I Carried a Lucky Pencil for 20 Years. This Semester, My Phone Replaced It

I have a pencil in a drawer at my parents' house in Ohio that I genuinely believe helped me pass AP Calculus. Yellow, chewed on one end, with a smiley face eraser worn down to a nub. I used it for every quiz, every unit test, and finally the AP exam itself. When I lost it for a terrifying week in April, my practice scores dipped — or at least I was convinced they did.

I was 17 and I knew, intellectually, that a pencil had no power over integrals. I used it anyway. So did half my class. Someone had a lucky flannel. Someone else wrote answers in the same blue pen every time. We were all performing the same small ritual, and nobody thought it was weird.

Turns out, we were participating in something much older and much bigger than a high school superstition.

The Omamori Problem: Charms Work, Just Not the Way You Think

In Japan, students visit Shinto shrines before big exams — most famously the entrance exams that can shape your entire life trajectory — and buy omamori, small brocade pouches blessed by shrine priests. There's a specific variety, gakugyō omamori, dedicated to academic success. Students tuck them into pencil cases, pin them to backpacks, and carry them into the exam hall. Shrines like Kitano Tenmangū in Kyoto, dedicated to a deified scholar, are flooded with students during exam season.

Here's the part that surprised me when I started digging into the psychology: nobody involved thinks the pouch itself magically fills in the right bubbles. The charm works on the person, not the outcome.

Researchers have studied this for decades. A well-known 2010 study from the University of Cologne found that people with lucky charms performed measurably better on tasks involving confidence, memory, and puzzle-solving — not because luck intervened, but because the charm reduced anxiety and boosted what psychologists call self-efficacy, your belief that you can actually do the thing. Superstition, in this framing, is a delivery mechanism for calm.

American students do the same thing with different props. The lucky pencil. The lucky hoodie. The roommate who has to text you "you got this" at exactly 7:45. The ritual is the point.

So when I heard the term "digital omamori" — using an AI study tool not just as a tutor but as a calmness anchor — I was skeptical. But I'd carried a pencil for 20 years. I figured I owed the experiment some rigor.

The Setup: Finals Season as a Laboratory

The rules I set for myself were simple:

1. Four weeks, three finals. Statistics, Organic Chemistry II, and a history elective I'd been ignoring.
2. One AI study companion. I used a general-purpose chatbot with a saved, persistent prompt that turned it into a quiz generator, a scheduling nagger, and — critically — a pre-exam pep talk machine.
3. The reframe. Every session ended the same way: the AI would summarize what I had genuinely mastered that day, in plain language, and I'd screenshot it. That screenshot was my omamori. Evidence, not vibes.
4. The measurement. I rated my anxiety 1–10 before each study session, before each exam, and immediately after sitting down. Old-school, but consistent.

The key insight I stumbled into: the content of the charm matters less than the certainty it represents. My pencil represented nothing. It was a feeling with no data behind it. The screenshots were different — they were receipts. "You have correctly answered 41 of the last 50 stoichiometry problems." That's a charm with a transcript.

What Actually Happened

Week one was ordinary. Anxiety ratings hovered around 7 before sessions. The AI was useful but felt like a tool, not a talisman.

The shift happened around week two, and I think it happened because of the ritual, not the AI. Ending every session with a summary of what I'd proven I knew started to rewire how I walked into study time. I stopped dreading it. My pre-session numbers drifted down to a 5, then a 4.

The night before my organic chemistry final — historically my anxiety apex — I did something new. Instead of cramming, I asked the AI to run a 20-minute review of only the topics I'd already mastered, then generate my summary screenshot. My pre-exam rating that morning: a 4. My previous semester's equivalent: a 9.

Did I score better? Yes, in every class, compared to my midterm performance in the same courses. Am I claiming causation? Absolutely not. Sample size of one, no control group, and I also simply studied more consistently than usual. But the anxiety data felt real in a way grades don't. I was calmer in the room. I read questions once instead of three times. I didn't blank on question one, which is my signature move.

Why the Reframe Beats the Tool

Here's my honest takeaway: the chatbot wasn't magic, and neither was my pencil. What worked was converting vague dread into specific, reviewable evidence of preparedness. The AI just happens to be embarrassingly good at generating that evidence on demand.

There's a lesson in the omamori tradition worth stealing: the shrine priests don't promise you'll pass. The charm promises you'll walk in with your best self intact. That's a much better deal anyway — it's the part you actually control.

If you want to try this for your next exam, the recipe is short:

• Tell your AI tool exactly what you're studying and when the exam is
• Drill with it daily, even for ten minutes
• End every session by asking: "Summarize what I've proven I know today"
• Screenshot it. Save it. Read them all the night before

Then, if you want, keep the lucky pencil too. I won't judge. I still have mine — right next to a folder of screenshots that finally outperformed it.