Doing science on dice superstition

[BreathingMeat]

So I had a particularly bad night with my dice last week. In a typical combat round, my dragonborn fighter would perform an Insightful Strike on an enemy and a Dragon Breath on a cheerful little group of minions - five attack rolls with approximately 60% chance of hitting, costing me two encounter powers, and a net result of zero hits and a lot of laughter from the other players.

My d20s were doing shifts in the freezer all evening as punishment - but to no avail. I carried on rolling 2s and 3s until the end of the session. It was terrible.

So over the next few weeks I'm going to do something about it. I'm going to do SCIENCE about it. Starting tonight, I'm going to apply one dice superstition per week and see which one results in the most success.

Tonight it's going to be the old standby: Store the d20s on the table with their 20s up all the time when they aren't being used, so that they get used to being in that position, and become more comfortable yielding high numbers. I shall publish the results of this method in this thread after tonight's session.

Next week I'll try something else. I'm very happy to take suggestions.

Now I don't want to be branded a cheat, so ideas like "blow on the 1 so that it gets sprayed with condensation and sticks to the table," which might have some actual physical reason to work, will not be considered.

Things I plan to try:

- Storing d20s with 20s up so they get used to being in that position and become more comfortable yielding high numbers Results

- Prepping dice by rolling all my d20s a few times before the session, and then playing with the one which has had all its low numbers "rolled out of it" during prep.

- Rolling a few d20s at the start of the session to see which one plans to co-operate

- Telling the d20s that the one which rolls worst this week will be kept in the freezer for a week.

- Carrying out my threat and putting the one that rolled worst last session in the freezer, and then seeing if the other dice become more co-operative.

- Rubbing the dice on the player who is rolling best to get some of their luck onto my dice.

- Buying a new die that is in the prime of its life and hasn't had all the good rolls rolled out of it.

[Edits]

- Rolling attack and damage dice together, but not from the same "set". This prevents them from becoming elitist.

- Storing dice with 20 down so the dice yield 20s out of curiosity or rebellion.

- Using an ugly die (ugly dice try harder)

- Rolling dice from a cup (because the dice don't like touching me)

- Speaking "lovingly" to the dice

- Writing "I will roll well tonight" on the battle mat

- Actually, I might also try writing "I will roll poorly tonight" on the battle mat and thus challenge the dice to defy me.

- Burning a particularly unlucky die to death and gathering the others around to watch.

- Dropping the dice from between cupped hands (hmm, this one might be a bit cheaty. I'll have to make sure there's a bit of spin on them as they tumble out)

- Storing the dice with some fresh playing cards to make them crazy.

Other suggestions?

[Logalmier]

You should use an online dice roller if this keeps happening. One of your dice may be uneven.

OR

Buy loaded dice.

EDIT: Ohhh, you said you DIDN'T want to be branded as a cheat? Never mind then.

[drengnikrafe]

Do it 3 times. We need replication. If all 3 trials turn out favorable, there would be science to it, in theory. If it's 2/3, try it twice more.
We're looking for statistical significance, which means less than 5% chance it occured by random.

[Myou]

Assuming the methods don't physically affect the dice, what on Earth are you expecting to find? They'll all average out exactly the same.

[Logalmier]

Assuming the methods don't physically affect the dice, what on Earth are you expecting to find? They'll all average out exactly the same.

But, but, SCIENCE!

[drengnikrafe]

This way, we can prove once and for... now... that dice superstition is (or isn't) real.

[AslanCross]

Dice lose their shape eventually, so their rolls are not guaranteed to be uniformly random.

I do respect your Mythbusters-like attitude, though.

I saw a guy "bribe" his dice with money. :P It worked. Also, microwaving the offending dice and making the others "watch."

[Myou]

But, but, SCIENCE!

Yes, the science of statistics! Which states that random die rolls will average out over time. And then there's simple logic that tells us that methods that do not affect the die or roll, will not affect the result.

So really all you'd be doing is trying to find patterns in random die rolls. And die rolls don't exhibit any recurrence relation.

