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.

 Pouncey wrote:

To go back to an earlier argument, I'm going to show you just how much you misunderstand probability.

*sigh*

However, I don't know how to calculate the unlikelihood of rolling 29% 1s instead of 16% 1s over the 72,000 rolls of GW and Chessex dice that averaged out to 29% 1s. I don't even know what to type into Wolfram Alpha. You claim to, since you said you know it's the same as 1 million dice all coming up 6, i.e. 6^1000000.

Million actually balanced dice will never come up with all sixes nor do 72000 actually balanced dice come up with 29% as ones. Granted, the latter is vastly more probable than the former, but both are still unlikely enough that they're practically impossible.

But that's not the claim. The claim being made is that the dice are not properly balanced, that is the claim that the study makes that is the claim you make, and that was my joke all-sixes claim. So indeed, if the million dice are weighted so that they always roll six, then the probability of them rolling six is 100%*.

I merely said this because you were willing to accept and outlandish claim because the person making the claim said they rolled a lot of dice, and you justified your belief in this claim by the large number of dice purportedly being rolled. Logically it should follow, that you would accept any obviously non-sense claim as long as the person making it said that they had rolled more dice than the person making the previous bogus claim you believed.

* (granted, it is most likely impossible to so heavily imbalance a die, in a million rolls even a heavily weighted die would fail occasionally, but that's besides the point.)

------

But my actual claim is this: If the study was correct, and ones were indeed 1.8 times as likely as they should be, this would be instantly apparent even in a small test. All these people who have made their own tests with hundreds or thousands rolls would have easily noticed this. They didn't, it's not real.

 Crimson wrote:

 Pouncey wrote:

To go back to an earlier argument, I'm going to show you just how much you misunderstand probability.

*sigh*

However, I don't know how to calculate the unlikelihood of rolling 29% 1s instead of 16% 1s over the 72,000 rolls of GW and Chessex dice that averaged out to 29% 1s. I don't even know what to type into Wolfram Alpha. You claim to, since you said you know it's the same as 1 million dice all coming up 6, i.e. 6^1000000.

Million actually balanced dice will never come up with all sixes nor do 72000 actually balanced dice come up with 29% as ones. Granted, the latter is vastly more probable than the former, but both are still unlikely enough that they're practically impossible.

But that's not the claim. The claim being made is that the dice are not properly balanced, that is the claim that the study makes that is the claim you make, and that was my joke all-sixes claim. So indeed, if the million dice are weighted so that they always roll six, then the probability of them rolling six is 100%*.

I merely said this because you were willing to accept and outlandish claim because the person making the claim said they rolled a lot of dice, and you justified your belief in this claim by the large number of dice purportedly being rolled. Logically it should follow, that you would accept any obviously non-sense claim as long as the person making it said that they had rolled more dice than the person making the previous bogus claim you believed.

* (granted, it is most likely impossible to so heavily imbalance a die, in a million rolls even a heavily weighted die would fail occasionally, but that's besides the point.)

------

But my actual claim is this: If the study was correct, and ones were indeed 1.8 times as likely as they should be, this would be instantly apparent even in a small test. All these people who have made their own tests with hundreds or thousands rolls would have easily noticed this. They didn't, it's not real.

I just want to point this out, to end this argument once and for all. The fact everyone else's results are different from the article's doesn't even mean someone has to be lying. Everyone can be reporting exactly what they see, exactly as they see it, and no one is lying at all. That is in fact something science would expect to happen in this case.

That article, even if everything it said was true, doesn't mean that the GW dice being rolled in it weren't perfectly balanced to roll each result 1/6 of the time. You can roll casino dice, in the way a casino requires you to roll, 72,000 times in a row and get 29% of those rolls being a 1. It would be very unlikely to happen, but if you repeat something enough, you get streaks of consecutive results that are highly unlikely.

Even something as rare as rolling 1,000,000 d6s and having them come up 6 every time doesn't mean the dice are necessarily loaded. That can happen, and statistically speaking, according to the exact math I just said for calculating how often that would happen, 1 in (number with over 700,000 digits) times.

What science uses to prevent their perfectly valid studies from reporting a random fluke, is called a replication study. Everyone who's tried to replicate this article has found completely different results that conform to what we already know. That doesn't mean the original article's results were wrong, it just meant that what the article recorded happening, was a statistical fluke.

You don't call a fluke the result of someone lying their ass off. When you think you can show that the exploration study was a fluke, you don't call anyone a liar at any point. The article wouldn't be lying. Everything they said happened, would've actually happened, exactly as they said it. It was just a fluke that no one's been able to replicate at any point. Which is what science does replication studies precisely to find out - whether the original results were a fluke or not. Flukes. Are not. Lies. They are flukes.

