A Field Guide to Dice Superstitions (Or: How We All Lost Our Minds)

An anthropological study of the bizarre rituals players perform with plastic polyhedrons. Yes, I'm included. Yes, I have a dice jail.

When I first started playing D&D, I was a rational person. A man of science and art. I understood dice were random number generators: small plastic objects governed by physics and probability.

Then I met my d20.

Now I own seventeen sets of dice, three of them in "dice jail" for rolling poorly, and I refuse to leave any d20 showing a 1 face-up between sessions, because obviously that's asking for trouble.

I am, as the youth say, "cooked."

The Taxonomy of Dice Rituals

Type A: The Pre-Roll Ritual

Characteristics: Must perform a specific action before every important roll.

Observed behaviours:

• Shaking dice in cupped hands while whispering encouragement
• Blowing on dice (breath does nothing to plastic)
• Kissing dice (as above, plus a hygiene problem)
• The "Dice Massage": rolling the die gently to "warm it up"
• Ceremonial lifting and dropping from specific heights

I fall into this category. Before any crucial roll, I tap the d20 on the table exactly three times. Why three? Five felt excessive and two felt insufficient. This is what passes for logic in my brain now.

We are not the first. In 1967 the sociologist James Henslin published a study of cab drivers who shot craps. They bet by the known odds, then spent the actual throw on magic: rituals meant to control how the dice fell, all perfectly logical inside the players' own theory of cause and effect [1]. Swap the cab stand for a folding table and you have us.

Type B: The Dice Collector

Characteristics: Believes different dice suit different jobs.

Observed behaviours:

• "These are my attack dice, these are my damage dice, and these are for saving throws."
• Separate sets for separate characters
• "Lucky" dice for specific situations
• Refusing to let anyone else touch their dice (reasonable, honestly)

My friend Gerald owns forty-seven sets. Forty-seven. He claims each set has a distinct "personality." Gerald, if you're reading this, please seek help. Also, can I borrow your sparkly purple set? It seems to roll well.

Gerald has a point, psychologically. Ellen Langer showed that people treat chance like skill once you give them a choice: lottery players who picked their own ticket felt more confident in it than players handed one [2]. Choosing which die to roll feels like control. It isn't.

Type C: The Dice Disciplinarian

Characteristics: Believes dice can and should be punished for poor performance.

Observed behaviours:

• Dice Jail: a small container where offending dice serve their time
• Forced Observation: making bad dice watch good dice roll well
• Threats: verbal warnings of consequences
• Exile: permanent banishment of repeat offenders

My dice jail is a small cardboard box labelled "SHAME" in red marker. Currently incarcerated:

• One d20 that rolled three natural 1s in a row during combat
• A full set that failed every death save last month
• A metal d6 I swear is cursed (it's not weighted; I checked. Twice.)

Here's the maddening part: dice jail works. Not because the die learned its lesson, but because of regression to the mean, first described by Francis Galton in 1886 [3]. A die only goes to jail after an unusually bad run. Its next rolls are ordinary, so they look better. The box labelled SHAME takes the credit.

Roll the crime, do the time.

Type D: The Dice Mystic

Characteristics: Believes in cosmic dice karma and a spiritual bond with polyhedrons.

Observed behaviours:

• Rolling only on specific surfaces for "proper energy flow"
• Cleansing dice with sage smoke between sessions
• Arranging dice in patterns when not in use
• Consulting the dice for life decisions (don't)
• Believing dice have souls (seriously, don't)

My mate Patricia runs a full cleansing ritual before every session: incense, crystals, the works. Does it work? Her character hasn't died yet, so I'm not saying she's wrong.

Science briefly sided with Patricia. A 2010 study found that people told they had a lucky golf ball sank more putts [4]. Then two larger, careful replications gave people the same lucky ball and found they putted almost exactly like everyone else [5]. The ball was just a ball. Patricia's crystals remain under review.

Type E: The Dice Atheist

Characteristics: Claims to have no superstitions whatsoever. Is absolutely lying.

Observed behaviours:

• "I don't believe in luck; it's just probability."
• Uses the same dice every session "for consistency"
• Definitely has preferences but calls them "practical"
• Rolls dice between sessions to "test them" (that's a ritual, mate, you've got one)

The most dangerous type, because they don't realise they've caught the madness. Ask one to roll your d20 and watch them wipe it on their shirt first.

