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.

 Peregrine wrote:

This is not necessarily true. It's possible that something about the manufacturing process causes distortion/air bubbles/whatever to form in a consistent manner, which would have the bias always favor the same side. For example, maybe the 1 side is at the top of the mold, and any air bubbles that form rise a bit before the plastic cools enough to make their location permanent. It wouldn't be a thing without precedent, cast resin models often have bubbles/mold slip/etc in the same place across multiple casts because that's where that particular mold doesn't work quite right.

Valid point. Still, the systematic imbalance as large as described in the article is implausible. I doubt you would get such massive imbalance without intentionally weighting the dice.

 Crimson wrote:

 Peregrine wrote:

This is not necessarily true. It's possible that something about the manufacturing process causes distortion/air bubbles/whatever to form in a consistent manner, which would have the bias always favor the same side. For example, maybe the 1 side is at the top of the mold, and any air bubbles that form rise a bit before the plastic cools enough to make their location permanent. It wouldn't be a thing without precedent, cast resin models often have bubbles/mold slip/etc in the same place across multiple casts because that's where that particular mold doesn't work quite right.

Valid point. Still, the systematic imbalance as large as described in the article is implausible. I doubt you would get such massive imbalance without intentionally weighting the dice.

I thought the article described what they found when they physically dissected all of their dice to see what was inside them.

Honestly I thought when that guy you just quoted was talking about air bubbles he wasn't speaking hypothetically about those air bubbles being inside GW dice. Because when you cut open GW dice, what you find are air bubbles and plastic seeds.

Also they were able to somehow correlate each die's air bubbles to how much more often than average each die rolled 1s.

And they don't have to really intentionally weight the dice to get imbalances like this. They just have to shave off the corners to save on material. If you're able to rebuild your GW dice's shaved-off corners, you get dice that roll pretty darned fairly, with about 19% 1s.

Also, shaving off the corners make these kinds of imbalances possible. The air bubbles in each die seem to directly affect how bad the imbalance is.

I've bought cheap dice like this ones


They cost like 4-5 cents per one, can be bought en masse for ~3 cents online. They've been pretty ok.

 koooaei wrote:
I've bought cheap dice like this ones

Spoiler:

They cost like 4-5 cents per one, can be bought en masse for 2-3 cents online. They've been pretty ok.

Yeah, that's fine.

I mean, even the article that claimed to have proved that GW dice roll somewhere between 23% and 33% 1s didn't say you should never use GW dice, because a game could easily be designed to accommodate that kind of imbalance. Like, for example, having a thing where rolling low is sometimes good (Leadership tests of any sort, also characteristic tests).

I think its main suggestion was that you should use the same dice for everything, and call out people who use casino dice normally and GW dice for Leadership tests as cheaters.

 Crimson wrote:

 Pouncey wrote:

Why shouldn't I?

The results of the study are pretty incredible and anyone can test themselves whether their dice behave like this test claims (mine certainly don't.)
Furthermore, there is a clear motive for this to be a hoax, the aricle directly links to a store selling 'perfect dice.' (Not saying that this necessarily is a hoax rather just an experiment with some systemic flaw, but that possibility certainly comes to mind.)

Yeah, I did my own 600 throw test (with Chessex dice) after reading the article. Sure that's a pretty small sample, but I marked down the results after every 100 throws, and the distribution went gradually towards what you would expect (16.67 etc). In the end, '1' was the LEAST common result.

It is of course POSSIBLE that had I continued the test, the '1' suddenly would have become much more common, but quite unlikely.
So yeah, the article was a hoax, or at very least completely irrelevant to an actual wargamer.

ps. I play Deathwing so I'd notice if every 4th roll was '1'.

Backfire wrote:

 Crimson wrote:

 Pouncey wrote:

Why shouldn't I?

The results of the study are pretty incredible and anyone can test themselves whether their dice behave like this test claims (mine certainly don't.)
Furthermore, there is a clear motive for this to be a hoax, the aricle directly links to a store selling 'perfect dice.' (Not saying that this necessarily is a hoax rather just an experiment with some systemic flaw, but that possibility certainly comes to mind.)

