Most players don't lose Magic games because they picked the wrong card. They lose because they didn't pause long enough to ask the right question before they tapped a land.
Critical thinking in MTG isn't about knowing every interaction or memorizing the perfect line. It's a habit — pausing, scanning, and asking yourself a few sharp questions every time the game asks you to make a real decision. Mulligan or keep? Cast now or hold? Tutor for ramp or removal? Counter this spell, or save the counter for something worse later?
This guide walks through the five decision points that quietly decide most Commander games — opening hands, tutors, sequencing, threat assessment, and politics — and gives you a single five-question framework you can run at the table. By the end, you'll have a tool that works for any deck in any pod.
The Real Skill in MTG Isn't Memorizing Plays — It's Asking Better Questions
Getting better at Magic isn't about cataloging every card or memorizing combo lines word for word. It's about asking sharper questions — and asking them at the right moment. The players who win consistently aren't the ones who remember the most plays. They're the ones who pause longest before tapping a land.
If you've already drilled your combo recognition through the Milkman Combo Trainer, you've solved half the problem. This article handles the other half: the thinking between the combos.
When Should You Mulligan in MTG?
The mulligan is the first decision of every game and the one most players get wrong. Keeping a borderline hand because you're tired of mulliganing has cost more games than any misplay during the game itself.
Magic uses the London mulligan: take seven, keep or mulligan, draw seven again, bottom one card per mulligan taken. You can mulligan as many times as you want. There's no penalty for going to six. The penalty is keeping a hand that can't win.
The Five-Question Mulligan Check
Before you keep, ask:
- Do I have lands? Two to four lands in seven is the safe range. One land or six lands almost always means mulligan.
- Do I have my colors? A two-color deck wants both colors accessible by turn three. A three-color deck wants all three by turn four.
- Do I have early plays? A hand of seven powerful six-drops is dead weight if you can't reach turn six.
- Does this hand advance my plan? If you're piloting a combo deck and your seven cards are all interaction, that's not your hand.
- Does this hand lose to one piece of disruption? If a single counterspell or removal spell breaks the hand, mulligan.
How Many Lands Should Be in Your Opening Hand?
Three is the sweet spot. Two with a cheap cantrip or a one-drop is fine. Four with strong plays in hand is fine. One land is a mulligan unless you also have ramp plus card draw — and even then, it's risky.
When a "Bad" Hand Is Actually Fine
A hand with no win conditions but smooth lands, ramp, and interaction is a great keep in Commander. Games go long. You don't need to draw your finisher in the opening seven — you need to survive long enough to find it.
How to Decide What to Tutor For
Tutors are the most punishing decisions in Magic because they reward planning and punish autopilot. "I'll tutor for my best card" is almost never the right answer.
The Tutor Framework
Before you search, ask:
- Am I behind, even, or ahead?
- What do I actually need — mana, cards, interaction, or a finisher?
- What beats me on the next turn cycle?
- What wins the game if no one answers it?
- What target stays good even if the board shifts before I can cast it?
Ramp vs. Removal vs. Combo Piece
Simple rule: tutor for the thing you can't draw into.
If you have eight cards in hand and three are interaction, you don't need to tutor more interaction. You need a finisher, or you need ramp to cast the finisher. If you're behind on mana with no card draw, tutoring up Demonic Tutor itself is meaningless — you need Sol Ring, Mana Vault, or a two-drop rock now.
The Trap of Tutoring Reactively
The biggest tutor mistake at every skill level is searching for an answer to a problem that doesn't exist yet. If the table is mid-development and no one is threatening to win, tutoring for removal wastes a card. Tutor for the thing that advances your plan unless someone's about to win.
A useful test: if you tutored for this card and immediately drew your next-best option off the top, would you still want the tutor target? If yes, tutor. If no, you're tutoring out of fear, not strategy.
Sequencing: What Order Should You Play Your Spells?
The right cards in the wrong order are the same as the wrong cards. Sequencing is where most "I had the cards but still lost" games actually fail.
When to Lead, When to Hold
A clean ordering principle: information first, commitments second.
- Cantrips and card draw before bigger plays. You see more before you commit.
- Ramp before threats, so threats land with protection mana up.
- Tutors after cantrips — you've narrowed what you actually need.
- Threats before tutors only if the threat draws out the specific answer you want gone.
When to Hold Mana for Interaction
Holding up mana for a counterspell or removal feels safe but costs you tempo. The rule: hold up interaction only when there's a specific spell you'd use it on. If no opponent has untapped mana hinting at a relevant threat, spend your mana advancing your own plan.
A common Commander mistake is holding up Counterspell on turn 4 in a pod with no win lines yet. You're trading your turn for a counter you might never cast. Spend the mana.
Threat Assessment in Commander
In a 1v1 game, the threat is whatever the opponent plays. In a four-player Commander pod, the threat is whoever can win first — and that's not always the player with the biggest board.
