Crafting an Arena Cube: A Dev Diary

The journey to bringing a Powered Cube to MTG Arena began during an internal innovation sprint known as an “Experimental Frenzy,” where developers explore new ideas. In April 2023, one engineer decided to investigate the feasibility of implementing a high-powered cube environment, starting with an analysis of Magic Online’s Vintage Cube as a reference point. This initial review helped identify which cards were already available on Arena, which could be added with minimal effort, and which presented significant implementation hurdles.

During these quarterly hackathons, team members step away from their usual tasks to experiment freely. Some projects focus on backend improvements or quality-of-life features, while others lead to visible additions like achievements or new draft formats. For the cube initiative, the first step involved cataloging hundreds of cards and assessing the engineering effort required to make them function within Arena’s rules engine.

The engineer’s initial breakdown revealed that around 215 cards were already present on the platform. Another 200 could be added with almost no additional rules coding, while roughly 20 needed a standard amount of work. A small group awaited mechanics like delve or suspend, which were expected to arrive soon. The remainder seemed unlikely to make the cut for various reasons. After this analysis, the engineer began prototyping cards like Fastbond and Dack Fayden, then secured approval to dedicate ongoing time to the project.

Implementing a card involves far more than just rules coding. A producer handles database organization and version selection, while the Art team prepares card frames and assets, sometimes creating new components from scratch. The Design team determines how players interact with card abilities, such as choosing between sacrifice or discard options. They also set “smart stops” for priority and confirmation prompts. The Duel Scene team ensures everything displays correctly during gameplay, and Localization translates all text into supported languages. The Content team manages details like wildcard crafting and format legality. Occasionally, other specialists contribute; for example, Grist, the Hunger Tide required collaboration with the deck-building team, and planeswalkers need voice-over work. Quality assurance tests every aspect before a card goes live.

While many cards parse automatically using Arena’s system, including complex ones like Emrakul, the Aeons Torn or Recurring Nightmare, others demand extensive manual coding. The rules team’s work is highly unpredictable; some cards take days, while others require weeks or months. By tackling the most challenging cards early, the team could schedule remaining work confidently and avoid last-minute exclusions of staples like Balance or Animate Dead.

Over eighteen months of part-time effort, several cards stood out for their complexity. Necromancy proved exceptionally difficult, combining flash timing permissions with a reanimation ability full of edge cases. The engineer manually wrote code for its rules, bypassing the usual automated process. Grist, the Hunger Tide revealed a subtle bug where library shuffling reset continuous effects, a issue that might have affected older cards like Sphinx of the Guildpact. Other bugs emerged during testing: Silvergill Adept incorrectly prompted for additional costs when flashed, Kytheon failed to count attackers that left the battlefield, and Urza’s Saga stayed in play if mana was tapped in response to its chapter trigger. Mox Diamond required multiple fixes for unusual interactions with tokens and face-down permanents.

One card, Booster Tutor, inspired a creative solution. Instead of literal booster packs, the team adapted Arena’s spellbook drafting mechanic. Initially, the engineer proposed using a 540-card spellbook, but colleagues pointed out practical issues like maintenance and visibility. The final implementation uses a pre-randomized, collated list for each match, ensuring players see fresh cards with each cast and maintaining cube pack balance. The Duel and Art teams later enhanced the visual presentation.

Several expected cube cards didn’t make the final list. Corpse Dance and Shallow Grave were excluded because they require graveyard order tracking, which would involve intrusive dialogs for players whenever multiple creatures entered the graveyard simultaneously. Since modern Magic design avoids graveyard order matters, the team decided the development cost outweighed the benefits.

Through the Breach sparked extensive debate due to its splice onto Arcane ability. While the splice effect is irrelevant in cube, there are no other Arcane spells, fully implementing it would demand significant resources across multiple teams. Options included releasing a modified version or using an Alchemy substitute, but the team ultimately chose to preserve paper accuracy and omit it for now, leaving Sneak Attack as the sole effect of its kind.

Playtesting uncovered unexpected issues, such as bots drafting multiple copies of Power Nine cards and deck validation falsely flagging cube cards as illegal. One memorable bug caused all power cards to disappear during deck construction because of a safeguard against uncollectible cards. These were resolved through iterative fixes and event tuning.

After years of development, the Powered Cube represents a milestone for the team. For its creators, the reward is the chance to finally draft their favorite format on Arena, even if it means piloting a five-color midrange deck to a modest 3-3 record.

(Source: Magic Wizards)