Frequently asked questions
How is this different from EDHREC?
EDHREC shows you what other people play with your commander, and it does that brilliantly. We cite its numbers as evidence throughout. EDHTuner starts from your actual 99: it reads your cards' rules text to find what your build is missing, names cuts from your own deck with reasons, and playtests every suggested swap in thousands of simulated games. Popularity data can't do those last two things, because they depend on the deck you pasted rather than the average deck.
How is it different from deck power-level raters?
Sites that grade a deck give you a score or a bracket, and a few suggest cuts that move the deck to another bracket. EDHTuner tunes the deck for the bracket you pick. It names specific adds and specific cuts, each with a stated reason and a playtest result, and it gives a land-count verdict measured on your own curve.
Is it a deck analyzer?
Yes, and it goes a step further by telling you what to change. It reads every card's rules text to find your deck's themes, counts your ramp, card draw, removal, board wipes and lands, and checks those counts against the numbers it aims for. Then it names cards to add and cards to cut, with the reason for each one.
Can I use it to upgrade a precon?
Yes. Paste the precon's decklist, or a Moxfield, Archidekt or MTGGoldfish link to it, and pick the bracket you want to play at. The cuts come from the precon's own cards, and each swap pairs a new card with the precon card it would replace.
Can I link straight to my deck?
Yes. Put your deck's link after ?deck=, like
https://edhtuner.com/?deck=https://moxfield.com/decks/your-deck-id. The page
opens with the link filled in, so you only pick a bracket and press Analyze. Add
&b=4 to the end to start on bracket 4. Archidekt and MTGGoldfish links
work the same way.
Can EDHTuner tell me what bracket my deck is?
Yes. Its bracket check reads your deck's Game Changers, mass land denial, extra-turn cards and two-card infinite combos against Wizards' bracket rules, and names the lowest bracket the deck fits. The rules also ask how fast a deck wins and what it's built to do, and that part is for you and your table. You still pick the bracket you play at. When your deck runs more Game Changers than that bracket allows, EDHTuner names the fewest cuts that get you down to the limit. The bracket checker page says how it reads each rule, and bringing a deck down to bracket 3 says how it picks those cuts.
How many lands should a Commander deck run?
Popular Commander decks on Moxfield run a median of 36 lands at brackets 1 and 2, 35 at bracket 3, 33 at bracket 4 and 28 in cEDH. Paste your deck into EDHTuner and it plays the deck at its own land count and at a few other counts on identical shuffles, then gives its verdict on your land count. These are solo games with no opponents. The page on how many lands Commander decks run splits the medians by curve, by ramp and by colors.
Do the playtests actually mean anything?
They're paired solo games, the same shuffles with and without a swap, so the measured difference really is the card. That's a strong test of speed and consistency and an honest zero on everything an opponent brings: removal, politics, blockers. We only let the simulator veto swaps that make your deck meaningfully slower. It never ranks the suggestions. The methodology page spells the whole thing out.
Does it see combos?
Yes. It reads Commander Spellbook's combos of two and three cards and shows the ones your deck holds, each with its cards and a link to its page on Spellbook. It never offers a combo piece as an ordinary cut, and a trim down to 100 takes them last. How it wins names the combos that win the game by themselves, and the bracket check counts two-card infinite combos against Wizards' bracket rules. At bracket 4 and up it can suggest the missing piece of a winning combo. Below bracket 4 it never suggests a card that would complete an infinite combo. Cast-loop engines still read smaller than they play in the goldfish games.
What do the brackets and "GC" mean?
The five Commander brackets from the official bracket system, with Game Changers (Wizards' official set of outsized cards) counting toward your bracket's cap. Over the cap, we name the minimum trims and their stand-ins. Under it, the deck panel just shows your count against the cap.
What leaves my browser?
The analysis itself runs entirely on your machine. No part of the engine runs on a server.
What a completed run sends to our own feedback store is the commander name, deck count, bracket, the price cap and the names in your never-offer box if you set them, app version, the deck you analyzed, and the analysis it produced: the swaps, the cuts, and the reasoning behind them. We read those to check what we're actually recommending to people, which is the only way to tell whether the suggestions are any good. Keeping the finished analysis means we can look at a run exactly as you saw it, instead of guessing from card names alone. If you analyze a deck exactly as it came in from a Moxfield, Archidekt or MTGGoldfish link, the run also sends that link, and so does an issue report about it.
The price cap and the cards you chose never to be offered are remembered in your browser so they are there next time. They are not part of a share link.
Each page load also sends how long the app took to start. Everything the page sends carries where your visit came from: the name of the site whose link brought you here, such as google.com, or the tag that link carried. If your visit started on another page of this site, like a commander's deck numbers, it also carries that page's address. When a link on one of this site's own pages opened the app, it carries a short name for that link, or for the kind of page it sat on, like the button on a commander page. The address of the page you were reading on the other site is not saved.
Clicking a Buy button sends which store and which card you clicked, along with the commander and bracket of the deck you were tuning. That tells us whether anyone uses those links.
If we can't read the deck you pasted, the error report carries back the first few lines we couldn't read, whatever those lines happen to hold. A broken paste is impossible to fix without seeing what broke it.
If the app can't start, the error report carries the error message it hit. When something on your network blocked part of the app, that message can include the block page's title and the server name it reported.
Voting on a swap, sharing a link, or sending an issue report sends the deck along with what you're reporting on.
All of these records expire on their own. Votes, reports, failed pastes and failed starts are deleted after 90 days, start timings after 30 days, and analyses and shared decks after a year.
Nothing is sold or shared. None of it is tied to a name, an email address or an account, because we don't have any.
Site traffic is measured with Cloudflare Web Analytics, which uses no cookies.
Why did it suggest cutting a card my deck is built around?
Sometimes the text read misses a mechanical relationship. That's why every card has a lock: hover a cut, click the padlock, and we never suggest cutting that card again. If a suggestion looked wrong, use "report it" on the page. Real reports are how this gets better, and past reports have turned into permanent fixes within days.
Is the card pricing an ad?
Buy links go to ManaPool, an affiliate link that may earn us a commission, plus TCGplayer and Scryfall with no affiliation. We compute the suggestions before any price is looked up, and no affiliate relationship influences them. You'd get the identical result with the links removed.
Is it free? Do I need an account?
It's free and there's no account to make. There are no ads either, and you can run as many decks as you like. The analysis runs in your own browser, so there's nothing to ration.
Any analysis can be copied as a link that replays the same result for anyone who opens it.