The argument you should've tried to get the point across to me that the article's conclusion that GW dice are imbalanced was wrong, isn't that they were lying. It's that scientifically, the replication studies being done by hundreds of other individuals, whose results, while not as expansive in the scope of the study, were still statistically valid, to confirm the original article, were showing that the original study was unable to be replicated and thus was probably a fluke.

Calling the guy who did that original article a liar should never have figured into it at all. You don't even need to do so to suggest it's probably wrong about the thing it was saying about GW dice. The fact that other people who try to do similar things don't get similar results is all you needed, you just needed to phrase it the right way.

And like I said. This thread is the first time I ever saw anyone even disagree with that article. And yes, I am a bit of a slow learner, even the IQ test I had done at a psychiatric evaluation said so. I am a very, very slow learner. My processing speed IQ (i.e. how quickly I learn) is an 86. That is between mental retardation and the absolute low end of normal, in a 10-point category between the two. So please understand if took me this long to finally learn that I might've been wrong, I am unable to learn at even a normal speed so you're probably going to have to explain things to me a lot.


Automatically Appended Next Post:
Also, if you ever want a completely undeniable argument to use against me the next time I decide to mention that earlier thing about the time I thought I could control the roll of dice, I'll give it to you, right now. This is a counter-argument that I know I cannot deny is an absolute counter-argument, whose logic even I cannot deny, to the point where there is something which I am totally unwilling to mention at any point where this fact is likely being considered inside the mind of someone I am trying to convince the thing I said about being able to roll the same die 15 times and get the result I wanted each time. Because I absolutely believe that the thing I am about to reveal will prove beyond a shadow of a doubt to any objective person, including myself, that my apparent claim about "mind powers" is not a thing that actually happened, despite happening right in front of me.

I am a schizophrenic.

I was still incredibly sick when that entire thing happened. Either I hadn't been diagnosed yet or I was so early in recovery I was still quite sick.

No one was even home when I rolled those dice. No one ever saw me do it who I could ask if it ever happened.

One of the things schizophrenia can do is cause you to see one thing, as another thing. That's called an illusion if you want to look it up, and yes, it is different from a hallucination where you see things that aren't there at all.

One of the other things about schizophrenia is that the symptoms fit into reality perfectly. Schizophrenics can't tell when they're hallucinating, and often those hallucinations blend into reality so much there are things I have seen when I was most ill where I'm pretty sure I hallucinated in there somewhere, but just from my memory I can't tell which part, just that two things didn't fit together that should have but I don't know which part was real and which part wasn't. Also, both things may not have been real.

One of the other things about schizophrenia is that it causes you to believe things that most people accept as not being true. For example, that you can control the results of dice rolls with your mind.

So even I, at some level, fully believe that the most likely explanation anyone can offer for why I was apparently able to roll 15 dice in a row the way I wanted to, is that I rolled 15 dice in a row and saw the results I was expecting, even though another human being who watched me do it would've seen me roll a random set of results including numerous scatter results.

Why did I bother arguing differently earlier?

Because part of the reason I type such weird posts is that writing things out is how I think things through. You are literally seeing my reasoning in-progress as I try to understand things when you read my posts. And what I just said there is not something I have ever described in full before. To anyone. Ever. Before then, that idea existed as a vague concept in my mind that if someone were able to link me being a schizophrenic with my apparent mind-dice powers with calling the latter to be BS. That's basically the extent of the entirety of this post where I explain why that thing I argued for a while about earlier on is probably BS. I didn't know it was BS then. I do now. And in the future, because I typed all this out, I won't ever be saying I can control the results of dice with my mind anymore, not because I'll get an argument, but because I don't believe it even happened at all anymore.


Automatically Appended Next Post:
I actually have a more plausible explanation than "random fluke" to explain why no one can replicate it.

That article was written 10 years ago.

Maybe GW and Chessex heard about it and started making their dice better since then?

Ok, thank you for your honesty.

I too first considered the random fluke explanation, but after seeing the purported scale of the experiment, I realised how mind-bogglingly unlikely it would be.

Often when scientific experiments give really strange results, it turns out that there has been some error in how the experiment has been conducted. However, considering how relatively straight-forward testing the dice is, I really cannot come up with any experimental error that would explain the results. Of course it is possible that such could exist and I'm just not clever enough to figure out what it could be.

