More information
Last updated: October 25, 2026
TLDR: every figure in the Regional Metagame window is counted from the public results of the tournaments on TopDeck.gg that EDHGlobe shows on the globe. Meta share is the percentage of decks (with a listed commander, from commanders that pass the Min. entries filter) that played a given commander, and conversion is the percentage of those decks that made the top cut, only counting tournaments that had one. Nothing is estimated or predicted, and nothing is added by hand (the smoothed trend lines in the two weekly charts are the only fitted values).
Tournaments per week
How many of the tournaments we count were played each week. It follows the same rules as the globe: the cEDH name filter is applied and only the players filter below changes what is included. The current week is left out until it ends. The orange line is a smoothed trend: it is fitted to the weeks around each point, so it ignores the ups and downs from one week to the next and shows how many tournaments a typical week has at that time, and whether that is rising or falling. It describes the past, it isn't a forecast.
Tournaments per week Trend
Unique players per week
How many different players took part in at least one of those tournaments each week. A player who plays several tournaments in the same week is counted once, so this is people, not entries. Players are told apart by their TopDeck player id. It uses the same tournaments as the chart above, so the Players filter above applies here too, and the current week is left out until it ends. The orange line is the same kind of smoothed trend.
Unique players per week Trend
Where the data comes from
- Tournaments and standings come from the TopDeck.gg API: name, date, location, number of players, size of the top cut, and for each player the final standing and the commander(s) listed in their decklist.
- Card art and color identity come from Scryfall.
- Country borders come from the open Natural Earth dataset.
- The data is refreshed daily at 05:00 AM UTC. The exact time of the last update is shown in the footer of the globe page.
Why commander data starts in June 2025
For tournaments before 1 June 2025, TopDeck mostly doesn't give us the commander of a deck. For those events, players submitted a link to a deck on another site (most of them on Moxfield) or nothing at all, and the API doesn't include the cards. From that date TopDeck reads the lists itself, so the commander is available. That is why the charts of a commander have little or no data before then, and why tournaments from that period show few or no decks in their standings or metagame.
Players who submitted no list at all, about half of them in every period, can't be recovered. For the links, we're trying to get access to Moxfield's data so that the commanders of older tournaments can be filled in. Until that works, the numbers before June 2025 should be read with care: they are based on a small and not necessarily representative part of the decks.
The charts of the commander view mark this date, and the ban list update of 23 September 2024 (Jeweled Lotus, Dockside Extortionist, Mana Crypt and Nadu, Winged Wisdom), with faint dashed vertical lines, because both change what a trend before and after them means.
Which tournaments are counted
The Regional Metagame uses exactly the tournaments that pass the filters of the window, which start from the ones you picked on the globe:
- cEDH only. TopDeck has no field that says whether an event is cEDH, so EDHGlobe hides tournaments by name: names mentioning things like Casual, Budget, Precon, Pauper or Bracket 1–4 are left out, unless the name also says cEDH or Bracket 5. This is a best effort and an event can occasionally be wrongly kept or hidden.
- Number of players (all, 30+, 50+, 100+): tournaments below the chosen size are ignored.
- Period (for example the last 6 months): counted back from today, with every month taken as 30.44 days.
- Tournaments for which TopDeck has no standings yet are skipped.
- Min. entries (All, 20, 60, 120) isn't about tournaments but about commanders, so it is explained below.
Decks and commanders
- Every player in the standings counts as one deck, represented by the commander(s) in their decklist.
- Partners count as one deck. A pair such as Kraum + Tymna is a single entry, and the order of the names doesn't matter. A commander played alone and the same commander paired with a partner are different entries.
- Players without a listed commander (no decklist, or one TopDeck could not read) can't be assigned to any deck. They are not counted in any percentage, and the window tells you how many there were. In practice this is a large part of all players, so keep it in mind when reading the shares.
- A commander is identified by its name, so a spelling variant in the decklist can show up as a separate entry.
Min. entries
This filter hides the commanders that were played fewer times than the chosen number (All, 20, 60 or 120) in the selected period and size, which helps leave out rarely played or casual decks that show up at cEDH tournaments. It starts at 20 every time the window is opened, and All turns it off.
- The count is made over the whole world, not inside the selected region. A commander with 20 or more decks worldwide stays in every region, even where it was played once, and one with fewer than 20 is hidden everywhere. This way a region and the one it is compared with always use the same list of commanders.
- The percentages are recalculated with what is left: the decks of the hidden commanders are removed from the total, so meta share, the average conversion and the deck counts in the window only describe the commanders that pass the filter. The window says how many decks were left out.
- A commander's own conversion rate doesn't change, because it only depends on its own decks.
Meta share
The meta share of a commander in a region is:
decks of that commander ÷ all decks with a listed commander that pass the Min. entries filter, both counted in that region.
- The color filter only hides commanders. Percentages are always a share of all the decks in the region, never of the selected colors only, so they don't change when you filter.
- The commanders are sorted by decks played by default, and every commander is listed, with no “Other” group.
Conversion rate
The conversion rate of a commander is:
decks that made the top cut ÷ decks played, only counting tournaments that had a top cut.
- A deck made the top cut when its player finished within the first N places of the standings, N being the size of the top cut that TopDeck reports for that tournament.
- Tournaments without a top cut are left out of both sides of the division. Those events, usually small ones, can't tell us anything about conversion, so counting their decks as “not converted” would lower every rate for something that isn't the deck's fault. The row shows the exact figures, for example “31 of 118”.
- The average conversion in the window's header is the same calculation over every deck of the region.
- Small samples. With few decks the rate says little (1 top cut in 2 decks isn't 50%). Commanders with fewer than 10 decks in tournaments with a top cut are marked small sample, shown dimmed, and listed after all the others when sorting by conversion.
Average draw rate
Shown at the end of the stats under the sort menu. It is the share of games that ended in a draw, averaged over tournaments:
- For each tournament,
games that were draws ÷ games played, taken from the win-loss-draw records of all its players. A drawn game adds one draw to every player at the pod, so the share of draws among all recorded results is the share of drawn games. - The region's figure is the plain average of its tournaments, so every tournament weighs the same whatever its size. Tournaments with no games recorded are skipped.
- When comparing, the other region's rate is shown next to it, with the gap in pp.
- The tournament window shows the draw rate of that single tournament, calculated the same way.
- It describes tournaments, so the Min. entries and color filters don't change it. The period, players and region do.
Seat win rate
Only the tournament window shows it, next to the draw rate. It is shown as S1 WR to S4 WR (win rate of seat 1 to 4) and says how often each seat of the pod won in that tournament, so you can see whether going first (1st) or last (4th) was an advantage there.
- It comes from the pairings TopDeck publishes for each round, which list the players of every table in seat order and say who won.
- Each number is
games won from that seat ÷ games played, counting every round (Swiss and top cut) and only completed four-player pods. Byes and pods of other sizes are left out. - A drawn game counts as played but is won from no seat, so the four percentages plus the draw rate add up to 100%.
- One tournament is a small sample (a few hundred games at most), so differences of a few points are normal luck. Don't read much into a single event.
- Tournaments loaded before this feature existed, or without round data, simply don't show the line.
The commander view
Clicking a commander in the Regional Metagame list opens its own view, with a banner, a chart over time and its numbers. It uses the same filters as the list (region, compare with, players, min. entries and period), so changing any of them updates it, and All commanders (or Esc) takes you back.
- The chart shows either the commander's meta share or its conversion rate for each period of time, for the region (turquoise line) and, if you pick one in Compare with, a second region (amber line). By default it is World with no comparison. Hover or touch the chart for the exact values and the number of decks behind each point. The same data is available as a table under the chart.
- Time periods: Group by chooses weeks or months; Auto uses weeks for periods of about four months or less and months for longer ones. A period with no tournaments in the region (or, for conversion, none with a top cut) is a gap in the line, not a zero.
- Meta share in each period is calculated like in the list: the commander's decks divided by all decks with a listed commander in that region and period, after the Min. entries filter. The cards under the chart give the figure for the whole period.
- Conversion uses the same rule as the list (only tournaments with a top cut, small sample under 10 decks), so a single week can swing a lot.
- Decks played is the number of decks of that commander in the period, and how many of the tournaments it appeared in.
- Top cut conversion is the commander's share of the top cut:
its decks in the top cut ÷ all decks (with a listed commander, after Min. entries) in the top cut, over the same tournaments. Compare it with the meta share: a higher figure means the commander reaches the top cut more often than its share of the field would suggest. Like conversion, it is dimmed as a small sample under 10 decks. - Seat win rate is the share of games that this commander won from each seat:
games won from that seat ÷ games it played from that seat, in completed four-player pods. It is built from the same pairings as the tournament window, assigning each game to the commander the player at that seat was using. A fair table would be a little under 25% for every seat, since draws are won by nobody. Counts are small for rare commanders, so check the number of games next to each rate. - Top Pilots lists the players who won a tournament or reached its top cut with this commander, inside the same filters. They are ranked by tournament wins, then top cuts, then best finish. Each pilot shows every tournament they played with the commander (not only the good ones), and each tournament links to the decklist they played there; the pilot's name opens their TopDeck.gg profile. Players are told apart by their TopDeck.gg id, so one person who changes their display name is still one pilot.
- Region performance ranks the countries by this commander's conversion, best first. For each country it shows the conversion (
top cut decks ÷ decks, only in tournaments with a top cut), in brackets how far that is from the commander's conversion in the whole world, and its meta share in that country. Countries with fewer than 10 decks in tournaments with a top cut are listed last and dimmed as small sample. Tournaments that can't be placed in a country are left out of the countries but count in the world figure. It follows the Players, Period and Min. entries filters; the region selector doesn't apply, because the tab compares all countries. - Seat data exists only for tournaments loaded after it was added to the site. Until all the data is rebuilt the seat table can be empty or based on fewer games than the other figures.
Commander view tabs
- Commander data has the chart and numbers described above.
- Tournaments lists every deck of that commander in the tournaments that pass the filters (region, players and period). Each row is one player's finish: place, tournament, date, location and size, player and record. They are sorted by final place, with the winners first; among equal places the tournament with more players goes first (then the most recent). Clicking a row opens that player's decklist on TopDeck.gg in a new tab. The place is the player's position in the standings TopDeck publishes.
- Spicy cards shows up to 5 cards that few decks of this commander play but that convert a lot better than the commander's average. See below.
Spicy cards
Card lists come from TopDeck, which only shows them once a tournament has ended or the organizer made decks public, so decks without a list are ignored here. Basic lands are not counted. A card is spicy for a commander when it is:
- Rarely played: in at most a quarter of the commander's decks with a list (and at most a third of the average card's play rate), and in at least 3 decks.
- Converting well: among its decks from tournaments with a top cut (at least 5), the conversion is at least 5 pp above the commander's own, and still above it after allowing for a small sample (the lower end of a 95% Wilson interval). That is why 7 of 7 can rank above 3 of 5.
Cards are ranked by how far that lower end is above the commander's conversion. Conversion (reaching the top cut) is used instead of win rate because it is what the rest of the site measures and is less noisy than game records. The view needs at least 20 decks with a list, and may show fewer than 5 cards, or none, when nothing qualifies. The pilots shown are the ones who played the card, with their best finish. Card images are loaded from Scryfall when you open the tab.
