How easy is to to cheat with dice?

June 14, 2013 8:19 AM

Rolling a standard six sided dice, how easy would it be for someone to roll it in a way that makes sixes (or any number) come up more frequently? Would it be obvious if someone was doing it?
posted by Cannon Fodder to Sports, Hobbies, & Recreation (17 answers total)

Best answer: Not easy, apparently. The Wizard of Odds has some stories on dice setting.
posted by photovox at 8:23 AM on June 14, 2013 [1 favorite]

Are you trying to do this or trying to figure out if someone else is doing this? You could try this: http://i.imgur.com/ceDW5.jpg
posted by ejazen at 8:23 AM on June 14, 2013

Standard dice? It's not very easy at all. Discretely palming them and swapping them for loaded dice? Much easier.
posted by leotrotsky at 8:24 AM on June 14, 2013 [4 favorites]

Also, see dice control.
posted by leotrotsky at 8:26 AM on June 14, 2013

I believe I heard Penn Jilette talking about this once. With years of practice, someone can increase their odds of rolling their chosen number. But not by very much- it's more like tipping the odds from 1:6 to 11:60. It's something you could use to have a better shot at a craps table, but not something you could use to call your shot and win a bet.

This is also why many dice games require a wall bounce for a roll to be valid.
posted by gjc at 8:29 AM on June 14, 2013 [1 favorite]

yup. loaded dice. and they are noticeable.
posted by QueerAngel28 at 8:32 AM on June 14, 2013

Best answer: There was a story a few years ago, I think in Harper's, about attending a dice-rolling camp, where the proprietors claimed they could teach you rolling techniques to subtly affect the outcomes of your rolls. It's behind a pay/subscription wall at the Harper's site, but it's from the December 2008 issue, titled "The Golden Touch" by Mattathias Schwartz. My recollection of the article is that the slant of the writing goes towards the training not really doing anything and not really being impressed by the skills or exhibitions put on by the proprietors, with the caveat that craps is a game where if played a certain way the odds are pretty good for the bettor, so even a very small percentage increase is helpful for someone who gambles a lot.
posted by LionIndex at 8:42 AM on June 14, 2013

Best answer: Back when I played a lot of D&D, I could reliably roll 1D6 and get the number I wanted. It required that the die lay in with the chosen number up, and that a particular surface was used (a spiral bound notebook) - and a lot of practice. But I was pretty good at it, given those constants.

With two dice it was harder, but possible - the key was to keep them from bouncing off each other during the throw. I never could manage to get three dice to reliably roll my way. Hence my preference for rolling stats on an individual D6. ;-)

You need a softish surface, so you can control the roll. If there is any elasticity at all (from a hard table top, for instance) it became much harder.

In craps, the dice have to bounce off the back wall - this adds a layer of complexity I don't think can be overcome. Also, It's a longish throw. It's much easier if you can just roll the dice in front of you.

FWIW, a D12 and D20 is pretty easy to do this with, too because of the uniform shape. I never could get a single D10 to roll predictably.
posted by Pogo_Fuzzybutt at 8:43 AM on June 14, 2013

I can usually fool my (young) kids by pretending to roll a die, but actually more-or-less placing it on the table. You can achieve a lot with distraction.
posted by pipeski at 8:58 AM on June 14, 2013

You might be interested in this documentary where Chow Yun Fat demonstrates his ability to roll a 1 with a pair of dice.

Like Pogo_Fuzzybut, I used to be able to reliably "roll" a 6 under certain very controlled conditions. It's not something you could use in gambling as it would be immediately obvious what I was doing.
posted by Eddie Mars at 9:58 AM on June 14, 2013 [1 favorite]

Best answer: The magnum opus text on these matters is "Scarne on Dice" (note that the title is misspelled in that link, and is correct as I've typed it here) by magician John Scarne. In it he discusses (at dissertation-length) the history of dice loading and dice manipulation.

I found a copy in a used book store as a child and was overwhelmed by both his historical and quantitative thoroughness. It introduced me to basic probability in ways that have stayed with me over a lifetime.

As others have said: dice loads can significantly alter outcomes, but the type of throw either A) alters it in very miniscule ways at best, or B) is so obviously a gimmick roll that everybody watching would call bullshit on it immediately (and would be rendered ineffective with a shaker cup or a wall-bounce).

