Some ideas are too good to abandon. They sit quietly in a developer's memory for years, waiting for the right project, the right team, the right moment. The dynamic scare mechanic that almost made it into BioShock 2 back in 2010 is a case in point, and the story of how it has finally found a home is a genuinely interesting window into how creative instincts survive the pressures of big-budget game development.
David Lindsey Pittman, who worked as an AI programmer at 2K Marin, writing AI code for BioShock 2, has revealed in a Bluesky thread that he and colleague Kent Hudson once discussed building what they called a dynamic "Big Sister scare" system for the 2010 sequel. The concept was straightforward but clever: the game's primary villain would appear at the edge of the player's vision whenever the game's tempo dipped, keeping tension perpetually alive rather than relying solely on scripted set-pieces. In the end, the idea was scrapped, and the Big Sister's appearances were all scripted.
That might have been the end of it. Instead, Pittman says the idea "stuck around in my brain ever since", and it has now resurfaced in a very different kind of game. That project is The Killing Stone, currently in early access development by Question, an indie studio formed by BioShock 2's director Jordan Thomas. Pittman has worked with the team at Question since 2017, contributing to The Killing Stone, South Park: Snow Day!, and The Blackout Club.
The Killing Stone is an occult horror deckbuilder from Question, featuring voice talent from Baldur's Gate 3 and Critical Role, and is available in Early Access on Steam for PC. A full 1.0 release is slated for later this year. The game's premise is as distinctive as its pedigree. It takes place in the 17th century and follows the Maven, who is investigating the death of Mariken Svangard; the investigation quickly reveals a curse that has afflicted the family for generations, and in order to break the cycle of damnation, the Maven engages in card battles with various demons, with the family's souls at stake.
The game has two halves: one sees players engaging in a virtual board game where cards are played to spawn living game-pieces to attack the opponent's pieces, while the other involves exploring an old, snowbound manor house and conversing with its inhabitants. It is in this exploratory half where the resurrected scare system almost certainly does its work. "CJ (artist) and Gretchen (engineer/designer) built a scare system, and it has added so much life to the level," Pittman writes. "It's very fun as a gamedev to be surprised by something in your own game."
The detail about being surprised by your own creation is worth pausing on. It speaks to a quality of emergent design that is genuinely rare, particularly in the indie space where teams are small and every system is intimately familiar to the people who built it. Question's games have tended to fly under the radar, but they have always had interesting ideas; The Blackout Club ran a fascinating "Enhanced Horror" experiment where the developers roleplayed as gods who would speak to players during play, which could then feed back into other players' sessions. A studio with that kind of experimental ambition is precisely the right environment for a mechanic that, had it shipped in 2010, would have been genuinely ahead of its time.
A dynamic villain-at-the-edge-of-vision system would have been an innovative approach to first-person horror in 2010, the same year that brought Amnesia: The Dark Descent. Amnesia became a touchstone for horror game design because it removed the player's ability to fight back, forcing reliance on evasion and atmosphere alone. A dynamic spawn system for BioShock 2's Big Sister could have achieved something similar within a more action-oriented framework: the threat ever-present, appearing not on a designer's schedule but in response to the player's own pacing. Pittman does not specify whether these dynamic scares would have resulted in an actual fight or chase sequence. The ambiguity makes the concept even more intriguing.
This is far from the only distinctive feature of The Killing Stone, which also lets players choose between modern English and a 17th-century dialect for its dialogue and voice acting. Voice performances come from Emma Gregory of Baldur's Gate 3 and Liam O'Brien of Critical Role. The dual-language option alone signals a studio willing to treat its audience as people who want genuine creative craft, not just genre comfort.
The broader story here is one that any creative industry professional will recognise: the sheer volume of good ideas that get cut from large commercial productions, not because they are bad, but because scope, time, and organisational priorities make them untenable. Question is a small independent studio known for The Magic Circle and The Blackout Club, with a design philosophy built around giving players agency to tell their own stories; the team is led by Jordan Thomas and Stephen Alexander, both known for their lead roles on the BioShock series. In the indie environment, with a small trusted team and no publisher demanding scope cuts, an idea like Pittman's dynamic scare system can finally be given the room it deserves.
Whether The Killing Stone ultimately delivers on that promise is a question only the finished product can answer. What is already clear is that the game represents something genuinely unusual: a creative lineage running directly from one of the most ambitious first-person horror titles of the last two decades into an occult card game that nobody saw coming. The Steam Early Access release suggests the community will have a meaningful hand in shaping what the final version becomes. The studio plans to give players access to over 90 per cent of the cards and mechanical combinations, making changes to balance and tuning based on community input. That collaborative spirit, combined with a decade-old idea finally getting its shot, makes The Killing Stone one of the more interesting early access propositions of 2026.