Skip to content

docs(roadmap): add I-28, a sourced guide to building for each bracket - #1296

Open
KaelSensei wants to merge 3 commits into
devfrom
chore/docs-commander-bracket-guide
Open

KaelSensei wants to merge 3 commits into
devfrom
chore/docs-commander-bracket-guide

Conversation

@KaelSensei

Copy link
Copy Markdown
Owner

Why

The owner asked for a roadmap section and a standalone document on how to build a Commander deck for each bracket (1 to 5). Getting that right requires knowing the format, the cards and how games really play out, so this PR is built on a dated research sweep (2026-11-08) rather than a summary of one source.

Wizards publishes the bracket rules and no deckbuilding guidance. EDHREC and the strategy press publish guidance without saying how strong the evidence is. Data sites publish numbers without explanation. No French step-by-step guide exists.

What is in it

File Change
docs/product/guides/commander-deckbuilding-by-bracket.md New, 1,150 lines. The guide. Every claim is labelled Official, Data or Community and cites one of 126 dated sources
docs/product/specs/i28-commander-deckbuilding-guide-by-bracket.md New. I-28 spec in the house format: market, verified code state, requirements, data model, non-goals, tests, 8 slices
docs/product/roadmap.md I-28 section (Next), portfolio entry, dated status note
docs/product/specs/README.md I-28 row
docs/references/game-changers.md, .agents/skills/mtg-learn/references/commander-guide.md Factual corrections (below)
docs/product/competitive-landscape.md New "Online play: adjacent, out of scope" section (MainTable, Playgroup Live, TableCommander, SpellTable); I-20 and I-28 rows

The guide covers

  • The official bracket table as of the 2025-10-21 text, with the 53-card Game Changers list of 2026-02-09, the official French names ("catégorie", "Cartes à impact") and a timeline of changes.
  • Foundations: The Command Zone template, Karsten's land and colour math, the free mulligan, and measured composition by bracket from two datasets (~25k decks). It also covers how games are actually won (combat accounts for 53% of wins in ~400k logged games).
  • One section per bracket with the same sub-headings: intent, official rules, target composition, card choices by role, combos (with Commander Spellbook tags), what to avoid, common mistakes, "wrong bracket if…", reference decks, budget and a pregame line. B5 adds cEDH archetypes, the current EDHTop16 field, mulligans and draws.
  • Moving between brackets (Rachel Weeks' theme → staples → speed → metagame arrows), self-assessment, 12 open debates presented on both sides, and implications for the product.

Sources reached

Official Wizards articles and the Bluesky graphic; EDHREC; CoolStuffInc; Card Kingdom; Draftsim; Commander's Herald; Sheldon Menery's archive; The Command Zone and EDHRECast (13 episodes transcribed locally); Play to Win; EDHTop16; Commander Spellbook; Scryfall; Playgroup.gg; EDHTuner and an Archidekt sample; Karsten; TappedOut; partial Reddit; French creators and the official French pages.

The guide states its limits:

  • Reddit was mostly blocked.
  • YouTube was readable only through podcast audio.
  • Some outlets returned 403.
  • Every dataset relies on self-declared brackets.

What the sweep found in the product

These are recorded as I-28 slices. No code is changed in this PR.

  • The deck wizard shows wrong bracket labels to players, in English and French (deck.json, rendered by DeckWizard.tsx):
    • Bracket 2 is "Casual" with "0–1 game changers". Official: Core, with zero Game Changers.
    • Bracket 3 is "Powered", or "Optimisé" in French. Official: Upgraded / Amélioré.
    • Bracket 4 is "cEDH". Official: Optimized; cEDH is bracket 5.
  • The copilot prompt is wrong. src/lib/ai/deck-plan.ts gives bracket 2 "at most 1 game changer" and forbids tutors at bracket 1.
  • The scorer is stricter than the official rules. src/lib/deck/bracket.ts forces any two-card infinite to bracket 4 under an "(RC rule)". Officially, bracket 3 allows combos that do not come up before turn 6, and the Rules Committee stopped managing the format in 2024.
  • The scorer's thresholds don't match measured decks. The ramp thresholds score a median bracket 2 deck as bracket 4. The land warning fires on the median cEDH deck.
  • The bracket model stops at 4, and bracket 1 is named "Casual" with a tutor rule that was removed in October 2025.

Slice 1 (labels and prompt) is small and should go first, as its own fix/* PR.

Reference corrections

docs/references/game-changers.md:

  • February 2026 added 2 cards (Farewell, Biorhythm), not 5.
  • Gifts Ungiven and Panoptic Mirror were unbanned on 2025-04-22, not in February 2026.
  • October 2025 removed 10 cards, not 11, for the official reasons. Fierce Guardianship was never removed, and Sway of the Stars is legal.
  • April 2025 added 23 cards (18 new, 5 unbanned), not 18.
  • Bracket 5 is now listed as unlimited.

Most of that file's diff is Prettier formatting applied by lint-staged.

The mtg-learn skill's bracket table now shows five brackets with the October 2025 rules.

Scope discipline

Documentation only, in a chore/docs PR because roadmap.md is touched. No initiative status changed and no scope box ticked. changelog.md and progress.md are untouched.

