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.

Personally, my opinions are based off of interactions of people at larger highly competitive events and people in my area.

I have had a bone to pick with people who use a pair of special dice for special rolls for over a decade waaaaay back when the only internet board was Portent.

CASINO DICE have no business in wargaming. They have no business being rolled on a wargaming surface, they need a bank, people can easily learn basic dice-rolling techniques and influence the outcome.

What has happened in wargaming is one person is a dice cheater. He cheats like a MFer. People *SEE* he prays to the dice gods and says 'ha! these are my lucky dice! I set them aside for special rolls! NON cheaters emulate behavior out of respect to the dice gods or ignorance.

Adepticon or 'ard boyz happens and all off a sudden, they are playing someone... "Wait a second... this is an important invulnerable save... let me roll this with my lucky dice."

Out come the square-edged balanced dice when all game he has been using chessex rounded wargaming dice. Without knowing it this player has basically changed the odds and given himself an unfair advantage. He needs a high number rolled and he changes knowingly or unknowingly to a dice that rolls less 1s.

The issue isn't with the ignorant people who don't know better. It is with the people who *DO* know better and go the extra distance to buy an advantage over others and then gloat how they deserve the advantage because they spent 40$ getting balanced dice and their opponent doesn't deserve it.

These people do exist. I have met them face to face. I disagree with them because IMHO a good sport playing a dice game wants no undue advantage outside winning through skill. If balanced dice is important, bring some, and share them so the whole game is statistically correct not just your actions.

I like tourneys who mandate Dice sharing, and in a competitive environment, I will apply scrutiny to segmented dice in games.

I don't think these opinions are unreasonable or stupid... Not sure what Stelek has to do with them either... I accept that rounded dice probably roll more 1s... I see why people would want to use different dice. I disagree that those who buy better dice deserve an in-game advantage.

nkelsch wrote:
I like tourneys who mandate Dice sharing, and in a competitive environment, I will apply scrutiny to segmented dice in games.

I agree with you on the whole "special dice" thing, but this just seems uncalled for. I don't want other people touching my dice, you probably don't want me touching your dice either. Call me crazy man, but I just couldn't get behind this even if I thought they were cheating. My dice are my dice, no touching!

People can touch my dice, just the more the get to touch them the easier it is for them to steal some without getting noticed.

RiTides wrote:
I also never used the phrase "dumb luck"... and the percentages I was referring to were the 97.5% / 99.999% that you included in your previous post (as the chances that the dice were unbalanced).

Hehe, we've reached the point in the debate where reality rewrite begins.

RiTides wrote:That explanation certainly could be true, but if I rolled thousands of dice, I'd be more shocked if I actually got completely even (16.6%) results for each face.

If it's random, shouldn't some faces end up with higher percentages? You can do a much smaller scale test and get very skewed results... how many dice do you need to roll until you get the "perfect" random result?

Imho, if the result is "perfectly random" then it's no longer random at least as far as the realm of rolling dice.

Just my $0.02, of course!

You most certainly did suggest that the results could have been just because of 'dumb luck.' As for the 97.5% and 99.9999% that's trivial math from the standard deviation, which I gave you the formula for. It's a ridiculously simple calculation, assuming normal distribution (which is a fine assumption here). Your results for standardized dice will not skew badly enough to break those confidence intervals.

Right up there at the top of the page.

No no no... The whole reason this thread has gone out of control is the original poster. Here is someone lamenting that something made by the hands of man is not completely perfect, so that when he is rolling his crucial save, the dice fail him. So, instead of doing what some people do here by blaming the rules, he blames the dice. So he plans to spend an inordinate amount of money to have precisely made random number generators.

It's a freaking GAME people! Everyone gets cursed by their dice from time-to-time, but the majority of us get past it. And when I go to this website, looking for some sage advice, or cool painting I get "why can't this system of random numbers turn up 6's more often!" Go master solitaire, then at least the only thing you would have to blame is your precision pack of cards that you over-spent on!

Everyone should buy a couple of dice cubes from GW. They are only £4 for 25 dice including the artillery die.

If you go to an event where nearly everyone is playing with them, it doesn't matter how much the dice get swapped around providing you come home with 25 dice including the artillery die.