As for older dice being faulty or there being faulty batches of dice, I considered that too. This article has been around for long time and people have reported results contradicting the article for a long time. Had such faulty dice indeed existed in the past, surely some of the people doing their own tests would have them too? Hell, it is quite possible that the dice I used for my tests are over ten years old. I've played 40K since the second edition, and while the dice I used definitely are not that old they're not new either. I cannot remember when I got them. So while faulty batches explanation is far more likely than the random fluke explanation it is still really not that likely.

So maybe I'm a cynic, but to me a hoax seems like most likely explanation. The motive is in the article, to sell the the 'perfect' dice. This sort of thing is hardly uncommon, it happens all the time. So whilst I certainly cannot say that this is definitely what's going on here, to me it seems like the most likely explanation.

I have one possible explanation that could explain a legitimately faulty test; the dice may not have been "rolled" in a conventional way, but rather mechanically agitated. More like being shaken rather than rolled. In that case, smaller differences in weight distribution would have an exaggerated effect.

To give an analogy, image a bucket of rocks and a bag of sand being poured into an empty bucket. If I pour them both in at the same time, there will be a "random" distribution of sand, rocks, and air pockets. However, if I vibrate and agitate the bucket I'm pouring both into, I'll end up with more sand on the bottom and less air pockets (which would be noticeable by having the bucket appear to be less full). The small imperfections in the shape and distribution of the mixture would have an exaggerated effect.

In this way, one could "coax" miniscule imperfections that normally have no nominally apparent bearing on the standard distribution to be exaggerated into appearing to be a much larger effect. Now, the link in the dakka article links to a non-existent page, so I can't confirm their methodology, but I can imagine that in order to keep dice rolls separate and to roll so many dice, that a tray constructed of small spaces for these dice may have been set up separately and independently from each other to allow for large numbers of individual dice and rolls to be made. Reading the dakka article, it appears that's what was done (edit: specifically, he said he made little plastic spaces in order to track individual dice, which would suggest this kind of methodology, but doesn't confirm it), and this could definitely allow for these conditions.

I think the reason dice roll so bad is because companies are more interested in making the dice look "pretty" that to roll well.

Due to construction methods of "pretty" dice they are not true cubes. They can be different dimensions on the X,Y, and Z axis. some dice are as much as 3% wider in one dimension than another. In addition some of their sides are not parallel to the opposing side.

Also, as most dice are made from injection molded plastic there is a dimple on one side (usually the one side) were the plastic is injected into the dice-mold. This is later trimmed off, but often a slight bump is still present. This is enough to bounce the one side up off a bottom facing. That makes a 1-roll more probable, which is what I think is happening with GW dice.

Companies also put their dice into a polishing machine. This makes them smooth, but also rounds the sides and corners to varying degrees.

Some of the worst offenders are the "clear" dice that have an ornamentation inside. These dice are rarely balanced correctly.

I can't remember the guy or his company, but he basically set out to make some of the best dice out there for true randomness. The best way of course is to make dice from milled metal, but that's expensive as heck. He instead chose to make "ugly" dice that are not polished. The "fill nipple" is trimmed off and ensured to not have a bump, but is not polished out either.

 Yarium wrote:
I have one possible explanation that could explain a legitimately faulty test; the dice may not have been "rolled" in a conventional way, but rather mechanically agitated. More like being shaken rather than rolled. In that case, smaller differences in weight distribution would have an exaggerated effect.

To give an analogy, image a bucket of rocks and a bag of sand being poured into an empty bucket. If I pour them both in at the same time, there will be a "random" distribution of sand, rocks, and air pockets. However, if I vibrate and agitate the bucket I'm pouring both into, I'll end up with more sand on the bottom and less air pockets (which would be noticeable by having the bucket appear to be less full). The small imperfections in the shape and distribution of the mixture would have an exaggerated effect.

In this way, one could "coax" miniscule imperfections that normally have no nominally apparent bearing on the standard distribution to be exaggerated into appearing to be a much larger effect. Now, the link in the dakka article links to a non-existent page, so I can't confirm their methodology, but I can imagine that in order to keep dice rolls separate and to roll so many dice, that a tray constructed of small spaces for these dice may have been set up separately and independently from each other to allow for large numbers of individual dice and rolls to be made. Reading the dakka article, it appears that's what was done (edit: specifically, he said he made little plastic spaces in order to track individual dice, which would suggest this kind of methodology, but doesn't confirm it), and this could definitely allow for these conditions.

That actually sounds somewhat plausible. I am not sure I am convinced that such a method could produce such a large error, but maybe it is possible. Good thinking in any case.

One way we could find out, if we had the original data, would be to see if other dice had preferential sides as well. The more the individual dice show preference towards certain rolls (not just 1) the more likely this potential mistake in the methodology would become.

EDIT: Or that the dice that did roll 1s very consistently rolled 1s.

 Crimson wrote:

 Yarium wrote:
I have one possible explanation that could explain a legitimately faulty test; the dice may not have been "rolled" in a conventional way, but rather mechanically agitated. More like being shaken rather than rolled. In that case, smaller differences in weight distribution would have an exaggerated effect.

To give an analogy, image a bucket of rocks and a bag of sand being poured into an empty bucket. If I pour them both in at the same time, there will be a "random" distribution of sand, rocks, and air pockets. However, if I vibrate and agitate the bucket I'm pouring both into, I'll end up with more sand on the bottom and less air pockets (which would be noticeable by having the bucket appear to be less full). The small imperfections in the shape and distribution of the mixture would have an exaggerated effect.

In this way, one could "coax" miniscule imperfections that normally have no nominally apparent bearing on the standard distribution to be exaggerated into appearing to be a much larger effect. Now, the link in the dakka article links to a non-existent page, so I can't confirm their methodology, but I can imagine that in order to keep dice rolls separate and to roll so many dice, that a tray constructed of small spaces for these dice may have been set up separately and independently from each other to allow for large numbers of individual dice and rolls to be made. Reading the dakka article, it appears that's what was done (edit: specifically, he said he made little plastic spaces in order to track individual dice, which would suggest this kind of methodology, but doesn't confirm it), and this could definitely allow for these conditions.

That actually sounds somewhat plausible. I am not sure I am convinced that such a method could produce such a large error, but maybe it is possible. Good thinking in any case.

Well, the error wasn't always 29% 1s. Some individual dice were as low as 23% 1s, others were as high as 33% 1s. I do admit that time-saving measures may have been employed which may have skewed the results though.

I wonder though. Apparently he then bought a new package of GW dice and rebuilt the corners with his lab's equipment, to be like the dice that had originally rolled 19% 1s. Then another 36,000 total rolls were made with GW dice whose corners had been rebuilt, and ended up with 19% 1s as a result.

Which means that the guy thinks that cutting off the corners of GW and Chessex dice is what unbalanced them, and simply not having done so would result in some decently-fair dice where he would've concluded from his original 72,000 GW/Chessex results turning out with 19% 1s, that GW dice are just as fair as other standard gaming dice, so it was all in everyone's heads the whole time.

So I have to ask, if not to save money by saving material, why exactly do GW and Chessex dice have their corners shaved off? What advantage does that provide to spend money on machines to do that instead of just leaving them with square edges?


Automatically Appended Next Post:

 Yarium wrote:
One way we could find out, if we had the original data, would be to see if other dice had preferential sides as well. The more the individual dice show preference towards certain rolls (not just 1) the more likely this potential mistake in the methodology would become.

EDIT: Or that the dice that did roll 1s very consistently rolled 1s.

Yeah, I did note that the Dakka article mentioned no other results but 1s.

Also, even the dice that rolled twice as many 1s as normal still usually didn't roll a 1 and were twice as likely to not roll a 1 than they were to roll a 1.

How consistent is "very consistent" to you?

I'm not a scientist so i couldn't say exactly. Best i could say is that it depends on the sample size. Generally though I'd say any die that rolls a 1 more than 1 standard deviations from the mean is suspect, and anything more than 1.5 would be almost definitely mean that something is wrong.

 Yarium wrote:
I'm not a scientist so i couldn't say exactly. Best i could say is that it depends on the sample size. Generally though I'd say any die that rolls a 1 more than 1 standard deviations from the mean is suspect, and anything more than 1.5 would be almost definitely mean that something is wrong.

I don't know what a standard deviation is, but I'm gonna guess it's doubling the odds of rolling a 1 in this case.

To which I say to you, 33% is less than double 16.7%.

33% is the highest any individual die in that article rolled 1s.

So basically the furthest deviation the article suggested was happening with GW/Chessex dice, from fair dice, was less than doubling the odds of rolling a 1.

The lowest on GW/Chessex dice was 23%, which is only 6.3% higher 1s than you would get from the 16.7% of 1s you normally roll with fair dice. Less than 1/3 more 1s than normal.

And the article basically said that if GW and Chessex stopped shaving the corners off their dice they'd roll as fairly as any other dice you're likely to be able to buy at your local gaming stores.

The article NEVER said that GW dice even MOSTLY roll 1s, much less always roll 1s. It said that GW dice roll somewhere between a third extra and double the number of 1s you'd get while rolling at a casino.

Every 100 rolls with the odds the article was saying GW and Chessex dice had with their corners shaved off, would get you somewhere between 23 and 33 1s instead of 16 or 17, depending on which dice you actually use from your collection to make those rolls.

The shape of the corners should have absolutely on effect on the randomness of the dice, as long as every corner is same the shape.

Pouncey, the study you keep referring to is not conclusive enough to prove that Chessex and GW dice are unbalanced and biased towards rolling a 1 because it has too small of a sample size. Not of sample size of rolls, but a sample size of the physical dice. The tester only used 36 dice from each variant (GW, Chessex, Casino, cornered) for a 144 dice total. Considering that each company probably manufactures hundreds of thousands of dice (if not millions) to test ONLY 36 dice is a ridiculously low sample size (that is only 0.003% of the dice tested at most)

 Crimson wrote:
The shape of the corners should have absolutely on effect on the randomness of the dice, as long as every corner is same the shape.

I forget what the article said about that, but apparently yeah, it does make a difference.

He even got an explanation from one point from a physicist who called bullgak on the fact he was getting more 1s than more 6s because both of them thought they should be rolling more 6s given how they were unbalaced. Loaded dice put the heaviest side on the bottom to weigh it down, so the heaviest side of a die with pips is the 1, since it has the fewest pips, should mean 6s came up.

After calling bullgak and trying to replicate the results himself, the physicist came back to him and said that the dice were so light that the mass lost from those extra pips was probably being outweighed by the centrifugal force of the die spinning as it rolled.

We're basically talking about a guy who, if what he's saying is true, arranged a total of 180,000 dice rolls to show the results being off was just in everyone's head, ended up proving they weren't, and phoned a casino, a dice manufacturer, and asked a physicist in his university for explanations of what he ended up showing.

I don't know WHY that is somehow less valid than some random person telling me they rolled 600 dice and then telling me they came up with a normal number of 1s.

You're not even offering me a combined total of more possibly-fake dice rolls than the number of possibly-fake dice rolls the thing you're trying to prove wrong with those possibly-fake dice rolls!

You're basically offering me less insubstantiated evidence than the insubstantiated evidence you're trying to prove wrong, without any sort of valid explanation, and expecting me to take what you say that I can't prove, over what someone else says that I also can't prove, essentially just because it's YOU saying it and not HIM.

Why should I believe what you're saying is true when I can't prove it happened any more than I can prove the thing that is obviously more statisticcally valid than what you're saying, and you're just offering me your own random guesswork which is obviously incredibly flawed given what I showed you on a recent page about the thing one of you said, as an explanation, over the things that casinos, dice manufacturers and physicists are saying about the randomness of dice rolls.

The reason I don't believe the article is wrong yet is that you haven't actually offered as much evidence as it has yet.


Automatically Appended Next Post:

 CrownAxe wrote:
Pouncey, the study you keep referring to is not conclusive enough to prove that Chessex and GW dice are unbalanced and biased towards rolling a 1 because it has too small of a sample size. Not of sample size of rolls, but a sample size of the physical dice. The tester only used 36 dice from each variant (GW, Chessex, Casino, cornered) for a 144 dice total. Considering that each company probably manufactures hundreds of thousands of dice (if not millions) to test ONLY 36 dice is a ridiculously low sample size (that is only 0.003% of the dice tested at most)

They tested 36 GW and 36 Chessex dice. 72 dice. All straight out of the package that was bought for this purpose only.

Then it rolled them 1,000 times each. 72,000 rolls.

Each die rolled at least 23% 1s. ALL 72 dice were unbalanced. Not even ONE was not.

Yes the article established that those 72 dice were unbalanced. But how is testing 72 dice out of a million proof that all million dice are unbalanced?

For there to be imbalance, there absolutely has to be some irregularity in either shape or in density. If the all sides of the die are symmetrical and the die is of uniform density it has to be in balance. If all corners are equally rounded the object is balanced. Rounding corners may introduce imbalance only if the corners are unevenly rounded.

 CrownAxe wrote:
Yes the article established that those 72 dice were unbalanced. But how is testing 72 dice out of a million proof that all million dice are unbalanced?

You do realize that the people who provide statistics for what people in one area think, don't actually ask each person in that area, right? Hell, they don't even ask the vast majority of those people. If you saw something that said that 52% of Americans believe that (thing) is real, they probably asked only about a thousand people in America in total before saying that. For a ccountry of, what, three hundred million?

72 dice ALL being unbalanced is 100% of the 72 dice tested. Tell me how many dice need to be unbalanced across a million dice boxes for buying two random boxes of dice off a shelf to give you two boxes where all the dice are unbalanced to be a likely possibility of random chance.

