build a revolutionary game on a broken foundation. They tried to build the engine and the game at the same time, which is basically like trying to renovate a house while the walls are still wet. It doesn't work.
The "Source 2" Bottleneck
Building a game engine is hard. Building one that satisfies the people who made Half-Life 2 is almost impossible.
The team kept hitting walls. In the early 2010s, Valve shifted heavily toward multiplayer. Dota 2 was exploding. Counter-Strike: Global Offensive needed a massive overhaul. The "cabal" system at Valve—where employees move their desks to the projects they find most interesting—meant the single-player team just kind of... evaporated. If you're a developer, do you want to work on a legendary sequel where the fans will hate you if it’s not a 10/10, or do you want to work on a skin system for CS:GO that makes the company millions of dollars a day?
It’s a tough choice.
The VR Pivot
Everything changed when Valve decided to get into hardware. When the Index headset was in development, they realized they needed a "killer app." They looked at their IP and realized Half-Life 3 was too big of a risk for a niche platform like VR. They didn't want to lock the "final chapter" behind a $1,000 headset.
So, they made Half-Life: Alyx instead.
📖 Related: When Was Monopoly Invented: The Truth About Lizzie Magie and the Parker Brothers
And here’s the thing: Alyx actually changed the ending of Episode 2. If you've played it, you know. If you haven't, stop reading. Seriously. Spoilers ahead.
By the end of the VR game, the timeline is altered. Eli Vance is alive. Gordon is back. The G-Man has a new pawn. Valve essentially hit the "undo" button on the 2007 cliffhanger so they could start fresh. They cleared the deck.
What's Actually Happening Now?
Is it coming?
Rumors have been swirling again because of a project codenamed "HLX." Data miners like Tyler McVicker have found references in Valve's recent code—things like "HEV suit" mechanics, gravity physics, and non-VR movement. This suggests a traditional, mouse-and-keyboard Half-Life 3 is actually in active development.
But we’ve been here before.
Valve operates on "Valve Time." They don't announce things until they are 90% done because they want the freedom to cancel a project if it sucks. They’ve cancelled more games than most studios have ever released. They cancelled a version of Left 4 Dead 3. They cancelled an RPG. They probably cancelled three different versions of Episode 3.
👉 See also: Blox Fruit Current Stock: What Most People Get Wrong
The Problem With Perfection
The biggest enemy of this game isn't the Combine. It's the legacy. Half-Life 1 reinvented how we tell stories in shooters. Half-Life 2 reinvented physics and facial animation.
What can a third game do today?
Ray tracing? Better AI? That’s not enough for Valve. They want a "paradigm shift." And in a world where games like Cyberpunk 2077 and Elden Ring have already pushed boundaries so far, finding that next "big thing" is a nightmare.
Dealing With the "Half-Life 3" Hype
If you're waiting for a trailer to drop during a random Tuesday, don't hold your breath. Valve usually reveals things when they have a piece of hardware to sell or a major platform update.
Honestly, the best way to track the progress of Half-Life 3 isn't by looking at "leaked" posters or "insider" tweets. Look at the Steam Database. Look at the updates to the Source 2 engine. When you see new strings of code for "Xen flora" or "gravity gun interactions" appearing in Dota 2 patches, that's when you know something is moving in the shadows of the Bellevue offices.
What You Should Do While Waiting
Don't just sit around refreshing forums. The community has actually done some incredible work keeping the spirit alive.
- Play Black Mesa. It’s a fan-made remake of the first game that Valve actually let them sell on Steam. It’s better than the original in many ways, especially the Xen levels.
- Check out Project Borealis. This is a group of fans trying to build the "Epistle 3" script into a real game. They recently released a tech demo, and the physics are surprisingly spot-on.
- Finish Half-Life: Alyx. Even if you don't have VR, there are "No VR" mods now. You need to see that ending to understand where the story is actually going.
- Read "The Final Hours of Half-Life: Alyx" by Geoff Keighley. It’s an interactive book on Steam. It’s the most honest look you’ll ever get at why the company struggled for a decade.
The reality is that Valve is finally in a position where they can make the game. The engine is stable. The story has been reset. The team is energized by the success of Alyx. It might not be called Episode 3 anymore, and it might not look like what we imagined in 2007, but the dream is arguably more alive now than it was five years ago.
Keep an eye on the "HLX" leaks. They are the most concrete evidence we've had in a generation.