I'm not sure why you insist on arguing against a straw-man that no one is espousing. You'd be much more convincing if you'd respond to what people are saying, intsead of going off about Stelek, "dumb luck", and other things that only you have typed...

[quote=RisingPhoenixYou most certainly did suggest that the results could have been just because of 'dumb luck.' As for the 97.5% and 99.9999% that's trivial math from the standard deviation, which I gave you the formula for. It's a ridiculously simple calculation, assuming normal distribution (which is a fine assumption here). Your results for standardized dice will not skew badly enough to break those confidence intervals.

Right up there at the top of the page.

Catyrpelius wrote:
What your refering to is Sigma specifacally Three Sigma, which states that 99.99% of all values rolled will fall within 3 standard deveations from the mean value. To use a quote from one of my favorite movies...."I don't think it meens what you think it meens."

No, three sigma is 99.7%. The certainty of the stated results would be beyond 6 sigma. The standard deviation of a result on a dice roll is trivially (N*P*(1-P))^0.5 (again...) and 29% is many many more than three standard deviations outside of expected results - it's about 10, in point of fact.

It does mean EXACTLY what I think it means. Maybe you should spend more time studying and less watching movies? A small suggestion

Modquisition on. All parties need to refresh their recollections with what Rule #1 entails. BE POLITE. Consider this a public warning to all posters on this thread. Posts after this warning will be viewed strictly and will receive an extraordinary punishment if found to be violative.

Seriously:

OMG it's a game! Buy a big ing box of dice and roll them.

This is interesting.

http://dicephysics.info/

Every time this thread comes up. I wonder, why is it okay to spend thousands on an army. But when someone suggests dropping $50 into dice so that their army plays correctly, people look at them like they are crazy.

Because I don't want to spend 1000s on my army, but I do my best to get them as cheap as possible. And dice are just generally cheaper.

Oshova

From KK's link:
http://dicephysics.info/

The Physics of Dice wrote:Low Quality Dice

A variety of dice were purchased from toy stores and variety stores in Kelowna, Canada. One group of dice came in a package priced at US $0.10 each. One was selected at random and rolled 21543 times. The table was covered with 21 ounce mali cloth (wool felt) as on a casino craps table. The results were as follows:

1: 2895
2: 4214
3: 3389
4: 3347
5: 4383
6: 3315

This is why I'm so skeptical of the Dakka article that says 1's are rolled more often by cheap dice. Here's an outside source testing cheap dice, and although it's only a single (group of dice? hard to tell from the wording), at least we have the totals for each number it rolled (not just the 1's).

Here's the percentages that that works out to:

1: 13.4%
2: 19.6%
3: 15.7%
4: 15.5%
5: 20.3%
6: 15.4%

Which supports my stance that cheap dice can be weighted so that they don't roll average, but not necessarily rolling more 1's. These dice apparently roll 5's and 2's like mad and 1's the Least of any number!

Nice link... will bookmark it for the next time this comes up

I'm not sure why you think that a set of dice will be skewed towards different numbers with equal likelyhood. It is much more likely that dice made with a particular manufacturing process will all be skewed towards the same number particularly within the same batch. It's pretty clear from the warseer study from ASU that the particular dice he was testing were heavily skewed towards ones.

Have you got a link for the ASU study?

RiTides wrote:From KK's link:
http://dicephysics.info/

The Physics of Dice wrote:Low Quality Dice

A variety of dice were purchased from toy stores and variety stores in Kelowna, Canada. One group of dice came in a package priced at US $0.10 each. One was selected at random and rolled 21543 times. The table was covered with 21 ounce mali cloth (wool felt) as on a casino craps table. The results were as follows:

1: 2895
2: 4214
3: 3389
4: 3347
5: 4383
6: 3315

This is why I'm so skeptical of the Dakka article that says 1's are rolled more often by cheap dice. Here's an outside source testing cheap dice, and although it's only a single (group of dice? hard to tell from the wording), at least we have the totals for each number it rolled (not just the 1's).

Here's the percentages that that works out to:

1: 13.4%
2: 19.6%
3: 15.7%
4: 15.5%
5: 20.3%
6: 15.4%

Which supports my stance that cheap dice can be weighted so that they don't roll average, but not necessarily rolling more 1's. These dice apparently roll 5's and 2's like mad and 1's the Least of any number!

It's almost like they're axially biased because of non-cubic nature.

However, this doesn't particularly address the pip issue or the rock tumbler issue.

The pip issue is addressed in the thesis available among those dice physics links.

He says that due to physical forces I don't understand -- something to do with the energy state model -- a dice with pips is unfair, though it wouldn't be noticeable in every play.

I believe, however, that this is the Emporer's revenge on Terminators.

I think asmith is referring to the study that is also described in the Dakka article, where he got 29% ones.

I might have miscommunicated... what I meant is that different batches could be skewed towards different numbers. Clearly, his batch was skewed to ones . And the cheap batch studied in the link KK gave was skewed away from ones... definitely one of those ymmv things!

My problem with the ASU research was that it was implying that all GW/chessex/mass-produced type dice would be biased towards ones, as it followed up the rolling tests by talking about the pips. While I'm sure pips have lots of different effects on rolls (and will read the main part of KK's article a bit later to see what they have to say about it) clearly on the dice they tested, this effect was not to bias them towards ones, but towards other numbers (in their case 2 or 5).

I should point out the scientific standard of unfairness of a die is that it does not roll within statistical limits of randomness.

The pipped dice don't, because the pips bias the roll, but they will roll differently enough that you will not notice a difference.

Then there are dice which are more strongly biased, usually due to being of unequal dimensions.

According to the paper, it is possible to make fair dice which are asymmetric.

It dose seem unllike that all Chessex dice would be biased towards 1. I should have thought the amount of tumbling required to abrade them smooth would remove any initial positional bias caused by the moulding facings.

I'm not sure how they actually make the dice, but if they are all molded into a near net shape (including pips) and then tumbled to polish, I think it is very likely they will all be biased the same way (assuming they are made in the same mold). A tumbling process should change surface finish not change dimensions.


Automatically Appended Next Post:
I think the article in warseer where the dakka article comes from has a link to the actual paper and raw data. I read it when it originally came out on warseer and not since so I could be mistaken.

That could be true, I'd be interested to see another test on GW dice. Might try it on mine if I get the chance. Chances are the Dice Physics article was using some other random batch of cheap dice (it even says so) which more than likely aren't GW's . So it is still possible that GW dice cubes are all skewed towards 1s. The only way for me to disprove this is to test my set... which I will probably do, given that anecdotally we have noted them to roll really, really high! So we will see...

Wow this is a complex topic.

Seems like if you grab some dice off the shelf and the dice accidentally roll better without you knowing then your not a cheat. However if you test them and find that they are better and knowingly continue to use them then you are a cheat.

Reminds me of Schrödinger's cat, you will simultaneously be a cheater and not a cheater at the same time if you don't test your dice. However its the act of testing the determines if you are a bad person or not.

doomworcs wrote:Wow this is a complex topic.

Seems like if you grab some dice off the shelf and the dice accidentally roll better without you knowing then your not a cheat. However if you test them and find that they are better and knowingly continue to use them then you are a cheat.

Reminds me of Schrödinger's cat, you will simultaneously be a cheater and not a cheater at the same time if you don't test your dice. However its the act of testing the determines if you are a bad person or not.

That actually sums up my points very well =D

However I'm now tempted to buy some GW dice purely for the purpose of testing out. As a student I have a lot of free time to sit on the floor and roll dice. =p

Oshova

doomworcs wrote:Wow this is a complex topic.

Seems like if you grab some dice off the shelf and the dice accidentally roll better without you knowing then your not a cheat. However if you test them and find that they are better and knowingly continue to use them then you are a cheat.

Reminds me of Schrödinger's cat, you will simultaneously be a cheater and not a cheater at the same time if you don't test your dice. However its the act of testing the determines if you are a bad person or not.

Well, I could agree with you for the most part, with two BIG exceptions:

First, there isn't an issue of cheating if your dice "roll better". The issue is if your dice roll better than average. The game assumes that your dice have an equal chance of landing on any number from 1 through 6. Otherwise, it would be fine to use loaded dice. If you have some dice that roll lower than average, and you swap them for dice that roll average, you aren't cheating. You are playing under the assumptions of the game.

