How Commander Players Rate Their Own Decks
May 4, 2026
A few months ago we added a feedback form to ScryCheck — a way for players to tell us whether their deck’s rating felt right. 184 responses later, the data says something genuinely interesting about how Commander players think about their own decks.
The breakdown
The responses split into three buckets: players who thought the rating was accurate, players who thought it was too high, and players who thought it was too low.
184 total responses, Apr–May 2026.
The first thing that stands out: players are far more likely to think their deck is underrated than overrated. “Too low” is nearly 1.5× more common than “too high” — 82 responses versus 53. And only about one in four thought the rating was accurate.
This is consistent with something we’ve seen in the broader deck data and that most regular players know from experience: Commander players tend to believe in their decks. Sometimes that’s well-founded. Often it’s not. But either way, the impulse is nearly universal.
It depends on which bracket you’re in
The aggregate skew toward “too low” masks a much more interesting pattern. When you break the feedback down by the bracket the deck actually received, the picture looks almost nothing like the headline numbers.
n=14
n=26
n=46
n=22
n=8
Bracket 1 and Bracket 2 are nearly unanimous: over 80% of players in each say their deck was rated too low. Bracket 1 has zero “too high” responses at all. Nobody wants to be told their deck is a Bracket 1.
Bracket 3 is the outlier — the only bracket where the “accurate” response is actually the plurality at 46%. That makes sense: B3 is where the format lives. 40% of all analyzed decks land there, the power range within it is the most varied, and players in B3 are perhaps the most calibrated about what that label means.
Bracket 4 and 5 flip almost completely. 68% of B4 players and 63% of B5 players say their deck was rated too high. This is where hard-gate escalations concentrate — and it tells a specific story about how players relate to those markers.
When the engine flags something objective
The more interesting pattern isn’t the overall skew toward “too low” — it’s what happens when a deck carries one of the WotC bracket markers.
The bracket system includes a set of markers — specific card types whose presence triggers a hard gate, automatically placing a deck at a minimum bracket regardless of anything else in it. A 2-card infinite combo, chaining extra turns, mass land denial, or four-plus Game Changers each function as markers. These are objective calls. Either the card is present or it isn’t.
When a deck in our feedback data triggered one of these markers, the “too high” responses didn’t disappear — they persisted. Players disagreed with the bracket their deck was placed in even when an infinite combo or a full Game Changer suite was the reason.
That’s not a calibration problem. The engine is doing exactly what the bracket framework says to do. The disagreement is coming from somewhere else.
Ratings are social, not just technical
Here’s the thing about a bracket rating: it tells you something about the deck in isolation. It doesn’t know your table.
Consider this deck — Sam, Loyal Attendant, built around Lord of the Rings hobbits, Food tokens, and lifegain. It runs 32 lifegain synergy cards, 31 token producers, and is themed around fellowship, second breakfasts, and the Shire. ScryCheck rates it at power level 5.3, which the engine calls an “upgraded/tuned build.” Its bracket is 4.
The reason: one 2-card infinite combo detected among its 99. The bracket system sees the combo and correctly escalates to Bracket 4 minimum. The player — who built a cozy hobbit deck and probably doesn’t assemble that combo most games — sees a Bracket 4 label and thinks the rating is too high.
Both of them are right, in different ways. The bracket is measuring what the deck is capable of. The player is measuring how it actually functions at their table.
This plays out in the other direction too. A player trying to get into a more competitive pod might want their deck rated higher than it scores — not to misrepresent it, but because they feel it plays up to that level against their usual opponents. The bracket number has social weight. Players interact with it accordingly.
The bracket system is a shared language for a personal game
Commander is, at its core, a social game. Every pod has its own norms — how hard people are trying to win, how much they care about the ending, what kinds of interactions are fun at that table versus frustrating. A bracket rating can’t encode any of that.
What the bracket system does well is create a common reference point that doesn’t depend on individual judgment calls. Infinite combos and Game Changers are enumerable. “This feels like a 7” is not. The framework trades precision for objectivity on purpose — and that tradeoff is worth it, because a shared vocabulary only works if it means the same thing to everyone at the table.
The feedback data suggests that players understand this intellectually but still filter it through their meta. When a marker triggers a hard gate, the question most players are really asking isn’t “is this rule correct?” It’s “does this label describe how my deck actually plays at my table?” Those are different questions, and no rating system answers both at once.
What this means for the pregame conversation
The takeaway isn’t that players are wrong to push back on their ratings. It’s that a bracket alone — or a power level alone — will always leave something on the table for the conversation to fill in.
“I’m Bracket 4, but it’s dragon tribal and I’ve never assembled the combo” is a more useful statement than “I’m Bracket 4.” It gives the rest of the table the context the bracket doesn’t carry.
The data reinforces something that experienced players already know: the pregame conversation isn’t a formality. It’s where the bracket label gets translated into something that actually fits the table. The number starts the conversation. It doesn’t finish it.
Run your deck through ScryCheck to get a bracket rating, power level, and full vector breakdown — useful data for that pregame conversation.
Analyze your deck →