Yeah, I did my own 600 throw test (with Chessex dice) after reading the article. Sure that's a pretty small sample, but I marked down the results after every 100 throws, and the distribution went gradually towards what you would expect (16.67 etc). In the end, '1' was the LEAST common result.

It is of course POSSIBLE that had I continued the test, the '1' suddenly would have become much more common, but quite unlikely.
So yeah, the article was a hoax, or at very least completely irrelevant to an actual wargamer.

ps. I play Deathwing so I'd notice if every 4th roll was '1'.

Wow.

You only wrote down the results after every 100 rolls?

You have an amazing memory to be able to remember 100 numbers each ranging from 1-6 before you can't remember anymore and have to write stuff down.

Like, seriously, I'm so amazed I literally don't believe it. And I have a pretty good memory myself.

 Pouncey wrote:

Backfire wrote:

 Crimson wrote:

 Pouncey wrote:

Why shouldn't I?

The results of the study are pretty incredible and anyone can test themselves whether their dice behave like this test claims (mine certainly don't.)
Furthermore, there is a clear motive for this to be a hoax, the aricle directly links to a store selling 'perfect dice.' (Not saying that this necessarily is a hoax rather just an experiment with some systemic flaw, but that possibility certainly comes to mind.)

Yeah, I did my own 600 throw test (with Chessex dice) after reading the article. Sure that's a pretty small sample, but I marked down the results after every 100 throws, and the distribution went gradually towards what you would expect (16.67 etc). In the end, '1' was the LEAST common result.

It is of course POSSIBLE that had I continued the test, the '1' suddenly would have become much more common, but quite unlikely.
So yeah, the article was a hoax, or at very least completely irrelevant to an actual wargamer.

ps. I play Deathwing so I'd notice if every 4th roll was '1'.

Wow.

You only wrote down the results after every 100 rolls?

You have an amazing memory to be able to remember 100 numbers each ranging from 1-6 before you can't remember anymore and have to write stuff down.

Like, seriously, I'm so amazed I literally don't believe it. And I have a pretty good memory myself.

Or...

It's entirely possible that the tester in this case was only counting results of "1", in which case, they only had to remember how many 1's were rolled. Everything else is "not 1" in this case. They are not stating that the dice aren't weighted or improper, but that a result of "1" verses "not 1" fits an approximation of a standard distribution.

This thread should be locked, as it really is just going in circles. This is the circle I've seen:

"The dice are rigged as per study."
"Study is no good, for reasons."
"That okay, study still generally good."
"Me do study that is generally good."
"Your study no good though, for reasons."
"But you say reasons still generally good!"
"Me be very specific now. The dice are rigged as per study."
"But study no good! We've said this already..."

No, its a frustrating thread because dice have nothing to do with GWs rules writing abilities - whether you view them as good or bad.
You can still build that deathstar with 2+++++ and roll a one. Pooof, goodbye deathstar.

You fail tactically if you roll a 1.

Aw hell bring back 1+ saves so.

 Yarium wrote:

 Pouncey wrote:

Backfire wrote:

 Crimson wrote:

 Pouncey wrote:

Why shouldn't I?

The results of the study are pretty incredible and anyone can test themselves whether their dice behave like this test claims (mine certainly don't.)
Furthermore, there is a clear motive for this to be a hoax, the aricle directly links to a store selling 'perfect dice.' (Not saying that this necessarily is a hoax rather just an experiment with some systemic flaw, but that possibility certainly comes to mind.)

Yeah, I did my own 600 throw test (with Chessex dice) after reading the article. Sure that's a pretty small sample, but I marked down the results after every 100 throws, and the distribution went gradually towards what you would expect (16.67 etc). In the end, '1' was the LEAST common result.

It is of course POSSIBLE that had I continued the test, the '1' suddenly would have become much more common, but quite unlikely.
So yeah, the article was a hoax, or at very least completely irrelevant to an actual wargamer.

ps. I play Deathwing so I'd notice if every 4th roll was '1'.