Verification

Run on this branch, rebased on current dev:

  • npx tsc --noEmit: clean
  • pnpm lint: zero warnings
  • TZ=UTC pnpm test: 344 files, 3,510 tests passing
  • prettier --check on every changed file: clean
  • Every relative link in the changed files resolves
  • Pre-push Docker E2E gate: passed (no SKIP_E2E)
  • pnpm sonar was not run, since the diff is Markdown only; SonarCloud CI is advisory

The lint-staged hook prints a cosmetic FAILED … .agents/skills is ignored when a commit touches the tracked mtg-learn reference, because the folder is gitignored even though the file is tracked. The commits are correct.

For review

  • Competitor and creator claims were read from public pages on 2026-11-08; they are not evidence. The guide labels each one, and transcribed podcast quotes are marked as speech recognition.
  • The Recommended composition ranges in the guide are its own synthesis, and each row names the evidence it rests on.

🤖 Generated with Claude Code

KaelSensei and others added 3 commits November 8, 2026 22:42
Wizards publishes the bracket rules but no deckbuilding guidance. EDHREC
and the strategy press publish guidance without evidence labels, data
sites publish numbers without explanation, and no French step-by-step
guide exists. I-28 publishes one guide that separates the three.

The guide (docs/product/guides/commander-deckbuilding-by-bracket.md)
comes from a dated sweep on 2026-11-08: official Wizards texts, EDHREC,
Card Kingdom, CoolStuffInc, Draftsim, Commander's Herald, The Command
Zone and EDHRECast, cEDH resources, EDHTop16, card-level status checked
on Scryfall, Playgroup.gg game logs, two deck-composition datasets and
Karsten's mana math. Every claim is labelled Official, Data or Community
and cites one of 126 dated sources.

The sweep found user-facing errors, recorded as I-28's first slices:

- The deck wizard labels bracket 2 "Casual" with "0-1 game changers",
  bracket 3 "Powered" and bracket 4 "cEDH" (official: Core with zero,
  Upgraded, Optimized; cEDH is bracket 5), in English and French.
- The copilot prompt in deck-plan.ts gives bracket 2 one Game Changer
  and forbids tutors at bracket 1, both outdated.
- The scorer forces any two-card infinite to bracket 4 under a
  Rules Committee rule that no longer applies.

No code changes here; no initiative status changed, no scope ticked.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Checked against the official announcements on 2026-11-08:

- February 2026 added two cards (Farewell, Biorhythm), not five.
- Gifts Ungiven and Panoptic Mirror were unbanned on 2025-04-22.
- October 2025 removed ten cards with the stated official reasons;
  Fierce Guardianship was never removed; Sway of the Stars is legal.
- April 2025 added 23 (18 new, 5 unbanned); the tutor restriction
  went in October 2025.
- Bracket 5 is unlimited, like bracket 4.

The mtg-learn bracket table now shows five brackets with the October
2025 restrictions and turn floors, and no longer cites the banned Mana
Crypt as a Game Changer. Most of the game-changers.md diff is Prettier
formatting applied by lint-staged.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
MainTable (maintable.app) began advertising to French Commander players
in November 2026. It is a free browser table for online play, which the
roadmap keeps out of scope; its deckbuilder overlaps only our non-AI
features, and its owner-approved suggestions on public decks are a
second shipped reference for I-20. Playgroup Live, TableCommander and
SpellTable are recorded alongside it so the boundary is explicit.

No roadmap change follows beyond the I-20 reference and the I-28 row.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
@vercel

vercel Bot commented Nov 8, 2026 •

Copy link
Copy Markdown

The latest updates on your projects. Learn more about Vercel for GitHub.

Project Deployment Actions Updated
magic-ai-builder Ready Ready Preview Nov 8, 2026 10:55pm UTC

@chatgpt-codex-connector

Copy link
Copy Markdown
Codex Review Summary: ❌ Failed

This comment shows the latest Codex review activity on this pull request.

Review Status Commit Review trigger
📝 Code Review ❌ Failed 2026-11-08T22:53:34.366731Z
We couldn’t complete this review. Comment @codex review to try again.
a1b205c PR opened
ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

  • Open a pull request for review
  • Mark a draft as ready
  • Comment "@codex review" or "@codex security review".

Codex reacts with 👀 while any review is running, comments if it has suggestions, and reacts with 👍 once all reviews finish with no findings.

@chatgpt-codex-connector

Copy link
Copy Markdown

You have reached your Codex usage limits for code reviews. You can see your limits in the Codex usage dashboard.

@coderabbitai

coderabbitai Bot commented Nov 8, 2026

Copy link
Copy Markdown

Important

Review skipped

Auto reviews are disabled on base/target branches other than the default branch.

Please check the settings in the CodeRabbit UI or the .coderabbit.yaml file in this repository. To trigger a single review, invoke the @coderabbitai review command.

Run configuration

  • Source mode: Auto
  • Repository: Unavailable
  • Branch: Unavailable
  • File: No configuration file from this repository applied
  • Commit: Unavailable
  • Layers: defaults
  • Review profile: CHILL
  • Plan: Advanced
  • Run ID: a3cdaa52-53dc-4003-8c98-9b93773d4292

You can disable this status message by setting the reviews.review_status to false in the CodeRabbit configuration file.

Use the checkbox below for a quick retry:

  • Trigger review
  • Autofix · Keep fixing CodeRabbit findings and required CI, and resolving merge conflicts

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

Share

@sonarqubecloud

sonarqubecloud Bot commented Nov 8, 2026

Copy link
Copy Markdown

This branch was successfully deployed

1 active deployment
Preview — a1b205c9 Deployed Nov 8, 2026 by vercel[bot]
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant