Metagame
What’s being played in Bracket 3. A commander can be built a ton and barely make it to tournaments —fun, not easy to win with— or the other way around, and here you can see both.
Color identities
Which colors are being played. It’s the first thing to stabilize, because it doesn’t depend on which commander each person chose.
- Five-color316 · 7.5%
- Grixis242 · 5.7%
- Esper239 · 5.6%
- Naya228 · 5.4%
- Mardu207 · 4.9%
- Izzet206 · 4.9%
- Dimir190 · 4.5%best 29th
- Jeskai188 · 4.4%
- Gruul177 · 4.2%
- Rakdos174 · 4.1%
- Golgari170 · 4.0%
- Mono-red167 · 3.9%
- Boros164 · 3.9%
- Mono-black155 · 3.7%
- Sultai147 · 3.5%
- Temur145 · 3.4%
- Mono-green139 · 3.3%
- Simic129 · 3.0%
Commanders
906 different commanders. The right-hand column is the one that answers "is this a tournament deck?": what share of the people who build it end up certifying it.
* marks the commanders that get certified at least at twice the overall rate (3.9%). It usually means that whoever builds it builds it to compete. The rate is only calculated with 8 or more decks built: below that, a single certificate would make it spike without meaning anything. There are 30 in this format.
Showing the first 40 of 906.
Individual lists are only published when their owner played a tournament with them. Everything else goes in here as a number, never as a list, and emails are never published.
How this is measured
4232 decks built: distinct lists that passed the checker clean, meaning they were legal when they were measured. Of those, 163 issued a certificate — an overall rate of 3.9%. A deck checked ten times while it’s being tuned counts once.
No deck is discarded for having aged: what gets discarded is the card. If a list includes something that is banned today, that card leaves the count and the rest keeps counting, because it is still a real signal of how people play. The normative lists last changed on August 20, 2026. Data cutoff 0a02ef1a7064.
3406 of these decks were checked before September 3, 2026, when the checker didn’t yet record which format each list had been checked against. The data isn’t made up: they count in both formats. Those checked since that date count only in their own.