Wow.

You only wrote down the results after every 100 rolls?

You have an amazing memory to be able to remember 100 numbers each ranging from 1-6 before you can't remember anymore and have to write stuff down.

Like, seriously, I'm so amazed I literally don't believe it. And I have a pretty good memory myself.

Or...

It's entirely possible that the tester in this case was only counting results of "1", in which case, they only had to remember how many 1's were rolled. Everything else is "not 1" in this case. They are not stating that the dice aren't weighted or improper, but that a result of "1" verses "not 1" fits an approximation of a standard distribution.

This thread should be locked, as it really is just going in circles. This is the circle I've seen:

"The dice are rigged as per study."
"Study is no good, for reasons."
"That okay, study still generally good."
"Me do study that is generally good."
"Your study no good though, for reasons."
"But you say reasons still generally good!"
"Me be very specific now. The dice are rigged as per study."
"But study no good! We've said this already..."

I'll try to break the cycle with something new then.

I'm completely uneducated in physics, but it seems to me that something that might have an effect of some sort on the result of a dice roll that is determined by physics, is having 9 identical dice in motion nearby during that roll that that single die might bounce off of at some point during the roll.


Automatically Appended Next Post:

 Ratius wrote:
No, its a frustrating thread because dice have nothing to do with GWs rules writing abilities - whether you view them as good or bad.
You can still build that deathstar with 2+++++ and roll a one. Pooof, goodbye deathstar.

That's the other thing. The OP was never even really about the dice thing at all. I wasn't expecting anyone who read this thread to even treat it as something they didn't already know about and accept as true, much less argue with.

The dice thing I mentioned in the OP was just background information to the actual thing I was wanting to say that relates to the title.

Fair enough but mentioning dice and your thought experiments to control them was only going to have the topic go one way......

Maybe next time just ask "whats wrong with 40ks ruleset?"

 Pouncey wrote:
Many years ago, probably during 5th edition, I learned of that scientific study where a Professor effectively proved that GW and Chessex dice are inherently unfair through 36,000 rolls of dice from 4 different brands.

Naturally, I set out to acquire some better dice for my games. I went to a local gaming store to order some dice from a catalog, and I ended up talking to the owner of the store working the counter about the study that led me to his store in the first place. He'd also heard of it, so we talked about it for a bit, and when I tried to buy some of those board game dice that were more fair than GW dice but not as expensive as casino dice, he asked if the dice were for Warhammer 40k. I said yes, because it was the only tabletop game I played of any sort. So he replied to me, something very similar to, "You don't need to buy new dice. If you're playing Warhammer 40k, you're not the kind of person who cares enough about game balance to need fair dice."

That's a pretty rough situation for your game's balance when people are saying things like that to convince customers not to buy stuff from them.

I didn't even argue or disagree, and immediately picked up 2 sets of Chessex dice off the shelves that looked pretty and bought them, since I like pretty dice.

I have never regretted my decision to buy inherently unfair dice for WH40k that have screwed me over a lot in my games by making my melta weapons roll 1s up to twice as often as they should be, over insisting that my dice be mostly fair when I play WH40k and buying some standard board game dice instead.

Also, that was during 5th edition. Things have gotten a LOT worse since then in terms of game balance.

1. Citation required or it never happened.

2. Have you considered chess? I mean it's not completely balanced either (White always go first) but the price of entry is much lower.

 koooaei wrote:
You fail tactically if you roll a 1.

I wonder how many games are won by rolling that tactical 6 on a stomp or D table.

 MarsNZ wrote:

 koooaei wrote:
You fail tactically if you roll a 1.

I wonder how many games are won by rolling that tactical 6 on a stomp or D table.

Around 1/6. If your dice are evenly weighted inside.

 Ratius wrote:
Fair enough but mentioning dice and your thought experiments to control them was only going to have the topic go one way......

Maybe next time just ask "whats wrong with 40ks ruleset?"

Yeah, I usually succeed in avoiding mentioning my attempts to control dice rolls over a decade ago on this forum I started posting on 5 years ago. Because I never really got the impression that people here would be open to the idea that that was possible and would probably just call me a liar the whole time.

