Forum adverts like this one are shown to any user who is not logged in. Join us by filling out a tiny 3 field form and you will get your own, free, dakka user account which gives a good range of benefits to you: No adverts like this in the forums anymore. Times and dates in your local timezone. Full tracking of what you have read so you can skip to your first unread post, easily see what has changed since you last logged in, and easily see what is new at a glance. Email notifications for threads you want to watch closely. Being a part of the oldest wargaming community on the net. If you are already a member then feel free to login now.

MikeMcSomething wrote:Here's the problem - you have a cube of dice in your hand, it's either a) suitable for randomizing or b) not, and should be thrown away, and you - this is really important - can't deduce that by just rolling the cube until your arms are tired.

Some guy rolling a million dice and getting lots of ones is not "These dice roll lots of ones" it's "I rolled and got a bunch of ones" the dice don't care what they rolled before, and the law of large numbers doesn't magically make it so that if you roll a bunch of times the universe owes you a given spread or the dice are inherently bad.

Sure, a single die could roll a 1 a thousand times in a row, and it could be random.

How do you propose to tell, then, if a die is actually random, or if there are impurities that make it non-random?

MikeMcSomething wrote:Here's the problem - you have a cube of dice in your hand, it's either a) suitable for randomizing or b) not, and should be thrown away, and you - this is really important - can't deduce that by just rolling the cube until your arms are tired.

Nobody here is willing (few are likely able) to go into a detailed enough analysis of the physics involved to determine the inertia acting on a die striking a given surface to determine whether manufacturing differences you would be measuring with a micrometer on a d6 are going to make a discernible difference on what you roll.

Some guy rolling a million dice and getting lots of ones is not "These dice roll lots of ones" it's "I rolled and got a bunch of ones" the dice don't care what they rolled before, and the law of large numbers doesn't magically make it so that if you roll a bunch of times the universe owes you a given spread or the dice are inherently bad.

29% ones! Really? You guys seriously think that's true? How often do you really think you would see failed leadership tests when a given roll has over a 35% chance of containing at least a single 1 result?

Okay. Fine. Your experience differs.
Look, man. I've got 3 cubes of Chessex 12mm dice. They all roll sh**. Excessive 1's. Lots of 2's. Just sh**.
I changed to Koplow dice with squared edges and *bam* average rolls. With a few exceptional games (good and/or bad), my games are almost never plagued by an excessive # of 1's.
That's not coincidence. That's fact. We're talking 2 to 3 games a week, 2000 points, probably 35 to 40 weeks a year. That's over the last 2 years.
I've used the Koplow dice enough to know they're as close to average as is reasonable to expect of these dice. Having used Chessek for longer than that, I have enough experience to have seen the *other side* of things.

Your experiences don't match that. Fine.

Anyone saying it's the rounded edges is wrong. I'll get on board with that. What IS the problem is the low quality of the dice. They're the McDonalds of dice manufacturers.

Eric

Lycaeus Wrex wrote:I use GW dice. I have never noticed, not even really cared to notice, whether they roll more 1s or 3s or 4s or whatever. I just roll the dice and accept the result. Furthermore, I shudder to even think of the sheer number of times I'd have to roll a dice to determine which side, if any, it is biased upon.

I kind of agree with Solofalcon and johnscott here; its just a game man. Do we all really need to take it *this* seriously? And if we do, is that really healthy?

L. Wrex

I don't get how voluntarily choosing to spend money on more dice is taking the game *this* seriously.

I mean, we are choosing to spend money on little tiny plastic men to move around on a table and pretend shoot other little tiny plastic men.

Is the money wasted on plastic cubes more ridiculous than the money wasted on plastic evil space elves?

Da Butcha wrote:I don't get how voluntarily choosing to spend money on more dice is taking the game *this* seriously.

I mean, we are choosing to spend money on little tiny plastic men to move around on a table and pretend shoot other little tiny plastic men.

I see your taking little plastic men seriously and raise you talking on a forum about buying dice to play with little plastic men seriously

MagickalMemories wrote:The price that Dash quoted is for 36 dice.
So, there you go. Six more dice than you'd get from GW, with the bonus of being better made and rolling closer to average.

http://godboma.com/12mm-Opaque-Dice-Set-of-36.html
You're welcome. ; )

Ah, I see! I guess the higher cost is for non-pipped dice, then.

And MM, I can agree with you about the cheap dice because I've had sets before that were prone to 1's. But my last set was prone to roll high. I think we can all agree that cheap dice can be biased towards one side or other... I just think it's a bit of wishful thinking that they're all slanted to 1's

Da Butcha wrote: Is the money wasted on plastic cubes more ridiculous than the money wasted on plastic misunderstood space elves?

