Guide

Commander brackets explained (1–5)

Commander brackets are Wizards of the Coast's five-tier scale for matching decks of similar power before a game: 1 Exhibition, 2 Core, 3 Upgraded, 4 Optimized and 5 cEDH. A deck's bracket comes from what it is built to do, checked against a few restrictions: how many Game Changers it plays, and whether it runs mass land denial, extra turns or two-card combos. Brackets 1 and 2 allow no Game Changers, Bracket 3 allows up to three, and Brackets 4 and 5 have no restriction beyond the banned list. Below: each bracket, how the rules have changed since February 2025, and how Lotus Vault's Analyse estimates one.

This guide was written entirely by AI, from Lotus Vault's code and the official sources listed at the end.

What the brackets are, and why they exist

Commander is a social format, and four players rarely share one idea of a good game. The brackets give them a common vocabulary — five named tiers with concrete restrictions — to compare decks before shuffling up rather than after someone has been run over. Wizards calls the system “an entirely optional way to help matchmake”: a tool for the pre-game conversation, not a rule.

It came with a change in who runs the format. On September 30, 2024, the Commander Rules Committee handed day-to-day management to Wizards' design team, its members staying on as advisers; on October 22, 2024, Wizards formed the Commander Format Panel, 17 advisers ranging from former committee members to content creators and cEDH specialists. The panel advises; Wizards decides. As Gavin Verhey put it: “Ultimately, myself and other designers at Wizards are going to make the final calls.”

The five brackets launched on February 11, 2025, together with a first Game Changers list of 40 cards. The system is still labelled a beta: the latest official update, on February 9, 2026, changed no bracket rules, and no later update had been published when this guide was last checked, on September 26, 2026.

The five brackets at a glance

Wizards splits the tiers in two: “Brackets 1, 2, and 3 are different levels of socially focused play. Brackets 4 and 5 are focused on a higher power or even a competitive experience.” The expected game lengths, added in October 2025, describe a typical game at that table, not a test a deck must pass.

The five Commander brackets — Wizards of the Coast's rules as of February 9, 2026
Bracket	Intent	Restrictions
1 · Exhibition	The lowest-powered tier of socially focused play; games expected to last 9 turns or more.	No Game Changers · no mass land denial · no extra-turn cards · no intentional two-card infinite combos
2 · Core	Socially focused play; games expected to last 8 turns or more. No longer defined by precon power since October 2025.	No Game Changers · no mass land denial · extra turns only in low quantity, not chained · no intentional two-card infinite combos
3 · Upgraded	The strongest tier of socially focused play; games expected to last 6 turns or more.	Up to 3 Game Changers · no mass land denial · extra turns only in low quantity, not chained · no intentional early-game two-card infinite combos (late-game ones are allowed)
4 · Optimized	High power; games expected to last 4 turns or more.	No restrictions other than the banned list
5 · cEDH	Competitive, tournament-minded play; a game can end on any turn.	No restrictions other than the banned list — intent, not a card rule, separates it from Bracket 4

The four restrictions, one by one

Every bracket rule comes down to four questions about a decklist — all about what a deck must not play, never what it must play to reach a bracket.

Game Changers. Cards that, in Wizards' definition, “can dramatically warp Commander games, allowing players to run away with resources, shift games in ways that many players dislike, block people from play, efficiently search for their strongest cards, or have commanders that tend to take away from more casual games.” None in Brackets 1 and 2, up to three in Bracket 3, any number in 4 and 5. The list held 53 cards in September 2026 and has been revised three times since launch; the Game Changers guide lists them all.
Mass land denial. Denying lands to every player at once — destroying, exiling or bouncing them en masse, or stopping them from untapping, as Armageddon or Winter Orb do. Not permitted in Brackets 1 to 3; permitted in 4 and 5.
Extra turns. None in Bracket 1. In Brackets 2 and 3, only in low quantity and not meant to be chained. Unrestricted in 4 and 5.
Two-card combos. No intentional two-card infinite combos in Brackets 1 and 2. In Bracket 3, none early in the game, though two cards that come together late are allowed. Anything goes in 4 and 5. Wizards sets no turn number for “early”.

Tutors used to be a fifth question — the launch text wanted them “sparse” in the lowest brackets — until Wizards dropped that guidance on October 21, 2025. The Game Changers list now restricts the most efficient ones: Demonic Tutor counts toward Bracket 3's three Game Changers, an ordinary tutor toward no limit at all.

How the rules changed, February 2025 to February 2026

The bracket rules themselves have changed only once; the Game Changers list, more often.

February 11, 2025 — The five-bracket beta launches, with a first list of 40 Game Changers.
April 22, 2025 — No change to the bracket rules. Game Changers: 2 removed, 18 added (Necropotence and Orcish Bowmasters among them), and 5 more added as they came off the banned list. The headline message: “Intent is the most important part of the bracket system.”
October 21, 2025 — The only change to the bracket rules so far. Tutor guidance is removed entirely; each bracket gets an expected game length; Core stops being defined by the power of preconstructed decks, which had grown too uneven to serve as a yardstick. Game Changers: 10 removed, from expensive spells such as Expropriate to legends valued mostly for what they do from the command zone, such as Yuriko, the Tiger's Shadow.
December 11, 2025 — The Commander Format Panel rotates members: six leave, two join. No format changes.
February 9, 2026 — No change to the bracket rules; the panel will “cool it on bracket-level changes for the time being.” Farewell and Biorhythm, the latter on its unbanning, join the Game Changers, bringing the list to 53. Lutri, the Spellchaser, unbanned the same day, stays off it: “Lutri was not banned because of its inherent power.”