I wasn't wrong, was I?

This time I mentioned it only because I misunderstood what I was replying to while I'd been awake for 3 days straight and thought someone wrote a paragraph-long post seriously discussing the possibility I'm not a liar, in response to a very minor mention I thought I remembered slipping into a post somewhere that I thought everyone would miss. I should've deleted that post entirely when I re-read what I was replying to and saw it was about something else, before anyone had responded to it, instead of editing it with a statement about how I misunderstood what I was replying to.

Also, I didn't think mentioning the thing about GW dice rolling tons more 1s than they should would be a big deal. The last I read on this forum, pretty much everyone was taking it to be completely proven. And I thought it was so unquestionably accepted around here that the only reason I mentioned it at all was because reading that study is what set me off to the store, where the thing the store owner told me was what I considered to be, and let me quote the title here: "The Ultimate Condemnation of WH40k's Rules".


Automatically Appended Next Post:

Diogenes wrote:

 Pouncey wrote:
Many years ago, probably during 5th edition, I learned of that scientific study where a Professor effectively proved that GW and Chessex dice are inherently unfair through 36,000 rolls of dice from 4 different brands.

Naturally, I set out to acquire some better dice for my games. I went to a local gaming store to order some dice from a catalog, and I ended up talking to the owner of the store working the counter about the study that led me to his store in the first place. He'd also heard of it, so we talked about it for a bit, and when I tried to buy some of those board game dice that were more fair than GW dice but not as expensive as casino dice, he asked if the dice were for Warhammer 40k. I said yes, because it was the only tabletop game I played of any sort. So he replied to me, something very similar to, "You don't need to buy new dice. If you're playing Warhammer 40k, you're not the kind of person who cares enough about game balance to need fair dice."

That's a pretty rough situation for your game's balance when people are saying things like that to convince customers not to buy stuff from them.

I didn't even argue or disagree, and immediately picked up 2 sets of Chessex dice off the shelves that looked pretty and bought them, since I like pretty dice.

I have never regretted my decision to buy inherently unfair dice for WH40k that have screwed me over a lot in my games by making my melta weapons roll 1s up to twice as often as they should be, over insisting that my dice be mostly fair when I play WH40k and buying some standard board game dice instead.

Also, that was during 5th edition. Things have gotten a LOT worse since then in terms of game balance.

1. Citation required or it never happened.

I'll get right on proving that one thing happened one time in a small store 5 years ago that I wasn't expecting to happen to the point I'd've brought a video camera to record it or have enough time to whip out a cell phone to record it happening before it was already over.

So basically, citation will not be incoming. Citation is not possible, unless you wanna track down the guy working the desk at some point around 5 years ago at a store called Fandom II in downtown Ottawa, give him a ring and ask, "Hey, did you say this one thing to this one guy who was in your store one time five years ago who you've never seen since and probably don't remember?"

2. Have you considered chess? I mean it's not completely balanced either (White always go first) but the price of entry is much lower.

Personally I decided to play video games instead of tabletop games.

I have 15,000 hours spent playing WoW since 2007. Also I was playing WH40k tabletop pretty much every week from 2006 up until 2011. Also I have over 50,000 posts on the WoW forums.

 Pouncey wrote:

You only wrote down the results after every 100 rolls?

You have an amazing memory to be able to remember 100 numbers each ranging from 1-6 before you can't remember anymore and have to write stuff down.

Like, seriously, I'm so amazed I literally don't believe it. And I have a pretty good memory myself.

No, I was unclear. What I did was that I marked down the result of each throw, and when I reached 100 throws, I'd add them up and calculate distribution. At first '100' there were some big differences in percentiles but they smoothed over as the trial progressed.

Backfire wrote:

 Pouncey wrote:

You only wrote down the results after every 100 rolls?

You have an amazing memory to be able to remember 100 numbers each ranging from 1-6 before you can't remember anymore and have to write stuff down.

Like, seriously, I'm so amazed I literally don't believe it. And I have a pretty good memory myself.

No, I was unclear. What I did was that I marked down the result of each throw, and when I reached 100 throws, I'd add them up and calculate distribution. At first '100' there were some big differences in percentiles but they smoothed over as the trial progressed.

Why calculate distribution every 100 throws instead of after the end total of 600 throws?

That's not really snark, I'm legitimately wondering if there's a reason to do so, since there could easily be a totally valid reason I don't know about.

About it mattering whether you roll dice separately or together, I doubt it would. Sure, if you roll them together the dice will bounce against each other but that is completely random too.

But, I tested that too.

500 rolls, ten dice used, but each rolled separately. Exactly 16% of ones.

The article is complete bolllocks. The real reason why some people think that their dice roll more ones is the confirmation bias. When I did my 2000 dice test, rolling ten dice at once, I manged to twice to roll five ones at once. Now, were that happen when I was rolling saves for my terminators I would remember that for a long time and might end up thinking that my dice are cursed or something. But that's random for you, sometimes these things happen.

As for the dice not being the original point, there was no original point. There was a story about some guy who thought that 40K rules were bad for unspecified reasons. So? And then there was a claim about having superpowers that allow controlling dice and then rambling about psychic rodents...

Also, you wanna know why I question you even think your thing where you rolled a die 600 times and recorded each result, and because your results after 600 rolls evened out to a fairly even distribution of results, means that the article claiming to show that GW dice roll around 26% 1s on average is wrong since it'd average out to 16.7% or so over long periods of time with hundreds of rolls?

Because it means you probably didn't even read the article at all. The article claimed they rolled each die 1,000 times. If 600 rolls is enough to even out the results, the article should've reached that point on every single die with 400 fewer recorded rolls on each die they tested. They tested 144 dice by rolling them 1,000 times each.

That article claims to have rolled each die they tested 400 more times than the 600 times you tested only a single die.


Automatically Appended Next Post:

 Crimson wrote:
About it mattering whether you roll dice separately or together, I doubt it would. Sure, if you roll them together the dice will bounce against each other but that is completely random too.

But, I tested that too.

500 rolls, ten dice used, but each rolled separately. Exactly 16% of ones.

The article is complete bolllocks. The real reason why some people think that their dice roll more ones is the confirmation bias. When I did my 2000 dice test, rolling ten dice at once, I manged to twice to roll five ones at once. Now, were that happen when I was rolling saves for my terminators I would remember that for a long time and might end up thinking that my dice are cursed or something. But that's random for you, sometimes these things happen.

As for the dice not being the original point, there was no original point. There was a story about some guy who thought that 40K rules were bad for unspecified reasons. So? And then there was a claim about having superpowers that allow controlling dice and then rambling about psychic rodents...

Sharing our random WH40k-related stories is just something we do around here on Dakka.

 Pouncey wrote:
I'll try to break the cycle with something new then.

I'm completely uneducated in physics, but it seems to me that something that might have an effect of some sort on the result of a dice roll that is determined by physics, is having 9 identical dice in motion nearby during that roll that that single die might bounce off of at some point during the roll.

I'll try to help you then. Yes, they will definitely influence the roll, but not in any meaningful way. Let's pretend you have two completely random real numbers that are truly random, and you multiply them together. Is the product more random, as random, or less random than either of the original random numbers? The answer is; exactly as random. I could also add the results together, subtract them, divide them, etc. So long as both of them are truly random, any possible outcome of either number is possible, which includes the outcome of both of them multiplied together. Did the act of them interacting influence the outcome? Yes. But did it do so in a meaningful way? No. I did not achieve a "more random" result.

Same happens for a die throw. The die is already interacting with an untold number of random factors; the sides of your hands, the table, the wall, the air current, the air temperature, the influence of Jupiter's incredibly tiny gravity at this distance, the slightly larger influence of your opponent's gravity at this distance... They're all random and influencing the result. Having it bounce a few times (or, in fact, a random number of times) off of some other dice will not influence the result in a meaningful way. The result of a die bouncing on the table is just as random as the result of a die bouncing on the table, and bouncing off another die.

 Yarium wrote:

 Pouncey wrote:
I'll try to break the cycle with something new then.

I'm completely uneducated in physics, but it seems to me that something that might have an effect of some sort on the result of a dice roll that is determined by physics, is having 9 identical dice in motion nearby during that roll that that single die might bounce off of at some point during the roll.

I'll try to help you then. Yes, they will definitely influence the roll, but not in any meaningful way. Let's pretend you have two completely random real numbers that are truly random, and you multiply them together. Is the product more random, as random, or less random than either of the original random numbers? The answer is; exactly as random. I could also add the results together, subtract them, divide them, etc. So long as both of them as truly random, any possible outcome of either number is possible, which includes the outcome of both of them multiplied together. Did the act of them interacting influence the outcome? Yes. But did it do so in a meaningful way? No. I did not achieve a "more random" result.

Same happens for a die throw. The die is already interacting with an untold number of random factors; the sides of your hands, the table, the wall, the air current, the air temperature, the influence of Jupiter's incredibly tiny gravity at this distance, the slightly larger influence of your opponent's gravity at this distance... They're all random and influencing the result. Having it bounce a few times (or, in fact, a random number of times) off of some other dice will not influence the result in a meaningful way. The result of a die bouncing on the table is just as random as the result of a die bouncing on the table, and bouncing off another die.

And the only way in which dice are random is in the human brain's inability to predict the result. The result you get is determined entirely by physics, and if you can influence a result controlled by physics by introducing more variables to the calculations, then you can alter the outcome. Having multiple dice bouncing against each other probably does make the results more even. The reason casinos require you to roll in a very specific way (two of the three requirements are that the dice bounce off the table and at last one wall, literally bouncing off of other objects) is because doing so results in a more even distribution of results, for casino dice that are already required to be perfectly balanced before they can leave the factory.

The thing you're discounting as a possible explanation for a non-even distribution of results in one case, and a non-even distribution in other cases, with the exact same dice, is already the reason that casinos require you do to things with dice that would require that explanation to be true to give you a reason to do them at all.

This thread makes me want to always blame my dice every time I lose. And now I have the proof! Thank you!

What if you rolled a dice in space and let it keep rolling until it reached the outer limits of our known Universe and was destroyed?
Would that be an unlucky 1 or a terrific 6?

 Ratius wrote:
What if you rolled a dice in space and let it keep rolling until it reached the outer limits of our known Universe and was destroyed?
Would that be an unlucky 1 or a terrific 6?

Well, for one thing, no human would care since the heat death of the universe would've happened already and humans will cease existing so long before that that the possibility that humans might still be around by then is ludicrous since even if we didn't get wiped out we'd've evolved into a species so different from humans that calling our descendants humans at that point is a dumb thing to say.

 Pouncey wrote:


And the only way in which dice are random is in the human brain's inability to predict the result.

Not necessarily. Many interpretations of quantum physics suggest that randomness is in fact real. Of course whether that results randomness on macro level is debatable.

The result you get is determined entirely by physics, and if you can influence a result controlled by physics by introducing more variables to the calculations, then you can alter the outcome. Having multiple dice bouncing against each other probably does make the results more even. The reason casinos require you to roll in a very specific way (two of the three requirements are that the dice bounce off the table and at last one wall, literally bouncing off of other objects) is because doing so results in a more even distribution of results, for casino dice that are already required to be perfectly balanced before they can leave the factory.

It is indeed possible to learn to somewhat control the results of small number of dice if you throw then carefully in a certain way and Casinos want to prevent people from doing that. But that's really not relevant unless people playing 40K (and the people who purportedly made the study.) were intentionally attempting to roll more ones.

 Crimson wrote:

 Pouncey wrote:


And the only way in which dice are random is in the human brain's inability to predict the result.

Not necessarily. Many interpretations of quantum physics suggest that randomness is in fact real. Of course whether that results randomness on macro level is debatable.