I'll give you a hint. The odds of these 72 dice being the ONLY unbalanced dice out there was incredibly low. There were and are plenty more like them out there.

Actually since they all came from the same box it would be odd if they weren't all the same, Being from the same package means they came from the same single production run and if that production run had an error in it they would all have it.

The is why the sample is too small. An single production run having an error is bound to happen (human error after all). It's no indicitive of all dice as a whole.

 Crimson wrote:
For there to be imbalance, there absolutely has to be some irregularity in either shape or in density. If the all sides of the die are symmetrical and the die is of uniform density it has to be in balance. If all corners are equally rounded the object is balanced. Rounding corners may introduce imbalance only if the corners are unevenly rounded.

READ. THE. ARTICLE. ITSELF.

They CUT the dice open. They found AIR BUBBLES and SEEDS inside ALL GW and Chessex dice. The air bubbles CORRELATED to HOW UNBALANCED each die WAS.

[quote=Pouncey 708883 9040449 c4dcc7aba1b12bbf127c68af0ea833bd.png

READ. THE. ARTICLE. ITSELF.

They CUT the dice open. They found AIR BUBBLES and SEEDS inside ALL GW and Chessex dice. The air bubbles CORRELATED to HOW UNBALANCED each die WAS.

I did. Yes, those are exactly the sort of things that can cause imbalance. Rounded corners are not.

 CrownAxe wrote:
Actually since they all came from the same box it would be odd if they weren't all the same, Being from the same package means they came from the same single production run and if that production run had an error in it they would all have it.

They're NOT all the same. They rolled different distributions of 1s.

The is why the sample is too small. An single production run having an error is bound to happen (human error after all). It's no indicitive of all dice as a whole.

A single production run does not produce only one box of dice.

HOW MANY dice are you saying were in this production run, exactly?

How many dice are you saying are like this?


Automatically Appended Next Post:

 Crimson wrote:
[quote=Pouncey 708883 9040449 c4dcc7aba1b12bbf127c68af0ea833bd.png

READ. THE. ARTICLE. ITSELF.

They CUT the dice open. They found AIR BUBBLES and SEEDS inside ALL GW and Chessex dice. The air bubbles CORRELATED to HOW UNBALANCED each die WAS.

I did. Yes, those are exactly the sort of things that can cause imbalance. Rounded corners are not.

Please quote for me the part of the article where he attempted to explain why the rounded corners caused them to be unbalanced while having square corners made them not unbalanced.

 Pouncey wrote:

 CrownAxe wrote:
Actually since they all came from the same box it would be odd if they weren't all the same, Being from the same package means they came from the same single production run and if that production run had an error in it they would all have it.

They're NOT all the same. They rolled different distributions of 1s.

All of which is within a reasonable margin of error of each other and all of which had the same skewed results. They are all functionally the same

The is why the sample is too small. An single production run having an error is bound to happen (human error after all). It's no indicitive of all dice as a whole.

A single production run does not produce only one box of dice.

HOW MANY dice are you saying were in this production run, exactly?

How many dice are you saying are like this?

There is no way to know with out contacting Chessex/GW and asking them how much dice they produce at a time and in total

But you can be sure that they didn't do all of there dice in a single production run. It's safe to say they've done a couple hundred at least and the study only tested one of them. They is not statistically conclussive

If we for the moment pretend that the part about melting more plastic to the dice and building the corners actually happened and it improved the balance, the reason would be that the corners were originally unevenly rounded (for some bizarre reason same way uneven on all the dice) and the new sharper corners were more symmetrical. The sharpness of the corners is not the thing that matters, it is the symmetry.

 CrownAxe wrote:

 Pouncey wrote:

 CrownAxe wrote:
Actually since they all came from the same box it would be odd if they weren't all the same, Being from the same package means they came from the same single production run and if that production run had an error in it they would all have it.

They're NOT all the same. They rolled different distributions of 1s.

All of which is within a reasonable margin of error of each other and all of which had the same skewed results. They are all functionally the same

For the love of god.

You should probably know something about probability if you're gonna be having this discussion with me. Like, anything at all.

The is why the sample is too small. An single production run having an error is bound to happen (human error after all). It's no indicitive of all dice as a whole.

A single production run does not produce only one box of dice.

HOW MANY dice are you saying were in this production run, exactly?

How many dice are you saying are like this?

There is no way to know with out contacting Chessex/GW and asking them how much dice they produce at a time and in total

But you can be sure that they didn't do all of there dice in a single production run. It's safe to say they've done a couple hundred at least and the study only tested one of them. They is not statistically conclussive

Then PHONE THEM AND ASK.