How to Tell Who's Actually Ahead
Look at three things, in order:
- Cards in hand. The player sitting on seven cards is winning the resource war.
- Mana available. The player who can cast their commander twice in one turn is ahead.
- Win conditions in play. The player two cards from a known combo is the real threat — even if their board is small.
The player with the giant army of creatures is rarely the actual threat at a competitive table. The player with Rhystic Study, Mystic Remora, and a tutor in hand is. EDHREC's inclusion data tells you which cards win games at the highest rate — and Rhystic Study isn't in the top inclusion bracket because it makes a big board.
When to Use Removal (and When to Save It)
Removal answers exactly one question: "Does this card win the game if I don't kill it?" If yes, kill it. If no, hold the spell.
Spending Swords to Plowshares on a 4/4 beater turn 3 is almost always wrong. Spending it on Thrasios, Triton Hero is almost always right. Removal is finite. Don't spend it on the first interesting thing — spend it on the thing that ends the game if it sticks.
If counterspells are dominating your meta, the answer often isn't "more removal" — it's restructuring the deck around abilities counters can't touch. We covered the full breakdown in How to Beat Counterspell Decks in MTG (Without Blue).
Commander Politics Without Painting a Target
Politics is the part of Commander no rulebook teaches. It's the difference between winning the game and winning the table.
The Politics Framework
Before any deal:
- Who is actually winning?
- Who can stop the leader?
- Does this deal help me more than it helps the table?
- Am I exposing myself by helping?
- Will this make me the next target?
When to Make Deals
Deals are most valuable when there's a shared threat. "I won't attack you this turn if you counter their commander" is fair — both players benefit, both face the same problem.
Deals are least valuable when only one side wins. If a player offers something that's mostly upside for them, the answer is no, even if their pitch sounds reasonable. Especially if it sounds reasonable.
When to Help Another Player
Help another player only when (a) you're behind, (b) the leader will win without intervention, and (c) helping doesn't expose you next turn.
The classic mistake is the kingmaker play — helping a player win because you're already losing. That's not strategy. That's exiting the game in the most expensive way possible. If you're behind, help yourself first.
The Five-Question Pause: A Universal Decision Loop
Here's the framework that ties all of the above together. Run this before every meaningful decision — every mulligan, every tutor, every removal spell, every counter, every attack:
- Who is winning right now?
- What kills me first?
- What wins for me if I'm left alone?
- What do I lose if I make this play?
- What do I gain if I wait?
Five questions. Maybe ten seconds at the table. The pause itself is the skill. Most players are reacting to the previous turn instead of looking at the actual board.
The players who win consistently aren't the ones who memorized the most combo lines. They're the ones who pause longest before tapping a land. Drilling combo recognition handles the second half of the pause — once you can spot a winning line in two seconds, you have eight seconds left for everything else. Run your deck through the analyzer to see exactly which lines you should be drilling first.
Frequently asked questions
How do I get better at decision making in MTG?
Pause before every action. Most players misplay because they're reacting to the previous turn, not analyzing the current board. The Five-Question Pause in this guide is the fastest way to build that habit.
How do I know what to tutor for in Commander?
Tutor for what your hand can't already do. If you have card draw and ramp but no finisher, tutor for a finisher. If you have a finisher but no protection, tutor for a force-through tool. Search for what you lack, not what's strongest in a vacuum.
How do I tell who's winning in a Commander pod?
Cards in hand first, then mana available, then win conditions in play. The biggest board is rarely the biggest threat. A player with seven cards in hand and a Rhystic Study is winning the game even if their battlefield is empty.
When should I mulligan in Commander?
Whenever your hand fails the five-question check: lands, colors, early plays, plan advancement, and resilience to one piece of disruption. Six-card hands aren't weak — six-card hands you actually want to play with beat seven-card hands you keep out of stubbornness.
How do I practice this stuff between games?
Drill the patterns when you're not playing. Go through your decklist, identify every two- and three-card win line, and practice spotting them as flashcards. When you can recognize a combo in under two seconds, the rest of the pause is free for politics, threat reads, and sequencing.
Should I proxy decks I'm still learning?
If you're trying to figure out whether an archetype clicks for you, yes. Sleeving up a few proxy decks is dramatically cheaper than dropping rent money on a deck you might shelve in two weeks.
Better Players Pause More
The players who get better at Magic faster aren't the ones with the biggest collections or the deepest knowledge of the format. They're the ones who pause before they act, ask sharp questions, and let the answers shape the play — instead of pattern-matching to the last game they remember winning.
Build the pause. Run the five questions. Trust the framework over the gut feeling. The wins follow.
If you're trying to figure out whether an archetype clicks for you, sleeving up a few proxy decks from Milkman Proxies is dramatically cheaper than dropping rent money on a deck you might shelve in two weeks. Combo databases like Commander Spellbook make the source material easy to pull.