What Actually Separates a Bracket 2 Deck From a Bracket 4 One
Most bracket advice is people describing their own decks. We have 5,375 Commander decks that have been rated and then corrected by the people who own them, so we can check which of the things players argue about actually track the bracket, and which do not.
Two of the most argued-about numbers in deck building turn out to have no relationship with bracket whatsoever. A few things matter a great deal. And one of the strongest signals is not a deck-building choice at all.
You can rate your own deck with the bracket checker; this guide is about the general pattern.
The corpus, and what the numbers are
5,375 decks rated through our API, where the rating was then confirmed or corrected by a human. Most corrections come through the deck tools we power for other sites, the rest from people using our own checker.
Two honest caveats before any number, because both change what the results mean.
These are bracket labels, not game results. Nobody watched these games. A deck is bracket 3 here because a person said so after seeing our rating, not because it won on turn six. That makes this a good measure of how experienced players classify decks, and not a measure of how fast decks actually win.
Some things correlate with bracket by rule, not by discovery. Game Changer count and two-card combo count set bracket floors under the WotC chart. Of course they track the label; the chart makes them. We report them separately and draw no conclusions from them.
Everything else below is a free variable: nothing in the rules says it should track the bracket.
The median deck at each bracket
| Bracket | Decks | Fast mana | Tutors | Avg CMC | Taplands | Lands | Interaction | Median price |
|---|---|---|---|---|---|---|---|---|
| 1 | 75 | 1 | 1 | 3.07 | 9% | 28 | 10 | $200 |
| 2 | 1,140 | 1 | 1 | 3.23 | 18% | 32 | 10 | $230 |
| 3 | 2,793 | 2 | 2 | 3.11 | 11% | 35 | 10 | $562 |
| 4 | 992 | 3 | 3 | 2.94 | 8% | 33 | 10 | $797 |
| 5 | 375 | 12 | 9 | 2.15 | 3% | 28 | 12 | $2,484 |
Two things that do not move your bracket
Removal and interaction. The median deck runs ten pieces of interaction at bracket 1, ten at bracket 2, ten at bracket 3 and ten at bracket 4. The correlation with bracket is 0.035, which is nothing. Only cEDH moves, and only to twelve.
This is the number people argue about most. "Your deck is too strong, it has too much removal" is a sentence said at kitchen tables every week, and the data does not support it at all. A bracket 2 deck and a bracket 4 deck carry the same amount of interaction. What changes is what they do with the rest of the list.
There is a small signal in cheap interaction, correlation 0.152: higher-bracket decks put their answers on cheaper cards. Five pieces costing two or less at bracket 4 against four at bracket 2. That is a real but modest effect, and it is about efficiency, not quantity.
Land count. Correlation minus 0.058, also nothing. Bracket 2 runs 32, bracket 4 runs 33. Bracket 5 runs 28, fewer than bracket 3's 35, which is the opposite of what you would guess if you thought stronger decks were doing more.
Cutting lands does not make a deck a higher bracket. It makes it a worse deck.
What actually separates bracket 2 from bracket 4
This is the comparison that matters, because it is where most real decks live and where most disagreements happen.
| Bracket 2 | Bracket 4 | |
|---|---|---|
| Fast mana | 1 | 3 |
| Tutors | 1 | 3 |
| Average mana value | 3.23 | 2.94 |
| Taplands | 18% | 8% |
| Median price | $230 | $797 |
| Average EDHREC rank of cards | 3,502 | 2,009 |
Fast mana is the biggest free signal, correlation 0.528. Two extra sources of fast mana is the single clearest structural difference between a bracket 2 deck and a bracket 4 one. Not twenty. Two.
Taplands are the cheapest lever nobody talks about. Bracket 2 decks run 18% of their lands entering tapped, bracket 4 decks run 8%. That is not a power difference, it is a money and attention difference, and it costs you a tempo hit every game. If you want your deck to play a bracket higher without changing a single spell, this is the place to look. Correlation minus 0.310, as strong as average mana value.
Curve matters, modestly. Average mana value drops from 3.23 to 2.94 between the brackets. Meaningful, but smaller than people assume.