Automatically Appended Next Post:

 Crimson wrote:
If we for the moment pretend that the part about melting more plastic to the dice and building the corners actually happened and it improved the balance, the reason would be that the corners were originally unevenly rounded (for some bizarre reason same way uneven on all the dice) and the new sharper corners were more symmetrical. The sharpness of the corners is not the thing that matters, it is the symmetry.

Pretend?

What does the article tell you the guy had at his disposal in terms of lab equipment near the start of the article?

This is why I say you didn't even read the article. You CLEARLY didn't. You ask about things the article already answered.

READ THE DAMNED THING.

 Pouncey wrote:

 CrownAxe wrote:

 Pouncey wrote:

 CrownAxe wrote:
Actually since they all came from the same box it would be odd if they weren't all the same, Being from the same package means they came from the same single production run and if that production run had an error in it they would all have it.

They're NOT all the same. They rolled different distributions of 1s.

All of which is within a reasonable margin of error of each other and all of which had the same skewed results. They are all functionally the same

For the love of god.

You should probably know something about probability if you're gonna be having this discussion with me. Like, anything at all.

I have a year's worth of statistics that i studied when i was in college. I pretty sure that's more then what' you've studied.

The fact that you think this single study is conclusive show's that you don't actually understand what you are talking about. Which considering how our discussion on telepathy went this wouldn't be the first time.

The is why the sample is too small. An single production run having an error is bound to happen (human error after all). It's no indicitive of all dice as a whole.

A single production run does not produce only one box of dice.

HOW MANY dice are you saying were in this production run, exactly?

How many dice are you saying are like this?

There is no way to know with out contacting Chessex/GW and asking them how much dice they produce at a time and in total

But you can be sure that they didn't do all of there dice in a single production run. It's safe to say they've done a couple hundred at least and the study only tested one of them. They is not statistically conclussive

Then PHONE THEM AND ASK.

I'm not the one trying to claim that a single study with such a small sample size is conclusive

I read it, I don't believe it. I find it unlikely that any of these tests actually happened.

Actually since they all came from the same box it would be odd if they weren't all the same, Being from the same package means they came from the same single production run and if that production run had an error in it they would all have it.



They're NOT all the same. They rolled different distributions of 1s.


All of which is within a reasonable margin of error of each other and all of which had the same skewed results. They are all functionally the same



For the love of god.

You should probably know something about probability if you're gonna be having this discussion with me. Like, anything at all.


I have a year's worth of statistics that i studied when i was in college. I pretty sure that's more then what' you've studied.

The fact that you think this single study is conclusive show's that you don't actually understand what you are talking about. Which considering how our discussion on telepathy went this wouldn't be the first time.

Crown - I think Pounceys very interesting if a tiny touch eccentric. But do not get into probability weighted distribution curve debates with him.
Aint listening bro.

 Ratius wrote:

Actually since they all came from the same box it would be odd if they weren't all the same, Being from the same package means they came from the same single production run and if that production run had an error in it they would all have it.



They're NOT all the same. They rolled different distributions of 1s.


All of which is within a reasonable margin of error of each other and all of which had the same skewed results. They are all functionally the same



For the love of god.

You should probably know something about probability if you're gonna be having this discussion with me. Like, anything at all.


I have a year's worth of statistics that i studied when i was in college. I pretty sure that's more then what' you've studied.

The fact that you think this single study is conclusive show's that you don't actually understand what you are talking about. Which considering how our discussion on telepathy went this wouldn't be the first time.

Crown - I think Pounceys very interesting if a tiny touch eccentric. But do not get into probability weighted distribution curve debates with him.
Aint listening bro.

Obviously. My knowledge of probability is mostly intuitive, it doesn't include any actual training.

But I'm dealing with you guys who are saying things like 29% of 72,000 rolls are 1s is equally as unlikely as 100% of 1,000,000 rolls being 6s, and even I know that's completely wrong just from the numbers you're comparing, while you guys seem to think it's perfectly okay to say and defend as a valid comparison!

So I know I don't know much, but I know YOU guys know WAY less than I do since you prove it by trying to defend statements like, "29% of 72,000 rolls are 1s is equally as unlikely as 100% of 1,000,000 rolls being 6s," after having said such an obviously absurd thing in the first place!

Honestly, even if I fully accept that just these 72 dice are imbalances and none others were, the only thing that meant those dice were used by the article and not every game I have ever played since 2006 is that he bought them and I didn't!