As a DE player, I felt the need to fix your post. ; )

RiTides wrote:And MM, I can agree with you about the cheap dice because I've had sets before that were prone to 1's. But my last set was prone to roll high. I think we can all agree that cheap dice can be biased towards one side or other... I just think it's a bit of wishful thinking that they're all slanted to 1's

You'll definitely get me on board with that statement. They can definitely be biased to any side of the die. I believe, however, that Chessex sees far more "1" bias than anything other.

Eric

Maybe instead of trying to use reason and common sense and explain that the physics of dice (a body moving through the air and striking another object, and those subsequent effects of those forces on the likelihood of a particular facing of said object ending up on top even when/if the facings are slightly different sizes) are likely far more complicated to model than "Hurp I get alot of ones" I should just open a dice shop, pick out all the ones that are measurably similar in size, and mark up the costs by ~12,000%, call them "Precise" and sell them to you guys.

I'll just walk over to this here closet to bust out my copy of Opening an internet busin--WHOOPS oh http://en.wikipedia.org/wiki/Analytical_dynamics jesu- http://en.wikipedia.org/wiki/Ballistic_conduction MY LEG! - http://en.wikipedia.org/wiki/External_ballistics oh sorry everyone tripped and fell on some science there my bad.

MikeMcSomething wrote:Maybe instead of trying to use reason and common sense and explain that the physics of dice...

To which I repeat my question. If you're so smart...

Ailaros wrote:How do you propose to tell, then, if a die is actually random, or if there are impurities that make it non-random?

You can pretend to be awesome by throwing down wikipedia links all you want, but unless you have something to actually positively contribute...

Edit - bleh it's inflammatory.

They do a study here:

http://dicephysics.info/0107.htm

where the abstract states that dice were found to be biased in direct proportion to the amount that they deviate from perfect cubicity - the example they use is a 3% height difference will produce a bias along that axis of (+6% of 16%) likelihood of rolling a result on that axis. If this carried through to the chessex dice they would need to be 15% shorter along the 1-6 axis (this would be hilariously noticeable), and would roll ones and sixes equally. From there, per their section on Loading, the hollow pips would then make the 6 more likely to land face up, by very little (0.8% is the example given where almost 1% more weight lies on a given side)

These magical 30% ones Chessex dice everyone keeps talking about would look more like small Frisbee discs and when stacked on their sides would take up only about half of the standard dice carrying cube's vertical space, or, be so much denser on one side that, in addition to rolling extremely awkwardly, the dice would crack and have huge chunks of the '6' faces falling out due to the forces the side that is 30%+ denser than the rest of the die was exerting on the strucure as a whole.

MikeMcSomething wrote:Edit - bleh it's inflammatory.

They do a study here:

http://dicephysics.info/0107.htm

where the abstract states that dice were found to be biased in direct proportion to the amount that they deviate from perfect cubicity - the example they use is a 3% height difference will produce a bias along that axis of (+6% of 16%) likelihood of rolling a result on that axis. If this carried through to the chessex dice they would need to be 15% shorter along the 1-6 axis (this would be hilariously noticeable), and would roll ones and sixes equally. From there, per their section on Loading, the hollow pips would then make the 6 more likely to land face up, by very little (0.8% is the example given where almost 1% more weight lies on a given side)

These magical 30% ones Chessex dice everyone keeps talking about would look more like small Frisbee discs and when stacked on their sides would take up only about half of the standard dice carrying cube's vertical space, or, be so much denser on one side that, in addition to rolling extremely awkwardly, the dice would crack and have huge chunks of the '6' faces falling out due to the forces the side that is 30%+ denser than the rest of the die was exerting on the strucure as a whole.

I'm not sure I absorbed all of that but if they found;

3% height difference = +6% result on that axis
16% = Nominal probability on a 6 sided die
30% = current result
30%-16% = 14% difference between nominal and claimed
14% / 6 = 2.33 (Shows us how many segments of 6% deviance there is)
2.33 * 3% = 7% difference in height

So a 7% difference in height will equate to rolling 30% on a given axis as opposed to 16%. I believe your mistake was in just dividing 30/6 rather than starting from a 16% supposed accuracy.

If a die is 12 mm in height that is .84 mm of a difference. That's pretty small and I'm not sure anyone could observe that with the naked eye. In addition, that 7% difference may be spread out along two sides (IE One axis is 4% while another is 3%, which results in a net 7% difference in height).



Regardless...of all that.... Dash's thread wasn't "Who wants to argue with me about the accuracy of dice?". He was just asking where to find some accurate/tested/precision dice. I'm not sure why anyone would care what he does with his money.

I started buying square-edged Bicycle brand dice last summer (saw them at dollar general selling 5/$1). Now i could be completely wrong here but i immediately noticed a better trend of dice rolls with these cheap, (obviously not precision) square dice. In fact it was so good that i claimed 'these dice are HOT'. My friends have actually agreed with that statement and begun using bicycle dice as well.