Quantum physics, by definition, applies only to things smaller than an atom. Dice are very noticeably larger than an atom. And you can't just scale up physics calculations from the quantum scale to the macro scale and expect to get the same results. Scaling up a model rocket the size of a toy that works fine, to a rocket the size of an ICBM you're expecting to put something into space, and expecting the physics to be the same that simply making it bigger means it should work just fine, is something that's already been proven wrong. If you want to see the results of that kind of thinking being put into practice with rockets the size of ICBMs, just look up the clip reel of early rocket designs failing incredibly spectacularly in giant explosions after failing to get off the launch pad, and other ways a rocket with incorrect physics can explode when it's test-launched. You've probably already seen it.

The result you get is determined entirely by physics, and if you can influence a result controlled by physics by introducing more variables to the calculations, then you can alter the outcome. Having multiple dice bouncing against each other probably does make the results more even. The reason casinos require you to roll in a very specific way (two of the three requirements are that the dice bounce off the table and at last one wall, literally bouncing off of other objects) is because doing so results in a more even distribution of results, for casino dice that are already required to be perfectly balanced before they can leave the factory.

It is indeed possible to learn to somewhat control the results of small number of dice if you throw then carefully in a certain way and Casinos want to prevent people from doing that. But that's really not relevant unless people playing 40K (and the people who purportedly made the study.) were intentionally attempting to roll more ones.

Setting aside the obvious tie-in to an earlier topic I don't want to discuss, the article's analysis of dice was done because people already believed that casino dice rolled 1s less often than GW dice that they were trying to cheat by using GW dice for stuff where they required a low result even though they also had enough casino dice to do everything else in the game.

Also, I roll single dice a lot as a Sisters of Battle player. And the results are pretty important. Like, figuring out how many shots my Exorcist gets that turn is controlled by 1d6 that can never be legitimately rolled alongside any other dice. Firing a BSS' only meltagun at a tank, also mean I only roll one die. Attempting an Act of Faith also requires me to roll one die. Situations where I roll 4-5 single d6s consecutively are something that comes up in pretty much every game, even when I'm rolling all the dice I'm allowed to roll at the time by the actual rules.

This popped up on my dakka ad feed just now. I kid you not.....

https://www.kickstarter.com/projects/drawlab/legendary-dice-throwers-the-evolution-of-dice-towe

 Ratius wrote:
This popped up on my dakka ad feed just now. I kid you not.....

https://www.kickstarter.com/projects/drawlab/legendary-dice-throwers-the-evolution-of-dice-towe

It's a fun way to roll dice being marketed as simply being more fun than rolling them with your hand.

I didn't read the whole thing though. Did they actually suggest it might result in your dice rolls having a different outcome at any point?

 Pouncey wrote:
And the only way in which dice are random is in the human brain's inability to predict the result. The result you get is determined entirely by physics, and if you can influence a result controlled by physics by introducing more variables to the calculations, then you can alter the outcome. Having multiple dice bouncing against each other probably does make the results more even. The reason casinos require you to roll in a very specific way (two of the three requirements are that the dice bounce off the table and at last one wall, literally bouncing off of other objects) is because doing so results in a more even distribution of results, for casino dice that are already required to be perfectly balanced before they can leave the factory.

The thing you're discounting as a possible explanation for a non-even distribution of results in one case, and a non-even distribution in other cases, with the exact same dice, is already the reason that casinos require you do to things with dice that would require that explanation to be true to give you a reason to do them at all.

By that extension, it is impossible for any roll to be random enough. Guess you can't play 40k then, since each and every outcome is knowable if you had perfect knowledge. Except, you don't have perfect knowledge, which is true even without bringing up quantum physics! At what point is something "random enough"? For a casino, it's what you have pointed out; it must go past a line, bounce off the table, and hit the wall (the walls of which are a special shape). At minimum this means 2 "bounces". This is done to ensure that someone can not learn to throw the dice in a specific way to achieve a specific outcome. Those wall liners are also changed regularly to change the axis at which the same roll will achieve the same result.

If two bounces is good enough for a casino, then I'm sure it's good enough for this test, and additional knocking into things will only make the results MORE accurate rather than less.