The Science (Such As It Is)

Rationally, we all know:

1. Dice are random
2. Past rolls don't influence future rolls
3. Plastic doesn't have feelings
4. None of our rituals matter

However:

1. My blue d20 rolls better than my red one (demonstrably true)
2. Dice in timeout DO improve their behaviour (see Galton, above, who ruins everything)
3. Blowing on dice clearly works (because I said so)
4. The dice gods are real and they're watching

So I ran the numbers on the crimes in my dice jail. Every figure assumes a fair d20 or d6, computed exactly rather than guessed:

The "evidence"	Real odds
Three natural 1s in a row	1 in 8,000
A run of three 1s somewhere in 1,000 d20 rolls (a long campaign)	about 11%
Rolling a 1 right after three 1s	1 in 20, same as ever
Dying from death saves with no help (SRD 5.2 rules: 10+ succeeds, a 1 counts twice, a 20 revives) [6]	about 40%
At least one face showing 3+ times in a 20-roll "test" of a fair d20	about 88%

That last line is why testing a die at the table proves nothing. Twenty rolls almost always produce a "suspicious" favourite. Tversky and Kahneman called this belief in the law of small numbers: we expect a small sample to look like the whole population, and it rarely does [7]. To catch a d20 that rolls 1s twice as often as it should (10% instead of 5%), you need roughly 150 rolls, assuming a one-sided test at the usual 5% significance with an 80% chance of spotting it. Gerald's forty-seven sets would take him until retirement.

Meanwhile, across a long campaign, a streak of three 1s is closer to a one-in-nine chance than a miracle. It happened to me. The die went to jail anyway.

A Personal Confession

Last month my d20 fell under the table mid-combat, so I borrowed a friend's. That borrowed d20 rolled a natural 20 on the most important check of the campaign.

I considered stealing it. I'm not proud of this. I'm just being honest.

Instead, I asked where he bought it. Then I drove forty minutes to that shop and bought five dice of the same brand, hoping they came from the same factory batch and shared the same blessed properties.

This is who I am now.

The Meta-Ritual

Here's the darkest secret: we all know it's nonsense.

Every person doing a dice ritual knows, on some level, it's ridiculous. We do it anyway because:

1. Routine is comforting. In a game where anything can happen, a ritual feels like control. Henslin's craps shooters had worked this out long before us [1].
2. Community bonding. Shared superstitions become table culture.
3. Entertainment value. It's fun to be a bit daft with your mates.
4. Plausible deniability. "I'm just joking... unless it works."

Besides, what's the harm? So what if I have a special rolling technique? So what if I won't use dice showing 1s? So what if I've named all my d20s and keep them on a rotation?

I'm not hurting anyone.

If your physical dice are all serving life sentences, the 3D dice roller will throw a fresh set with no criminal record.

The Verdict

Are dice superstitions irrational? Absolutely.

Will I stop? Absolutely not.

Will I judge you for yours? Only if they're different from mine.

Do I genuinely believe dice have moods and personalities? [Glances at dice bag] No comment.

---

May your d20s roll high and your d4s be easy to find when you inevitably drop them in the carpet.

—Arthur

P.S. If anyone asks, I'm being satirical. I definitely don't talk to my dice. That would be absurd.

Table Talk

Tia: Why does a grown adult put a d20 in a box labelled SHAME?

James: Because it rolled three 1s. It knew what it did.

Tia: It didn't. That's Langer's point: choose the die, and chance starts to feel like skill.

James: But jail works. The post says so.

Tia: Jail works because the next rolls are ordinary. Galton, regression to the mean. The box takes the credit.

James: So my ritual does nothing.

Tia: Nothing to the die. Plenty for you. Routine feels like control.

James: Then I'm keeping it. Tip for everyone: twenty test rolls won't convict a die.

Tia: To catch one rolling 1s twice as often as it should, you need about 150. Most dice go to jail without a trial.

James: No trial? You'd think a criminologist would object.

Tia: I do. Free the d20.

References

1. Henslin, J. M. (1967). Craps and magic. American Journal of Sociology, 73(3), 316–330. https://doi.org/10.1086/224479
2. Langer, E. J. (1975). The illusion of control. Journal of Personality and Social Psychology, 32(2), 311–328. https://doi.org/10.1037/0022-3514.32.2.311
3. Galton, F. (1886). Regression towards mediocrity in hereditary stature. Journal of the Anthropological Institute of Great Britain and Ireland, 15, 246–263. https://doi.org/10.2307/2841583
4. Damisch, L., Stoberock, B., & Mussweiler, T. (2010). Keep your fingers crossed! How superstition improves performance. Psychological Science, 21(7), 1014–1020. https://doi.org/10.1177/0956797610372631
5. Calin-Jageman, R. J., & Caldwell, T. L. (2014). Replication of the superstition and performance study by Damisch, Stoberock, and Mussweiler (2010). Social Psychology, 45(3), 239–245. https://doi.org/10.1027/1864-9335/a000190
6. Wizards of the Coast (2025). System Reference Document 5.2, "Playing the Game": Death Saving Throws. CC-BY-4.0. https://www.dndbeyond.com/srd
7. Tversky, A., & Kahneman, D. (1971). Belief in the law of small numbers. Psychological Bulletin, 76(2), 105–110. https://doi.org/10.1037/h0031322