PDA

View Full Version : Weighted Dice



Lappy9000
2008-10-16, 04:46 PM
So, I got a 1-pound bag of randomized dice for my birthday over the summer; including a little bit of everything, from inversed d4's to a blank d6 and percentile. However, I'm pretty sure that at least one of the dice in the grab bag is weighted. Is there any official way to determine this?

fractic
2008-10-16, 04:48 PM
Hold it between thumb and index finger on opposite corners and spin it. It shouldn't wobble. Works better if you don't use your fingers but something that produces less friction.

Also dropping it repeatedely in a glass of water.

sleepy
2008-10-16, 04:51 PM
...gyroscope?

Emperor Tippy
2008-10-16, 04:53 PM
Roll it a couple thousand times and record all the rolls, see if it comes up with 1 number significantly more often than the average.

LibraryOgre
2008-10-16, 04:53 PM
You can also roll it three to four hundred times, recording the results of each roll, then averaging. It should be close to (die size + 1)/2.

fractic
2008-10-16, 04:53 PM
...gyroscope?

How would a gyroscope help?

Defiant
2008-10-16, 04:54 PM
Roll it 10000 times and see what the standard deviation is.

Rei_Jin
2008-10-16, 04:54 PM
Using water would be the easiest method.

Y'see, the heavier side should end up on the bottom 99% of the time, as long as the drop in water is deep enough.

If you have a glass flower vase, you could use that to test it out.

Lappy9000
2008-10-16, 04:55 PM
Roll it 10000 times and see what the standard deviation is.

O_o



Thanks for the suggestions, I'll give 'em a try.

fireinthedust
2008-10-16, 04:57 PM
how would a gyroscope NOT help? It's a gyroscope!!! Those things have mad skillz, yo.

weighted vs. filed down: does it weigh physically more than another die? Or are the angles of the die weird on one side?

if you sand down one side then redo the 20, you can get this. Do the numbers look good or crappy?

If there's a ballbearing in the die... let me know where to get one, as it'd come in handy for my GMing to have better control when I absolutely need to fudge. (PCs shouldn't fudge, but GMs are expected to, y'know?)

fractic
2008-10-16, 05:02 PM
If there's a ballbearing in the die... let me know where to get one, as it'd come in handy for my GMing to have better control when I absolutely need to fudge. (PCs shouldn't fudge, but GMs are expected to, y'know?)

Hence the DM screen.

Defiant
2008-10-16, 05:03 PM
Hence the DM screen.

That is usually composed primarily of my left hand and my players' honesty. :smallwink:

Lappy9000
2008-10-16, 05:06 PM
how would a gyroscope NOT help? It's a gyroscope!!! Those things have mad skillz, yo.

weighted vs. filed down: does it weigh physically more than another die? Or are the angles of the die weird on one side?

if you sand down one side then redo the 20, you can get this. Do the numbers look good or crappy?

If there's a ballbearing in the die... let me know where to get one, as it'd come in handy for my GMing to have better control when I absolutely need to fudge. (PCs shouldn't fudge, but GMs are expected to, y'know?)

It's actually a d12, of all things. The dice are from Chessex (http://www.chessex.com/) in the completely awesome Pound o' Dice (http://www.chessex.com/Dice/poundofdice.htm). One die has the paint chipped away around the "1" and I can clearly see where the weight is. There's another one like it, but the paint hasn't chipped away there, so I wanted to test it out. I suspect that they both are, since they roll differently from a normal d12. They don't flounder near the end, like the usual 12-sider does.

Mushroom Ninja
2008-10-16, 05:07 PM
how would a gyroscope NOT help? It's a gyroscope!!! Those things have mad skillz, yo.


I dunno, I've always been a little creeped out by them (http://imgs.xkcd.com/comics/gyroscopes.png)

sleepy
2008-10-16, 05:53 PM
I thought that was how casinos tested dice? They mount them in the middle a gyroscope and spin it, and it will wobble or something if it's not balanced? The earlier reply of holding opposing corners and spinning it is the same principal... but worse for many reasons, not the least of which is the lack of gyroscope.

xPANCAKEx
2008-10-16, 05:56 PM
smash it and see if it holds any hidden treasure

fractic
2008-10-16, 05:58 PM
I thought that was how casinos tested dice? They mount them in the middle a gyroscope and spin it, and it will wobble or something if it's not balanced? The earlier reply of holding opposing corners and spinning it is the same principal... but worse for many reasons, not the least of which is the lack of gyroscope.

That's how a casino does it allright. But that's not a gyroscope (http://en.wikipedia.org/wiki/Gyroscope).

Prometheus
2008-10-16, 06:04 PM
I say don't get rid of the weighted die, weight all the other ones to different random values.

sleepy
2008-10-16, 06:06 PM
Yeah, I'm aware of what a gyroscope is. I am, however, saddened by the implication that there is no removeable-dice-core version of this contraption.

DM Raven
2008-10-16, 06:41 PM
You could do that, or you could just buy or use a different die, I hear they're pretty cheap these days...