Card choice matters more than card power. The average EDHREC rank of the cards in a deck is one of the strongest free signals in the whole set, at minus 0.420. Higher-bracket decks play more commonly played cards. They are not full of secret technology; they are full of the cards everyone already knows are the best ones.
The uncomfortable one
Median deck price runs $230 at bracket 2, $562 at bracket 3, $797 at bracket 4 and $2,484 at cEDH. Correlation 0.555, the strongest free signal we measured, ahead of fast mana.
We are not going to dress this up. Bracket tracks money more closely than it tracks almost any individual deck-building decision. Some of that is real: fast mana and efficient lands are expensive, and those genuinely move the bracket. Some of it is that expensive cards tend to be efficient cards.
It does not mean you have to spend to play at a bracket. The bracket 2 and bracket 3 decks in this data are separated by about $330 at the median, which is a manabase, not a collection. But anyone telling you that bracket is purely about deck-building skill is describing a world the numbers do not show.
Bracket 5 is a different game
Look again at the cEDH row. Twelve fast mana sources against three at bracket 4. Nine tutors against three. Average mana value 2.15. Twenty-eight lands.
Bracket 4 and bracket 5 are not neighbours on a scale, they are different formats. A bracket 4 deck is a strong Commander deck. A cEDH deck is a tournament deck that happens to have a hundred cards. If you are wondering whether your optimized deck is secretly cEDH, this row is the answer, and it is almost certainly no.
Bracket 1 is unusual, not weak
The average EDHREC rank of cards in a bracket 1 deck is 6,625, against 3,502 at bracket 2 and 1,236 at cEDH. Bracket 1 decks are built from cards almost nobody plays.
That matches what the bracket is supposed to mean. Exhibition is not the bucket for bad decks, it is the bucket for decks built around an idea rather than around winning. The data shows exactly that: they are not running weaker versions of normal cards, they are running different cards entirely. Our bracket guide makes the same point from the rules side; it is nice to see the decklists agree.
What to do with this
If your deck is rated higher than you expected, look at your fast mana count and your Game Changers before you look at your removal. Cutting interaction to seem more casual does not work and makes your deck worse.
If you want to play up a bracket, fix your lands first. Taplands are the difference the data shows between bracket 2 and bracket 4 manabases, and unlike Game Changers they carry no chart consequences at all.
If someone tells you your deck is too strong because of what it kills, the numbers are on your side. Ask them about your fast mana instead.
Frequently asked questions
Does adding more removal raise my bracket?
No. The median deck runs ten pieces of interaction at every bracket from 1 to 4, and the correlation between interaction count and bracket is 0.035, which is no relationship at all. Only cEDH decks run more, and only slightly. Cheap interaction is mildly associated with higher brackets, but that is about efficiency rather than quantity.
How many lands should my deck run?
The data has nothing useful to say about bracket here, which is itself the answer: land count does not track bracket. Bracket 2 decks run a median of 32 and bracket 4 decks 33. Build the manabase your curve needs and do not cut lands hoping to look like a different bracket.
What is the biggest difference between a bracket 2 and a bracket 4 deck?
Excluding the things the chart mandates, fast mana. The median bracket 2 deck runs one source and the median bracket 4 deck runs three. After that: tapland share, which falls from 18% to 8%, and how commonly played the deck's cards are.
Is bracket just about how much money you spend?
Not just, but more than people admit. Median deck price is the strongest free signal we measured, correlating 0.555 with bracket, ahead of fast mana. Bracket 2 sits at $230 and bracket 4 at $797. Some of that is genuinely causal, because efficient mana is expensive, and some is that expensive cards tend to be good ones.
Are these real game results?
No, and this matters. These are bracket labels assigned by people, either confirming or correcting our rating of their deck. Nobody recorded who won or on what turn. It is a strong measure of how experienced players classify decks and not a measure of how fast decks actually win.
Why are Game Changers and combos not in the findings?
Because they set bracket floors by rule. A Game Changer forces bracket 3, four of them force bracket 4, and a two-card combo forces bracket 3. They correlate with bracket because the chart says they must, so reporting that as a discovery would be circular. Every finding above is a variable the rules say nothing about.
How do I find out my own deck's bracket?
Paste your list into the bracket checker. It rates contents and expected speed, tells you which one set the bracket, and names the specific cards responsible.