Automatically Appended Next Post:

 CrownAxe wrote:
I'm not the one trying to claim that a single study with such a small sample size is conclusive

Correct. The guy in the article is. So when he needed Chessex to answer a question about why they manufactured their dice a certain way, he picked up the phone and called them to ask.

Now realize that what I replied to there was you saying:

"There is no way to know with out contacting Chessex/GW and asking them how much dice they produce at a time and in total

But you can be sure that they didn't do all of there dice in a single production run. It's safe to say they've done a couple hundred at least and the study only tested one of them. They is not statistically conclussive"

That is something YOU are trying to prove. Not ME. Not the ARTICLE. YOU.

Also, if you want to compare number of dice rolled, do you want me to go a few pages earlier in this thread and find people arguing AGAINST the article who talk about rolling 10 dice 200 times?


Automatically Appended Next Post:

 CrownAxe wrote:
Which considering how our discussion on telepathy went this wouldn't be the first time.

You know when I stopped arguing that? When I was proved wrong.

I am one of the most honest people on this board. I don't lie, since I don't ever see the point of lying about things that don't matter anyways.

When I am proved wrong about something in a way I accept as meaning I am wrong, I stop saying the wrong things I am saying. I don't argue points for no reason. I don't say things I know are wrong.

And you, have not proved this article wrong in my eyes. You are saying things about probability that I know are wrong from being able to compare two different numbers to each other and coming to the conclusion they are incredibly different.

So I don't believe you know probability better than I do. You have not proved to me that the article is wrong, and often I know you haven't read large parts of it because you make references to things that are already covered in the article.

I am an intellectually honest person who doesn't believe in lying on the Internet. And I think you are wrong.


Automatically Appended Next Post:

 Crimson wrote:
I read it, I don't believe it. I find it unlikely that any of these tests actually happened.

"I don't believe you, so you're lying."

That is your argument to ignore 72,000 dice rolls that show GW and Chessex dice are poorly-made.

You are saying things about probability that I know are wrong from being able to compare two different numbers to each other and coming to the conclusion they are incredibly different.

I would say the same thing about you. You are saying things about probability that I know are wrong. And I know they're wrong because I studied statistics for a year. You just said yourself that your understanding of probability is mostly intuitive and doesn't include any actual training. How can't you truly say you know what you are talking about when you haven't actually learned about the subject at hand?

That's what I was referring to when i brought the telepathy debate. You yourself said that you and other people who don't fully understand the science of the rat experiment called it telepathy. That is the same thing here. You and people who don't understand statistics are calling this article proof. When it is not proof, just like the rat experiment is not telepathy.

Honestly, here's how you prove to me that the article is probably lying. I will prove it to you without even mentioning the results at all.

Why does this article involve 36 GW dice and not 25? GW dice are sold in cubes of 27, 2 of which are scatter dice. What dice were added to this to give him an extra 11 GW dice? What is so special about having 36 of each that GW dice were either added to the box he grabbed, or excluded from the 2 boxes he grabbed? Why not just include all of the GW dice you bought to test this and no more?

And the only answer I can come up with is because human brains perceive evenness as good, so he was willing to have 36 of each be the results he was sharing mean the results would be more acceptable. And that's a terrible way to do any sort of analysis that you are excluding or adding things to the test, just to make it happen. So you're probably full of it about something.

You DON'T prove something is wrong by just saying you don't believe it, so it's wrong. You DON'T prove dice being unbalanced is impossible by saying that it would require those dice to be unbalanced physically because yes, if a die were rolling extra 1s, it would have to be unbalanced physically since a die's physical characteristics affect how it rolls and changing physical things about it is the only way to unbalance it.

You DON'T prove the article is wrong by saying the manufacturing process would have to be defective when GW made Finecast, one of the most defective products known to wargaming.

You DO prove to me that the article is wrong by pointing out things about it that mean it cannot be true.

You DON'T prove to me that the article is wrong by offering your word that fewer rolls were being made than it said it did and those fewer rolls mean it's wrong as proof it is wrong.

And frankly, I'm wondering why you guys started arguing about it again after a page or two ago when I said its results were probably a fluke that meant GW dice were actually balanced like you were saying they were. Why did you start arguing with me again after I said you were right? Do you just like to argue or something?

I'm not trying to prove the article is lying. In fact I think those 72 dice are in fact unbalanced.

I'm saying that single article is not enough evidence to claim that Chessex and GW dice are on the whole unbalanced. A basic understanding on manufacturing and producing goods is enough to prove that it's sample size is too small to be conclusive.

In science you can't prove any hypothesis with a single study, you have to have multiple people repeat that experiment to verify those findings. That has not be done with this article.