Abstract
Problem: Why did The Temple of Elemental Evil become Troika Games' worst-reviewed title and the lowest-scoring game of Tim Cain's career?
Approach: Tim Cain walks through the full development history, taking personal responsibility while explaining the cascade of decisions, contract limitations, and external pressures that shaped the final product.
Findings: The game suffered from a compressed 18-month timeline (later extended to 20), a split team (with Leonard Boyarsky and Jason Anderson on Bloodlines), mid-development rule changes from D&D 3.0 to 3.5, retroactive content removal by the publisher at month 16, inadequate publisher QA, contractual blocks on post-launch patches, and Tim's four-month kidney stone ordeal during the final stretch.
Key insight: Tim identifies two core lessons — he's a terrible businessman who shouldn't negotiate contracts, and he's not a strong narrative writer (though excellent at lore, quest ideas, and systems). Failures teach more than successes.
1. How Troika Split Into Two Teams
After Arcanum shipped, Troika's pitches for Journey to the Center of Arcanum and a Lord of the Rings game both fell through. In early 2002, Activision and Atari both approached Troika wanting licensed games — Vampire: The Masquerade and D&D respectively. Both required all three principals (Tim Cain, Leonard Boyarsky, Jason Anderson) as key men.
Rather than choose one, they decided to do both to avoid putting all eggs in one basket. Leonard and Jason took Bloodlines; Tim took Temple of Elemental Evil. This meant Tim lost his two longtime collaborators and had to hire new staff — many of whom had never played D&D. He ran Wednesday afternoon D&D sessions for a month or two to teach the team, improvising a short adventure about recovering an ancestral tapestry from bandits and a necromancer. Several team members later said they wished they'd made that game instead.
2. The Contract Problems
Tim chose Temple of Elemental Evil out of nostalgia — he'd been playing D&D since age 14 and wanted to do a Gary Gygax module in Greyhawk. He assumed that since Wizards of the Coast published the module, everything in it would be pre-approved. He was wrong, and he didn't put that assumption in the contract.
The timeline was 18 months — extremely tight. For comparison, Fallout: New Vegas was also made in 18 months but had over twice as many developers, a working engine with existing system mechanics, and pre-made art assets. Troika was building on a modified Arcanum engine, adding 3D characters, and implementing all of D&D 3rd Edition from scratch — every class, spell, and monster. Tim asked to drop Druids and Bards due to their complex special abilities. Atari said no.
3. Mid-Development Rule Change
Partway through development, Wizards of the Coast announced D&D 3.5 Edition. Tim believes they must have known about it when the project started. WotC pushed for the game to use 3.5 instead of 3.0. Tim asked for three additional months; they gave him two (extending the project to 20 months total). The differences between 3.0 and 3.5 were substantial.
4. The Alignment Vignette System
The team developed a group alignment system to prevent incompatible party compositions (e.g., paladins and assassins together). Players chose a group alignment, then individual alignments within one step of it. Tim wanted a one-hour opening vignette for each of the nine alignments, explaining why that group was traveling to Hommlet. They ended up being only a few minutes each. Tim considers this a feature he should have cut — he admits to latching onto ideas and refusing to let go.
5. Retroactive Content Removal
At month 16 of 20, Wizards of the Coast retroactively unapproved previously approved content. Some changes were minor (NPC name swaps), but two were devastating:
Children had to become invulnerable. This created cascading mechanical problems — do children provide cover from ranged attacks? Do they count toward cleave requirements? Rather than chase every edge case, Troika simply removed children from the game, leaving dialogue references to kids that don't exist in the world.
Nulb content was gutted. An entire brothel and associated quests, dialogues, and characters had to be removed. Rewriting everything would have taken too long, so they sealed off buildings. This is why Nulb feels incomplete with inaccessible structures.
6. Publisher QA Failures
Around month 18, Tim was playing an all-halfling, all-bard party and hitting bugs from the start. He contacted Atari's QA department — he had given them specific test plans covering classes, races, abilities, and encounters with special monster mechanics. Their response: "We don't like playing bards. We haven't played them."
The game shipped with a critical bard bug: transitioning maps while a bard song was playing would cause every subsequent map transition to crash. Tim had never encountered it because he always turned off songs before transitioning. The same QA problem recurred a year later on Bloodlines, where Nosferatu weren't being tested.
7. The Kidney Stone
A few months before shipping, Tim developed a kidney stone. Most pass in days; his took four months. A nurse told him she'd had three children and a kidney stone, and would rather have another child than another stone.
Tim needed painkillers so strong he couldn't drive. His routine: arrive at work early with his dog, take a painkiller, wait 30–60 minutes, work all day, drive home, take another, eat dinner, often work again from home, take a third before bed. Team members from both the Temple and Vampire teams grew concerned enough to consider an intervention. The painkillers affected his personality and output. The stone (7mm — large) finally passed one morning at the office. Tim ran around showing it on a napkin to the only two people there: Long Win (an artist he later worked with at Carbine and Obsidian) and Daniel Alpert (an intern who became art director on The Outer Worlds). Neither wanted to look at it. His doctor's office performed a destructive test and wouldn't return it.
8. The Patch Fiasco
Troika developed multiple patches after launch — at least two never shipped. The contract required publisher approval for every patch. Atari decided they didn't want to pay for more QA, and without QA signoff they wouldn't approve patches. Troika offered to pay for QA themselves — Atari said no. They asked to release unofficial patches — Atari said no. When this leaked, it was twisted into "Troika won't make patches unless Atari pays," making the studio look negligent. Tim couldn't publicly correct the narrative. The same situation later repeated with Vampire: Bloodlines.
9. Lessons Learned
Tim takes full responsibility: every problem stemmed from either a bad decision he made or a contract he read, agreed to, and signed. He identifies two personal takeaways:
- He's a terrible businessman — he should not be allowed to read or sign contracts.
- He's not a strong narrative writer — he can write excellent lore, item descriptions, and quest concepts (he once created 100 quest ideas for Arcanum in a file), but character dialogue and full quest writing aren't his strengths. His Fallout intro narration resonated not because of his writing but because Ron Perlman could read the phone book and make it sound good.
Despite everything, Tim likes the game. He thinks it feels like real D&D, has gorgeous art, and is fun — even if the quests and dialogue aren't great and the combat spaces are too tight. As he and Leonard Boyarsky often discuss: you learn more from failures than successes. Success tells you people liked something but not why; failure makes the reasons painfully clear.
10. References
- Tim Cain. YouTube video. https://www.youtube.com/watch?v=t4718r53zAs