Second, while you might not be testing dice to select the highest rollers (which is definitely cheating), if you purchase dice which you know to be more likely to be imperfectly random (such as polished dice), you are voluntarily purchasing dice which COULD bias your dice rolls higher or lower. Choosing to purchase dice which are more likely to be perfectly random would indicate that you are attempting to follow the rules and expectations of the game more closely.

As an example, you could make measurements with a good quality tape measure, or with a piece of string with knots tied at every inch. While there is no particular reason to suspect the "knotted string" user is deliberately cheating, it would seem to be clear that someone using a more conventional tape measure would be trying to follow the rules a bit more closely.



The problem is that it is very difficult to quantify exactly how 'non-random' a particular set of dice are, and even more difficult to quantify exactly how valuable, financially, a more perfectly random set of dice are.

The fact that gaming surfaces are complex and irregular is not a valid point in the argument, since you would presumably be rolling the polished and unpolished dice on the same surface.

Oshova wrote:

doomworcs wrote:Wow this is a complex topic.

Seems like if you grab some dice off the shelf and the dice accidentally roll better without you knowing then your not a cheat. However if you test them and find that they are better and knowingly continue to use them then you are a cheat.

Reminds me of Schrödinger's cat, you will simultaneously be a cheater and not a cheater at the same time if you don't test your dice. However its the act of testing the determines if you are a bad person or not.

That actually sums up my points very well =D

However I'm now tempted to buy some GW dice purely for the purpose of testing out. As a student I have a lot of free time to sit on the floor and roll dice. =p

Oshova

Cheating is deliberately breaking the rules. You aren't cheating unless you select biased dice because you know they are biased.

The main causes of bias in dice is irregular dimensions, and uneven weighting.

Don't waste time rolling billions of dice. Simply grab a bunch of supposedly identical dice and make a vertical stack with all the facings in the same orientation. For instance, all ones on top, and twos facing forwards. Measure the height of the stack. Now stack the dice again, with twos facing up and ones facing forwards. Measure the new stack and see if it is different to the other one. Now do a third stack with the third axis of the dice, threes up and twos forwards.

Ideally you need 100 "identical" dice, since the standard for fairness is that the dice would have identical dimensions to 0.01 of a mm. In other words, if the dimensions are out by more than that, and the bias is regular, you should find one stack is 1mm or more taller than another. You may easily find a bigger bias, though.

See the Lou Zocchi videos for a demo of this technique.

If you do need to test dice empirically by rolling, then a dice rolling and counting machine is a good idea.

RiTides wrote:
My problem with the ASU research was that it was implying that all GW/chessex/mass-produced type dice would be biased towards ones, as it followed up the rolling tests by talking about the pips. While I'm sure pips have lots of different effects on rolls (and will read the main part of KK's article a bit later to see what they have to say about it) clearly on the dice they tested, this effect was not to bias them towards ones, but towards other numbers (in their case 2 or 5).

Yiyiyi

There's three sets of bias that can occur. The first is axial, when the die isn't cubic. If you think about squashing a die f lat, as the other sides get smaller and smaller, the sides that aren't squashed come up more and mores 2 and 5 are... on opposite sides. The second is pips, the third is edge rounding.

Those are all different.

Edge rounding is only a cause of bias if it is uneven. There are precision dice for Backgammon, which have rounded corners.

A micrometer screw gauge would be useful for measuring dice in all their dimensions.

A chemical balance would be useful for weighing them.

You don't need to worry about the pips. They don't make any significant difference unless millions of dollars are involved.

I think the 'conventional wisdom' put forth by some is that all 'cheap dice' share the same flaw which is more 1s than anything else due to the drilled pips.

I think what http://dicephysics.info/ shows is that the pips are not necessarily the contributing factor to the flaw and flawed dice can be flawed to roll *ANY* side as the unbalanced side. So it is very possible to have dice that roll more 5s than anything. And in the example on the site, that particular dice will roll 4+ more than 50% of the time which would make that a nice dice to have for special saves...

Hence segmented dice being an issue. Being explicitly tested in a college physics lab, being able to control the outcome with rolling techniques or even just using a dice long enough to get a feel that it rolls better, segmenting the dice is a problem. *ANY* dice should be good enough or bad enough to roll for every roll for either player.