Only recent play counts here. Card lists exist only for tournaments from about June 2025, when TopDeck started reading decklists itself. Decks from older tournaments, even those whose commander we can fill in from a link, have no card list, so Spicy cards leaves them out. That is on purpose: what people play changes a lot with new sets and bans, and comparing a card against lists from a long time ago wouldn't be fair, so this tab describes recent play only. The period and region filters of the window still apply on top of that.
The metagame of one tournament
The View metagame button in the tournament window swaps the standings for the commanders of that tournament. For each one:
- Entries are its decks in the tournament, and the percentage next to it is their share of the entries that list a commander. Partners are one commander, whatever order they are listed in.
- Conversion is the decks that finished inside the top cut ÷ its decks. A tournament without a top cut shows a dash.
- Draw rate is the drawn games ÷ games played by that commander's decks, from their win-loss-draw records. With only a few decks it can jump a lot.
- Ordered by entries, then conversion.
Regions
- A tournament belongs to the country whose borders contain the location TopDeck lists for it. It is the location of the event, not where the players live. If the point falls just outside every border (a coast, a small island), it is assigned to the nearest country if that is within 50 km.
- Tournaments that can't be placed in a country count for World only.
- Each country belongs to a continent, according to the same borders dataset. Nothing is merged below that: Catalonia is part of Spain, for example, and regions inside a country aren't separated.
Colors
- The color of a commander is its color identity on Scryfall. For a partner pair it is the union of both commanders.
- The filter matches the exact identity: picking W and U shows only white-blue commanders, not every commander that contains white. With only C selected you get the colorless ones. No color selected shows everything.
- The five-color button (the colour wheel) is a shortcut for picking all of W, U, B, R and G at once. It and the colorless button are exclusive: choosing either clears the other colors, and choosing any single color clears them.
- Commanders whose colors couldn't be looked up are hidden while a color filter is active, and the window says how many decks that affects.
Comparing regions
- When you compare a region with another one (or with the world), each row also shows the other region's meta share and conversion for the same commander.
- The difference is given in pp, percentage points: the plain subtraction of the two percentages. 7.1% against 5.1% is +2.0 pp, not +40%.
- Biggest differences sorts by the size of that gap, including commanders that only one of the regions plays. To keep one lucky deck from topping the list, commanders with fewer than 3 decks (counting the decks the other region's share would predict) are left out.
Why the numbers can differ from other sites
If you compare EDHGlobe with a site such as edhtop16, the numbers can legitimately differ even when both read the same TopDeck data. The usual reasons are:
- Meta share tends to be higher here. We divide by the decks that have a listed commander only, and many players have none. On top of that, the Min. entries filter removes the rarest commanders from the total, which makes the others' shares grow. A site that divides by every entry will show lower percentages for the very same decks.
- Conversion tends to be higher here. We ignore tournaments without a top cut. A site that counts every entry as “played”, even in events with no top cut, will show lower rates.
- Different sets of tournaments. Each site decides what counts as a cEDH event, which sizes and which dates are included, and how long a “month” is. Our name filter hides some events that another site may keep, and the reverse.
- Different update times. A site that updates more often will already include tournaments we'll only add on our next daily refresh.
- Different grouping. Partner pairs, spelling variants and double-faced cards can be grouped differently, and some sites list a pair under each of its commanders.
- Small samples. Percentages from a handful of decks move a lot with one result, in any site.
Because of this, compare trends and rankings across sites rather than exact percentages, and make sure the same period, size and definitions are used on both.
Limits to keep in mind
- The results reflect what players and tournament organizers entered on TopDeck. Missing or wrong decklists, or a mislabeled event, end up in the numbers.
- A region with few tournaments gives noisy results, and one big event can dominate a country.
- Top cut size and standings are taken as TopDeck reports them. We don't have the results of the cut itself, only who got into it.
Found something that looks wrong?
Tell us on Discord or open an issue on GitHub. The counting code is public: scripts/build-meta.mjs builds the data and site/js/regional.js does the maths.