[Totally Guy]

This is one benefit of using a dice pool system.

The players don't spontaneously personify the dice because now they all look the same and it can't all be blamed on just one die. As it was always several of them one single traitor cannot be fully blamed.

Whenever we go back to using just one D20 everybody gets silly and irrational.

Edit: Who's up for proving dice pool hive-minds?

[BreathingMeat]

@drengnikrafe: Yeah you are right, but I want to address the issue broadly before I do it deeply - I'll repeat my methods but only when I've tried out lots of different ones. It doesn't matter what order I use to collect my data, and it's more fun to try different things each week

@Myou: You do know what a superstition is, don't you?

[Draz74]

Assuming the methods don't physically affect the dice, what on Earth are you expecting to find? They'll all average out exactly the same.

No, they probably won't. Not unless his sample size is infinitely large.

Whether any conclusions he might draw are valid bases on which to build expectations of future dice performance is, of course, questionable. But he almost certainly won't get exactly the same averages from various methods.

Edit:

Whenever we go back to using just one D20 everybody gets silly and irrational.

Wait, how is this a disadvantage?

[Myou]

No, they probably won't. Not unless his sample size is infinitely large.

Whether any conclusions he might draw are valid bases on which to build expectations of future dice performance is, of course, questionable. But he almost certainly won't get exactly the same averages from various methods.

Edit:


Wait, how is this a disadvantage?

By average out I mean over extended use. I thought that was obvious.

[Darrin]

This may get you some results a little quicker: buy some epsom salts, and dissolve them in a tub of hot water. This should increase the density of the liquid to the point where your dice will float.

Drop a die in and see which side "rolls" to the surface. If your dice are lopsided, worn on one side, or have hidden air bubbles, then this should show you how biased your dice are. The side that's heaviest will sink towards the bottom, and the side facing up is the side that's more likely to come up than the other sides.

Note: give your dice a good clean wash afterwards, otherwise the dissolved epsom salts will dry and recrystalize on your dice, and this can throw off their balance as well.

[Thursday]

- Storing d20s with 20s up so they get used to being in that position and become more comfortable yielding high numbers

Never! they should always rest in any position except 20 up, so they get curious/rebellious and want to try it out.

'Course, I am convinced my dice hate me, and will want to rebel..

[erikun]

Haha, this is crazy. Seriously, if your dice are actually rolling poorly, it's probably because they are rounded or weighted off-center. You can learn a bit more from YouTube, although I don't claim it an expert opinion. (It certainly sounds reasonable, though.)

I have terrible luck with dice myself, either rolling 1-10 or above 15 at random intervals. I have a set of precision dice (sharp-edged) for when I "need" a good, fair roll.

As for superstitions I tend to use (remember, though, that my luck is terrible):

- Store dice together in the same dice bag. This gets the dice familiar with each other, so that they are used to being rolled together.

- Roll dice together as a set. This is more relevant in D&D, where you roll d20 and another die for damage.

- Roll dice from different sets together. This keeps them from becoming elitist.

[BreathingMeat]

By average out I mean over extended use. I thought that was obvious.

Dice don't really get "extended use" during a gaming session. You'd throw a d20 maybe a couple of dozen times at most, depending on what type of character you are? I don't know; I haven't tried counting ... yet.

But anyway, this nebulous thing called "luck" plays a big part in it.

[Gamgee]

Yes, I do have something relevant for this topic.

[UglyPanda]

OP, casinos use gyroscopes to see if their dice are rigged. You could probably get the same type of device if you have an engineer friend.

[Logalmier]

Yes, the science of statistics! Which states that random die rolls will average out over time. And then there's simple logic that tells us that methods that do not affect the die or roll, will not affect the result.

So really all you'd be doing is trying to find patterns in random die rolls. And die rolls don't exhibit any recurrence relation.

Blasphemy! Obviously the only way to solve this problem for sure is by the careful application of very very dangerous chemicals!

Eye of newt and frog toes ought to do the trick...