Basically, since the square dice don't roll all over the table when they're tossed, it is a much neater system than the rounded dice. As has been stated before, however, without a dice cup, tower, or backboard to slam against, a shrewd player could practice his rolling techniques with square dice and never fail an armor save (see Jigga2Jones on youtube for an example of this).

Totally the wrong thread, my bad.

I'm not too fussed about the maths involved or any of that science mumbo jumbo.

Square dice are poop, they roll about as well as a D4, i.e without lots of space may as well not bother.

I for one welcome my rounded edge overlords.

rolled my dakkadakka dice 600 times today rounded corners and logo instead of 6.
one came up 98 times.

probibility expects 100 times.

moral of the story become a DCM buy dice get fair awesome dice.

Wow... good thing this is just a game... I'd hate to see what would happen if there were real consequences to these dice rolls...

SoloFalcon1138 wrote:Wow... good thing this is just a game... I'd hate to see what would happen if there were real consequences to these dice rolls...

What, like playing for $1000 cash? Or a steady stream of store credit? Or smaller wagers?

Define *real consequences*

Where do you get 1000 dollars cash for playing?

Make sure you report it to the IRS, by the way.

skkipper wrote:rolled my dakkadakka dice 600 times today rounded corners and logo instead of 6.
one came up 98 times.

probibility expects 100 times.

moral of the story become a DCM buy dice get fair awesome dice.

skkipper, you're my hero

My motivation for getting some precision dice is different- I don't like it when people attribute a game result to bad (or good) dice, and since I'm thinking of going to my first GTs this summer, I'd like some that everyone would know are completely fair. Obviously, you can still go on a good/bad spree just like with any dice, but can't fault or credit the dice themselves.

I've also seen in some packets where you can ask your opponent to use your dice- so if I was playing against a person using full-sized, square-edged casino dice, I would feel fine asking them to use mine- since they would be just as precision balanced, but of a more manageable size and with feathered edges.

These are the only dice I use.

Monster Rain wrote:Where do you get 1000 dollars cash for playing?

Get invited to the Nova Invitational.

dash have you rolled you brick of dice and counted? your current dice are probally just fine.

skkipper wrote:dash have you rolled you brick of dice and counted? your current dice are probally just fine.

Not that hard counting a single roll (what would you even be counting for? an even 123456123456 spread is one of the more rare results) would do anything, I imagine it would be difficult to appeal to Dash anyway, as I believe he truly thinks that if he were to pick up a set of Chessex dice, he would have about 1/3 chance of getting a 1 on every die roll.

Got my Koplow Square edged dice in the mail yesterday (20 8mm, 20 12mm). Will play my first few games with them today and let everybody know how they go. Not necessarily if i win or lose (i'm using a wolfstar SW list) but rather if i notice a general trend of 1's, or 5's & 6's.

Dashofpepper wrote:

Monster Rain wrote:Where do you get 1000 dollars cash for playing?

Get invited to the Nova Invitational.

I'll pass.

Other than Adepticon there's no way I'm traveling that far for a tournament.

It doesn't have to be a "high stakes" vs. other issue. It can simply be that someone wants to make sure they're getting balanced results... although it's nice to hear the Dakka Dice show up being pretty balanced (I believe most people take issue with the smaller, 12mm rounded chessex dice. Dakka Dice are 16mm).

So, when the expensive dice become chilly, as all dice do from time to time, are you going to decry the price of precise dice and call for Games Workshop to issue digitized number generators?

SoloFalcon1138 wrote:So, when the expensive dice become chilly, as all dice do from time to time, are you going to decry the price of precise dice and call for Games Workshop to issue digitized number generators?

That just means your dice aren't ''precise'' enough! Time to pay more money for a set of cubes that have MORE unicorn witches in them!


Automatically Appended Next Post:

AgeOfEgos wrote:

MikeMcSomething wrote:Edit - bleh it's inflammatory.

They do a study here:

http://dicephysics.info/0107.htm

where the abstract states that dice were found to be biased in direct proportion to the amount that they deviate from perfect cubicity - the example they use is a 3% height difference will produce a bias along that axis of (+6% of 16%) likelihood of rolling a result on that axis. If this carried through to the chessex dice they would need to be 15% shorter along the 1-6 axis (this would be hilariously noticeable), and would roll ones and sixes equally. From there, per their section on Loading, the hollow pips would then make the 6 more likely to land face up, by very little (0.8% is the example given where almost 1% more weight lies on a given side)

These magical 30% ones Chessex dice everyone keeps talking about would look more like small Frisbee discs and when stacked on their sides would take up only about half of the standard dice carrying cube's vertical space, or, be so much denser on one side that, in addition to rolling extremely awkwardly, the dice would crack and have huge chunks of the '6' faces falling out due to the forces the side that is 30%+ denser than the rest of the die was exerting on the strucure as a whole.

