How to tell your Commander deck’s power level

The “what’s your power level, 1 to 10?” question fails because nobody shares a scale. These signals are checkable, so two people can reach the same answer.

Ask five Commander players to rate the same deck out of ten and you will get five answers. The number is doing no work, because there is no shared anchor: your 7 is my 5. The way out is to stop rating vibes and start counting things that are actually visible in the decklist.

Signals that genuinely predict power

• Game Changers. The single most reliable marker, because the list is curated centrally — counting them is objective, not a matter of taste.
• Combo presence AND compactness. A deck that wins with two cards is in a different league from one that needs four pieces plus a turn to set up. Compactness matters more than whether a combo exists at all.
• Manabase quality. Fetches, duals and fast mana are expensive and unglamorous, and they separate tuned decks from casual ones more sharply than the flashy cards do.
• Average mana value. Cheap curves act sooner. This is a weaker signal than the ones above, but it is the one that tracks “how fast does this deck do its thing”.

Signals that sound right but mislead

This is where most power-level heuristics go wrong, and it is worth being specific because these mistakes are extremely common.

• Raw tutor count. Tutors do increase consistency, but counting them punishes fair decks that tutor for a land and ignores decks with one tutor that finds a game-ending card.
• Keyword-matching card text. Scanning oracle text for “destroy all lands” or “extra turn” produces false positives constantly — a planeswalker ultimate that destroys lands is not a mass land destruction deck.
• Price. Expensive decks are often strong, but plenty of budget builds are brutal and plenty of expensive ones are precious piles of pet cards. Price is a weak proxy, useful only in aggregate.
• Win rate in your own pod. This measures your playgroup, not your deck. It is the most tempting wrong answer of all.

A five-minute self-assessment

• Count your Game Changers. Zero keeps you in casual territory, up to three is a tuned deck, and four or more is high-powered — the same count the bracket system uses to draw its lines.
• Ask what your fastest realistic win looks like, not your best-case dream draw.
• Count the cards in your shortest win line. Two is a different deck from four.
• Look at your lands. If most of them enter tapped, you are not as fast as you think.
• Say the answer out loud to someone else in your pod. If they laugh, they are probably right.

How ScrySpire scores it

Paste or import a decklist and ScrySpire computes a Deck Power score out of 100 plus a tier, using the signals above — Game Changer flags and card data from Scryfall, combo detection from Commander Spellbook. The tier is the headline; the number gives finer texture inside it.

It deliberately does not merge the two lowest brackets into separate scores, because no composition signal reliably separates them — the difference between a bracket 1 and a bracket 2 deck is play intent, not card choice. Being honest about what the model cannot see is part of what makes the rest of it trustworthy.

Frequently asked questions