[Godskook]

First of all, there's a problem with your method. Any superstition worth its salt will take at *least* a few weeks to calibrate. Ergo, week-by-week testing will merely keep your dice decalibrated to any superstition.

Second, before you start applying superstitious science to your die rolls, why not try standard science? See if your dice have a bias towards certain numbers by rolling it out a few times.

[BreathingMeat]

This may get you some results a little quicker: buy some epsom salts, and dissolve them in a tub of hot water. This should increase the density of the liquid to the point where your dice will float.

Drop a die in and see which side "rolls" to the surface. If your dice are lopsided, worn on one side, or have hidden air bubbles, then this should show you how biased your dice are. The side that's heaviest will sink towards the bottom, and the side facing up is the side that's more likely to come up than the other sides.

Note: give your dice a good clean wash afterwards, otherwise the dissolved epsom salts will dry and recrystalize on your dice, and this can throw off their balance as well.

But that's not doing science on the superstition! That's doing science on the actual dice physics! What kind of a scientist are you anyway??

[Myou]

Blasphemy! Obviously the only way to solve this problem for sure is by the careful application of very very dangerous chemicals!

Eye of newt and frog toes ought to do the trick...

And dice whittled out of alkali metals by increasing period number! :D

[BreathingMeat]

First of all, there's a problem with your method. Any superstition worth its salt will take at *least* a few weeks to calibrate. Ergo, week-by-week testing will merely keep your dice decalibrated to any superstition.

Second, before you start applying superstitious science to your die rolls, why not try standard science? See if your dice have a bias towards certain numbers by rolling it out a few times.

Pah! Standard science is for people who are PAID to do it! My way is much more fun.

I'd argue that any superstition worth its salt would have an immediate effect.
Walk under a ladder: BAM. Smacked in the head by a cricket ball from a nearby staduim.
See a black cat: BAM. Sucked into a wormhole and emerge in the ice seas of Phobos.
Treat your dice wrong: BAM. Fail a save and get eaten by a slime.

Now THAT's science.

[Mongoose87]

Most dice are too well-made for it now, but, in the old days, leaving you dice 20 up would cause the plastic to shift towards the 1, making 20s more likely.

[pffh]

Let no one, except for you of course, touch your dice (our GM has this rule for sessions when he wants the baddies to roll well).

I'll post more when I think of them.

[ondonaflash]

Hey! Hey! What is your control!? You can't have science without a control group!

[Swordgleam]

I suggest buying some ugly dice, or bartering away other people's dice which they dislike. Ugly/unloved dice try harder.

This is one benefit of using a dice pool system.

The players don't spontaneously personify the dice because now they all look the same and it can't all be blamed on just one die. As it was always several of them one single traitor cannot be fully blamed.

Nonsense. I have 2d6 in bronze, 4d6 in gold, and 2d6 green with scarabs that I used in a SWd6 game. I would swap out the scarabs whenever a roll was important - they always rolled low.

[Jack_Simth]

Let's see...

The biggie:
Concretely track rolls. Whenever you roll a die, write down which die it was, and the result. Preferably in a spreadsheet or some other electronic manner that will let you do math on them.

Otherwise, you won't have a decent method by which to determine which hypothesis has the most impact on which die.

Other things to try:
Roll out of a cup, rather than out of your hand. Some dice might just not like touching you.

[MickJay]

Try holding the dice in your hands, and speak to them about how you love them, stroke them gently, and ask, in soft voice, if they could please roll high, just for you.

Might work, might not, but be ready with a camera to take photos of people who'll see you doing this.

[drengnikrafe]

Let's see...

The biggie:
Concretely track rolls. Whenever you roll a die, write down which die it was, and the result. Preferably in a spreadsheet or some other electronic manner that will let you do math on them.

Otherwise, you won't have a decent method by which to determine which hypothesis has the most impact on which die.

Other things to try:
Roll out of a cup, rather than out of your hand. Some dice might just not like touching you.

Also, track what they were doing. The dice, that is. Attack rolls, specific skill rolls...
Maybe your dice are fine, but just don't want you to jump.