The October 2025 update also floated open questions — Rhystic Study, Thassa's Oracle at casual tables, a sixth bracket, hybrid mana in colour identity — none of which has changed the rules; hybrid mana was shelved in February 2026. Check the date on any bracket advice, this guide's included.

Choosing a bracket: intent before checklist

The restrictions give the lowest bracket a deck may enter, not the one where it belongs. Wizards' April 2025 update put intent first, asking players to judge their decks honestly rather than tally them mechanically: a deck that plays stronger than its list suggests should move up. That cuts both ways — a deck with no Game Changers can outclass a Bracket 2 table if every card is tuned, and one with three can sit comfortably in Bracket 3 if it plays fair.

In practice, a short conversation before the game does most of the work:

Which bracket would you call your deck, and on which turn does it usually threaten to win?
Which Game Changers does it play, if any?
Can it win with two cards, and how early can they come together?
Does it lock lands, chain extra turns, or do anything the table would rather not face?

Since Brackets 4 and 5 differ only by intent, a strong deck that is not built for competitive play usually belongs in Bracket 4, however powerful its cards.

How Lotus Vault's Analyse estimates a bracket

Analyse reads a Moxfield or Archidekt link, or one of your saved decks if you have an account, and returns a bracket from 2 to 5, a + or − within it, and its reasons, naming the cards behind each Game Changer, land-denial or extra-turn flag. It never returns Bracket 1: Exhibition is a table agreement, not a property of the cards. Game Changers come from the official list, refreshed from Scryfall at most weekly; mass land denial, extra turns and nonland tutors from each card's rules text; combos from Commander Spellbook. Then, in order:

Nothing restricted. No Game Changer, mass land denial, extra turn or two-card combo, and at most two tutors (a house threshold): the deck is Core or Upgraded, and the tuning test below decides which.
Within Bracket 3. Up to three Game Changers, no mass land denial and no early two-card combo — Wizards' rule as written — plus two house caps: at most two extra-turn cards and four tutors.
Competitive shell. Seven or more Game Changers, an early two-card combo, and at least four fast-mana cards and five tutors: Bracket 5. Another house rule, since Wizards defines cEDH by intent, not by a card count.
Everything else. Bracket 4.

“Early” is Lotus Vault's reading too, since Wizards defines none: a combo Commander Spellbook tags Spicy or Ruthless in its own vocabulary. When Spellbook cannot be reached, a curated list of 35 well-known two-card combos stands in, a pair counting as early when its two cards cost six mana or less together.

The line between Core and Upgraded needs the most explaining. Wizards' rules say what a Bracket 3 deck must not play, not how sharp it is: without a test, every clean deck would fall to Bracket 2, and a precon would look no different from a deck built to beat it. Lotus Vault's house test counts five tuning signals, and a deck needs three of the five to read as Upgraded:

an average mana value of 3.4 or lower across its nonland cards;
9 or more pieces of interaction — removal, counterspells and board wipes;
12 or more card-draw effects;
a deck-total EDHREC salt score of 25 or more (see the salt score guide);
2 or more fast-mana cards and tutors, counted together.

These thresholds are judgement, not measurement: there is no labelled corpus of decks to fit them against. On real decklists run through the test, a tuned build fires all five signals (Bracket 3), a recent precon three (3−), and an older precon or a jank pile one (Bracket 2). Nothing detects precons as such; they are analysed like any other deck.

Then Commander Spellbook rates the deck in its own seven-tag vocabulary, which Lotus Vault maps onto the brackets: Powerful to 3, Spicy to 4, Ruthless to 5. If that rating is higher, it wins, because Spellbook knows combo speed better. It never lowers an estimate, and its Banned tag, a legality problem rather than a power statement, sets no floor.

Last comes the + or −, which places a deck within the bracket reached so far, Spellbook's floor included. Each level counts its own pressure signals, drawn from tutors, fast mana, combos, Game Changers, extra turns and the power score: three or more earn a +, close to the next bracket, and a low-end deck gets a −. A deck that saturates essentially its whole budget (four signals in Bracket 2, five in 3 or 4) is promoted and marked −, so it reads “3−” rather than a misleadingly soft “2+”. The About page's analysis section explains these choices in the first person.

What no estimator can tell you

Any bracket calculator, Lotus Vault's included, reads cards, not intent — and intent is what Wizards says matters most. Use the estimate to open the pre-game conversation, not to close it.

Detection reads rules text. Tutors, extra turns and mass land denial are found through text patterns (plus, for land denial, a short list of named cards); an unusual wording can slip through or be miscounted.
Combos depend on Commander Spellbook. Without it, only the 35 curated pairs are checked.
The house rules are judgement. The tutor and extra-turn caps, the tuning signals and the Bracket 5 floor are Lotus Vault's numbers, not Wizards', and they are meant to move when real decks disagree with them.
The rules move. The Game Changers list changes every few months, and the system is still a beta.

Bracket and power level are different readings, too: Analyse's power score out of 10 weighs combos, interaction and card draw most, tops out at 7/10 by design without a combo or a Game Changer, and can put a tuned Bracket 3 deck around 6. Read both before you sit down.

Paste a Moxfield or Archidekt link into Analyse to see the deck's estimated bracket, its + or −, and the cards behind the verdict — or add a second deck to compare the two side by side.

Open Analyse

Related guides

Sources

Last updated: September 26, 2026