Thirty Days of Bad Keeps: A Field Report From the Mulligan Trenches

Let's be honest about something: when we keep a bad hand, we don't think we're keeping a bad hand. We think we're keeping a hand with potential. There's a whole vocabulary of self-deception built around the mulligan — "it's close," "I've got the land coming," "I want to see what the deck does." So I decided to run the experiment nobody asked for: for thirty days, every time I got a genuinely questionable opening hand, I kept it on purpose and wrote down what happened.

The rules were simple. If a hand was a clear keep or a clear mulligan, play it normally. But if I hesitated — that little flicker of "hmm" — I had to keep it and log the result, my reasoning, and my emotional state at the time. Thirty days, forty-one logged keeps across paper FNM, Arena, and kitchen table games.

The Numbers Weren't the Interesting Part

The bad keeps lost. A lot. Something like 68% of them, which won't shock anyone who's read a single article about mulligan math. One-land hands with three expensive spells lost basically every time they got drawn, which is also not a revelation.

But here's what the win rate didn't capture: how those games were lost.

Games after a reluctant keep featured more misplays. Not slightly more — noticeably, obviously more. I'd fire removal at the wrong target. I'd hold up countermagic I needed to spend. I'd make greedy attacks trying to "make the keep worth it." The bad hand didn't just put me behind on resources; it put me behind on judgment.

The Tilt Tax Is Real and I Was Paying It

The pattern that emerged from my logs was embarrassingly consistent: the hands I kept angry played worse than the hands I kept curious.

When I kept out of frustration — usually after a previous loss, or because I was down a game in a match — my post-keep decision quality cratered. I was playing to justify the mulligan decision instead of playing the game in front of me. Every draw became evidence in an argument I was having with myself. When the missing land didn't come, I didn't think "that's variance." I thought "of course," and the next three decisions got worse.

Meanwhile, the hands I kept because I was genuinely experimenting — same quality of hand, different emotional state — lost at nearly the same rate but I played them cleanly. I made good plays with a bad board. I found the cheap wins that bad hands sometimes offer.

Same cards. Same odds. Completely different pilot.

What Was Actually Driving the Keeps

Reading back through my notes, the "reasons" I logged for keeping were almost never about the hand. They were about me:

• "Don't want to go to five" (fear of losing card quality — even when the six was worse than five)
• "I've lost two mulligans in a row already" (gambler's fallacy, mulligan edition)
• "This deck is supposed to mulligan well" (brand loyalty to my own decklist)
• "I want to actually play Magic tonight" (impatience, wearing a strategy costume)

Not one entry said something like "the mana ratios support this keep" or "the matchup rewards this shape of hand." The math was never in the room. The feelings were.

The Heuristics I Kept After the Experiment Ended

I don't keep bad hands on purpose anymore — the data did its job. But the month left me with three rules that have genuinely changed my mulligan game:

1. Name the emotion before you name the keep. If your justification starts with "I've already" or "I don't want to," you're describing your mood, not the hand. Mulligan.

2. Ask what turn the hand does something. Not what it might do. Not what it does with the perfect draw. Turn count, concretely. If the honest answer is "five or six, maybe," the honest answer is mulligan.

3. Track the keeps you regret, not the hands you lose. Losing with a good keep is variance. Losing with a keep you hesitated on is a decision you can fix. Only one of those is worth your practice time.

The Hand Was Never the Problem

Here's the punchline of the whole month: I blamed the shuffle for years, and the shuffle was almost never the problem. The problem was that mulligan decisions are made by a person, and that person is carrying the last game's result, the match pressure, and a stubborn belief that this particular pile of seven cards owes them a win.

The deck doesn't tilt. You do. I did. The good news is that once you see the pattern in your own keeps, you can't unsee it — and a mulligan you take with clear eyes is the cheapest edge in Magic.

Ship the bad hand back. Your ego will recover. Your win rate will too.