Commander Finder
What commander
should you build?
For your first deck or your next one. There are two ways in. Each takes about a minute, needs no account, and ends with five commanders and the reason each one suits you.
Every legal commander is eligible: 4,376 of them.
Question 1 of 7
Build around
What are you building around?
Add up to 15 cards you own or love. We'll find commanders that can legally hold all of them.
Picking a first Commander deck
The advice a new player gets is usually a list of somebody else's favorite commanders. That advice answers a different question from the one that matters most, which is what you want a game of Magic to feel like. A leaderboard can't tell you that.
This tool asks about the feeling and works backwards to the colors, because that is the direction the decision actually runs. Nobody starts out wanting Golgari; they start out wanting things to come back from the graveyard, and Golgari is where that lives. If you already know your colors, the last question takes them and the rest still does the work.
The question everything turns on
How much bookkeeping you want is the question most commander quizzes skip, and it is the one that decides whether the deck gets finished. A commander with four triggers and a board state to track is a good card and a bad first deck, and being handed one is why a lot of people bounce off the format. It is a hard filter here. Answer "minimal" and a complicated commander cannot appear in your results at all, however well it matches everything else.
Color identity, and why your pile of cards matters
A card is legal in your deck only if every colored mana symbol on it also appears on your commander. So the cards you already own set a floor: your commander's color identity has to cover the union of all of them. One off-color card can be the difference between a clean two-color deck and a four-color mana base you will fight for a year.
That is what the second door is for. Add the cards, watch the required colors appear, and if they spread across too many colors, the tool names the single card to drop and what dropping it would open up, instead of returning worse answers without saying why.
Where the numbers come from
Every commander in Magic is eligible here, scored on what its own rules text says. Where a commander has been played enough in TurnZero to have a meaningful record, the result card shows its win rate across tracked games. That takes at least 40 games, because below that the number is mostly noise. Short of the threshold, the card shows how many tracked decks run it instead. Most commanders have no tracked play at all; that is normal and counts against nothing. Around 584 have some.
13 of them also carry a hand-written line on what a turn looks like. The rest get a plainer description generated from what the card does. Those descriptions are kept literal on purpose, so nobody mistakes a generated sentence for a hand-checked one.
Common questions
Is this free?
Yes, and there's no account. The quiz, the card-first search, the results and the share image all work signed out. Building the deck in the app is where an account comes in, and that's a separate step you can ignore.
How many commanders can it recommend?
Every commander that legally exists: 4,376 of them at the last build, from Magic's whole card catalog. A hand-picked shortlist of favorites would answer "which of ours suits you" rather than the question you asked, and it would go stale the day a new set released.
Where do the recommendations come from?
Each commander is scored on six things read off the card itself: what it wants the deck to do, how it survives removal, how much bookkeeping it asks for, how the table reads you, what it costs, and how many colors it needs. Complexity is a hard filter rather than a preference. If you ask for minimal bookkeeping, you will never be shown a commander that needs more.
Why do some results show a win rate and others don't?
A win rate is only printed once a commander has at least 40 completed games tracked in TurnZero. Below that the sample is too small to mean anything, so the card shows how many tracked decks run it instead, which is a real number and harder to over-read. Most commanders have no tracked play at all, which is normal and doesn't count against them.
I already have some cards. Can it use those?
Yes, that's what the second door is for. Add up to 15 cards and it finds commanders whose color identity can legally hold them, ranked by how many they cover. It also names the ones that don't fit and why, which is the fastest way to learn the color identity rule.
Can I just pick a playstyle, like mill or spellslinger?
Yes. The first question is written in plain language because most people starting out don't know the jargon yet, but "Or pick a style by name" underneath it opens all 20 named styles (mill, blink, aristocrats, stax and the rest), each with a one-line explanation of what it does. Every one links to a full guide, so you can read up on a style before committing a deck to it.
It asked about brackets. Does it know what bracket a commander is?
No, and nothing does. A bracket describes the deck you build, not the commander. Almost any commander can be built at Bracket 2, and a great many can be built at Bracket 4; what changes is the ninety-nine cards, which this tool doesn't pick. The answer is used as a nudge: aiming casual leans toward commanders that come as precons and cost little to start, aiming optimized leans toward ones with a real tracked record. Both questions are optional and skipping either costs you nothing.
What is color identity and why does it matter?
A card can only go in your deck if every colored mana symbol on it also appears on your commander. So a green card needs a commander with green in its identity. It's the rule that catches out every new Commander player, and it's why adding a single off-color card to your pile can wipe out most of your options.
Does it tell me what to put in the deck?
Not yet. This finds the commander. Building the other 99 is what the Deck Lab in the TurnZero app does, and 'Build this deck' on any result is the way through to it.