I'm not sure I absorbed all of that but if they found;

3% height difference = +6% result on that axis
16% = Nominal probability on a 6 sided die
30% = current result
30%-16% = 14% difference between nominal and claimed
14% / 6 = 2.33 (Shows us how many segments of 6% deviance there is)
2.33 * 3% = 7% difference in height

So a 7% difference in height will equate to rolling 30% on a given axis as opposed to 16%. I believe your mistake was in just dividing 30/6 rather than starting from a 16% supposed accuracy.

If a die is 12 mm in height that is .84 mm of a difference. That's pretty small and I'm not sure anyone could observe that with the naked eye. In addition, that 7% difference may be spread out along two sides (IE One axis is 4% while another is 3%, which results in a net 7% difference in height).



Regardless...of all that.... Dash's thread wasn't "Who wants to argue with me about the accuracy of dice?". He was just asking where to find some accurate/tested/precision dice. I'm not sure why anyone would care what he does with his money.

If all of your dice had an axis that was 7% shorter, when you placed them in a dice cube every pillar would be a noticeably different size (if you stacked a couple dice on the short axis you would be able to, for example, look at the side of one die's 5 side, and see half of the top two pips of the 6 side of the dice behind it, or be able to align them in the cube so they formed a very visible staircase, etc) and if you stacked 6 of them up on their bad axis next to six good ones, you would be able to see almost half of an entire die on the other side. These are very noticeable differences.

We aren't discussing the accuracy vs cost of dice for Dash's benefit at this point (as he is a completely lost cause - notice he hasn't even acknowledged anything in the thread that shows a conclusion different to the one he has reached) we're throwing out stuff that people reading the thread, that want to actually learn something instead of being weirdly superstitious and closed-minded (anyone else notice this thread parallels a religious discussion in many ways?) can use to expand on what they might already know about the hobby. Too many wargamers have seriously flawed concepts of things like how numbers actually work, and this is a valuable place to help someone come away from the whole thing better-informed.

I'm not quite sure if you read my post. 7% is a percentile which is entirely based on the measurement in question. For example, 7% of 2 inches would likely be conducive to view from the naked eye. 7% of 12 mm = .84 mm, which again may be spread across two different planes (3% on one plane while another has 4%). 3% of 12 mm = 0.36 mm and 4% of 0.48 mm. Get out your mm ruler and try to locate 0.36 mm. In addition, your staircase analogy is flawed. For that example to hold true, each die on the right side would need to be 7% taller on only one plane (Which we have established may not be the case) and would need it's opposing die to not have deviation....which I would agree would be very unlikely. Again however, that is for a 30% deviation which I agree is very high....but not impossible. It's much more likely of 25% deviations which would equate into very small height differences in terms of .2 mm (Or less!). I'm not sure why you linked that study either. If you read the thesis/results of the link you posted they showed a axis deviation with non precision dice which would support Dash's motivations.

Again though, this was Dash's thread asking for dice information and if you feel that strongly about the subject, perhaps you should initiate a thread discussing dice statistics.

/Unsubscribed

Ailaros wrote:I'd also like to note that random number generators used by computers are not actually random. They have to have some sort of number base to seed their calculation. This number seed is usually based on the clock.

Having your numbers determined from the starting point of what time it is isn't truly random. That said, I could definitely see arguments for it along the lines of being more convenient than actually rolling dice, especially when things like FRF start coming up.

By calculation, do you mean:

(call time in nanoseconds % 6) + 1?

It's amazing how simple these things really are. The funny thing is that unless a human being can develop the precision to tap the button at precisely the right interval, it really is random. While it depends on a definite time, the unit is so small that no intentional human interference could possibly have an impact on the result.

Whilst I appreciate the thread in that it has pointed out to me many dice suppliers, I personally do not find the idea that Chessex dice roll 29% 1's to be believable.

I do believe that Chessex dice, and all other rounded dice, each may favor certain numbers due to the slightly random nature of their manufacture. However I find the idea that the number most favored by all Chessex dice is 1 to be a little absurd.

Firstly, the experiment done did 1000 rolls of each of 36 chessex dice. Whilst that may seem like a lot, it is hardly a conclusive experiment. I just wrote a small program to randomly roll a dice 1000 times. Using a computer's random number generator, which should give nearly a proper distribution, I was able to achieve results between 11% and 24% within a handful of attempts.

Whilst i have no doubt that the results were as stated, I hardly think that one experiment on one box of dice is enough to brand all Chessex with a bad brush. Ideally, I would be interested in seeing a set of 10000 results for 1000 boxes of dice - if that comes up favoring ones I will go to buy a set of casino dice immediately.