I'm remembering imperfectly, and the book is in storage, but the most effective (well... "least ineffective") gimmick throw was called the "whip... something", where you threw the two dice together -- with the desired pip facing up on the bottom die -- and tried to have them hit in a way that would freeze the bottom die on that pip. As a child I practiced that throw for maybe 40 hours total, and couldn't make it work much more often than chance.
posted by jjjjjjjijjjjjjj at 10:42 AM on June 14, 2013

In the course of my job, I have digitized the entire library collection belonging to a recognized expert on casino gambling cheating, and though I didn't of course read the several hundred volumes word-for-word, I absorbed enough to be able to say that: it is impossible to reliably throw a pair of non-loaded dice in a craps setting and have them land in a manner of the thrower's choosing. In the long run, dice control does not work. There are, however, no end of guides and instructions that will teach you how to do it -- for a price -- because the idea is so seductive.
posted by Devoidoid at 12:15 PM on June 14, 2013 [1 favorite]

I have perfected a method of rolling a pair of dice and having 7 come up more often that any other number.
posted by kenaldo at 12:22 PM on June 14, 2013 [6 favorites]

It is surprisingly easy to not notice a loaded die if you aren't looking for it: The DM of our GURPS game accidentally mixed some funny d6s into his big bowl of dice, and one player played with them for a bit before the DM noticed which pair he was rolling with. He has some that don't have all the numbers on them (They have two of one number), which is really obvious if you inspect the die, but not hard to miss when grabbing three dice out of a giant bowl. Heck, I once found I was rolling a misprinted die as I'd gotten it as con swag and never checked it, because why would I? I finally tossed it, as I had two dice, one good and one bad and kept getting them mixed up. He also has one that he added a little bit of weight on one face, so that they roll a bit higher on average then they should. Not enough to visibly roll funny (I've seen those before), but enough to bias them. (Said player was not happy; The dice were biased high, and you want to roll low in GURPS)

Nothing that would pass at a real casino, or probably not if there was money on the line between friends, but easy to do when playing GURPS and not paying attention.

Oh, and I once went for half a game session rolling a d10 instead of a d20: My Dad gave me some of his old dice, collected back when a d10 was a 20 sided die numbered 1-10 twice, before they made d20s. This was D&D, and I was wondering why I couldn't roll above 10, but why would anyone make a d20 that didn't go up to 20?
posted by Canageek at 10:26 PM on June 14, 2013

It's do-able to basically drop the die on a given face and make the lead-in to that look a little like you rolled the die across your hand and aren't just dropping it... but even if you're not cheating, that's a lousy way to roll a die, and it's not anything that stands up to the first suspicion you're up to something.

why would anyone make a d20 that didn't go up to 20?

Kids these days. Why, in my day we didn't have those newfangled ten-sided trapezohedrons the kids play those vampire games with; we didn't have d20's numbered 1-20; an icosahedron numbered 0-9 twice, hand-colored with different crayons was the only game in town. AND WE LIKED IT!
posted by Zed at 3:36 PM on June 15, 2013

I forgot: Back in high school one of my players developed a rolling method where he set a d20 on one of its vertexes and spun it, that SEEMED to give him a higher chance of a 20, before I banned it. But we didn't play with it for long, so it could have been coincidence.

Also watch out for countdown dice: Normal d20s have the faces scattered around randomly. However, d20 made for scoring have the numbers in sequential order, making the throw much easier to bias to high or low numbers.

(I knew about the d10/d6 way of rolling, as my Dad didn't own a d20 until 2002 when he went to a con and a DM made him use one, as he couldn't understand what he was doing (Um, 1-3 is 0, 4-6 is +10 to the d10, not that hard...) but I hadn't seen the icosahedrons. I'm not sure how he got the trapezohedrons without also getting a d20 though...
posted by Canageek at 10:41 AM on June 16, 2013

Response by poster: Ok, that helps thank you. I'm not interested in doing this myself, curious about the difficulty. The consensus seems to be thats its pretty tough to do without being super obvious, and bouncing off a surface will kill the roll. Yeah, I imagine loaded dice would make things easier!
posted by Cannon Fodder at 11:27 AM on June 17, 2013