Tokiko Mima
2008-10-16, 07:29 PM
Or you could just watch them as they roll. Weighted dice roll funny and curve off in odd directions because their center of gravity isn't where it should be. Roll them each a good distance a couple of times on a flat surface, and if you can't seem to roll consistantly in a straight line, it's probably weighted.

Epinephrine
2008-10-16, 09:43 PM
If you can see the weight, it's weighted.

The question though is whether it's weighted enough to matter. I think a chi-square test is your best bet though, not looking at the average. Roll it 50 times per side (so 600 times for a d12), record the number of times each comes up, and plug the result into a chi square test. Your expected number of results is 50 times for each number.

Here is a chi square tester:

http://www.physics.csbsju.edu/stats/chi-square_form.html

Your number of items is 12 for a d12, the X01, X02, ... are the number of times the result came up, the E01, E02, ... are the expected number of times (in this case, 50 for each, though you can use any number really, the more you roll the more sensitive the test will be).

JaxGaret
2008-10-16, 09:48 PM
Is there any official way to determine this?

Yes. They're weighted.

Unless you're buying premium dice, they're not going to be as random as they should be. There's two good videos on youtube that Matthew posted last week about this.

Mewtarthio
2008-10-16, 10:55 PM
Have one of your players roll it in-game. If it comes up "12," it's weighted.

Darrin
2008-10-17, 05:20 AM
There's an easier way to tell if your dice are biased. Buy some epsom salts, and dissolve them into a bowl/glass of hot water. This increases the density of the water and causes the die to float. Drop the die in, and watch which face floats to the top. The heaviest end of the die will sink towards the bottom.

Unless you're using casino dice, pretty much all plastic dice are going to show a bit of a bias, because the shape and density of the plastic is going to be a little different in each die. Some of the solid dice may even have air bubbles inside that you can't see. This is why transparent dice are supposed to be "fairer", although even transparent dice can have differences in shape and density. When I did this, I was surprised to find that my oldest 20-sider that I'd had since junior high was biased toward the number "16"... which may not be a bad side to be biased toward.

Haven't tested my current set... but then I hardly ever buy epsom salts.

Attilargh
2008-10-17, 06:37 AM
One die has the paint chipped away around the "1" and I can clearly see where the weight is. There's another one like it, but the paint hasn't chipped away there, so I wanted to test it out.
If you have an appropriately accurate scale, you could just weigh the weighted and a regular d12 and compare the suspect's weight to theirs.

Kaiyanwang
2008-10-17, 07:07 AM
Definitely X-rays.

Telonius
2008-10-17, 09:32 AM
It's actually a d12, of all things.

Then it should have just practically no in-game effect regardless. :smallbiggrin:

Talya
2008-10-17, 09:49 AM
Then it should have just practically no in-game effect regardless. :smallbiggrin:

that depends on game systems!

I'm playing in a deadlands campaign, and frequently must roll 5d12.

valadil
2008-10-17, 09:54 AM
that depends on game systems!

I'm playing in a deadlands campaign, and frequently must roll 5d12.

Even without weighted dice I wish I had a pull like that.

Blackfang108
2008-10-17, 09:58 AM
Roll them each a good distance a couple of times on a flat surface, and if you can't seem to roll consistantly in a straight line, it's probably weighted.

I can't roll a consistant straight line with unweighted dice...

you can try spinning the die on a corner and see its behavior.

The Glyphstone
2008-10-17, 10:08 AM
There's an easier way to tell if your dice are biased. Buy some epsom salts, and dissolve them into a bowl/glass of hot water. .

I read this, thought you were talking about the dice, and wondered if maybe that wasn't a bit of an extreme method to determine if the die was loaded...:smallsmile:

EvilElitest
2008-10-17, 03:49 PM
So, I got a 1-pound bag of randomized dice for my birthday over the summer; including a little bit of everything, from inversed d4's to a blank d6 and percentile. However, I'm pretty sure that at least one of the dice in the grab bag is weighted. Is there any official way to determine this?

wait, weighted as in trick dice or what?
from
EE

LotharBot
2008-10-18, 01:01 AM
The best way to test it is to float it in water (with an appropriate amount of dissolved salt, in order to make the die float). See if it tends to turn itself toward one side, and how quickly it does so. Most dice are going to be a little bit unbalanced, but some will be very unbalanced.

Of course, this only helps you test for unbalanced weight. If you want to test for overall bias, there are other factors -- if one face is smaller than others, or if its edges are more rounded than others, it's less likely to land down (and therefore you're less likely to get the number on the opposite face landing UP.)

The best way to test for overall bias is just to roll it a bunch and look at the statistics. How often does it roll each number, versus how often it should? (Use the chi squared test presented above if you want to be statistically precise. A badly unfair die will be somewhat obvious even without going through the calculations.)

EvilElitest
2008-10-18, 09:16 PM
IF worst comes to worst you can simply use the useless dice as ammunition for a slin shot
frm
EE

Lappy9000
2008-10-18, 09:52 PM
Yup. It's a weighted die. Thanks for the suggestions guys/gals.


IF worst comes to worst you can simply use the useless dice as ammunition for a slin shot
frm
EE

Wait, why on earth would it be considered useless? I find it rather awesome that I have at least 1 weighted die none of my players know about *summoning blackspeeker through sheer force of irony*