Streaky Dice Question

[natty dread]

Ok, let me put it this way:

What kind of results would you want the dice to give so that you would be satisfied with them?

[SirSebstar]

the anticipated and announced soon to be finished update of a dicetracker has not been finished just yet. it ran into some troubles while testing it.
with regards to the dice analyzer, i know nothing

[musicalmaven]

to sir sebstar - thanks for the update on the dice tracker program. that it ran into some problems is not surprising considering the complexity of such a task. good luck on resolving it.
as for the dice analyzer - well, that makes two of us.
to natty - to whom are you asking your question?
if it is me - my answer would be "normalcy".
it may be simplistic, but to me dice results should be the same for everyone.
if you did a chart of your results it should look like an oscillating line. the base line would be the mathematical average, about 54% for 3 vs 2. sometimes we would be above it and sometimes we should be below.
naturally, the length of time between being above and being below average would vary each time we crossed the "base-line" as well as for each individual and the height and depth above the norm and below it would also vary, but it should still be oscillating.
as i said, simplistic, but that is my answer. (and just in case you are unaware of it - my records indicate my line to be a generally steady line, below average. i have never been above average in the over 3 years that i have been keeping records.)

[wmdvanzyl]

This is a dangerous place to jump in, but i'm going to anyway.

There are a lot of discussions about dice randomness and that by itself should be an indicator that something is wrong. OR at least that it needs good investigation. I have only been here a while, but i know the sights and sounds of an unhappy community, both online and in real life. Let me also state that bad dice rolls hurt a lot more than good ones - that accounts for some of the unhappiness, but not all of it. I don't have the answer, but i am curious about something, thus the post.

There seems to be a lot of emphasis on the "randomness" of the dice. I have to take a step back and ask whether the underlying assumption has been tested. The underlying assumption is of course that if the dice rolls are random then they would be fair. Now everyone's definition of "fair" probably differs and therefore i suggest that the best measure of fair is correspondence to actual dice rolls - yes the ones with six-sided blocks. Generated randomness is a tricky business and seems to have taken the focus off the real question. Does the online experience correspond with actual risk games in real life?

At this point i feel more inclined to say no. I play a reasonable amount of Risk IRL and i don't get the same "feel" from the dice i get online and i don't mean just an occasional streak either. A lot of people have made the point that the feel done in by the dice and considering all aspects i must agree. There are some pretty complex mathematics involved in the outcomes of dice rolls IRL (the amount of dice being rolled, tipping points of equal and over combinations). The same is true for generators, but the issues are different, such as the type of generator (pseudo, quasi, etc) and how it is being modulated to a range of 1 to 6. Interesting the issues on each differ and the question becomes obvious - are the results comparable?

So what we probably need is a mathematician to study the outcomes and issues or RL dice rolls and then compare them with the results of generated dice rolls on this site. Someone could probably do a thesis on this...

[SirSebstar]

The problem is that a mathmatician has already gone over it. It is a sound as a bell...
The problem is the perception. I think the player who can come up with a dicesystem that makes everybody happy... well a nobelprize will be the first to follow.. no joke. I really think its that hard..

its psychological, people remember the bad more then the good, so they complain. so they see someone complain, thus it must be true. it is not a positive way to go forward, which is why i think it is good to once and a while wrap up all dicethreads into one. since they all are about the same subject.

This is a dangerous place to jump in, but i'm going to anyway.

There are a lot of discussions about dice randomness and that by itself should be an indicator that something is wrong. OR at least that it needs good investigation. I have only been here a while, but i know the sights and sounds of an unhappy community, both online and in real life. Let me also state that bad dice rolls hurt a lot more than good ones - that accounts for some of the unhappiness, but not all of it. I don't have the answer, but i am curious about something, thus the post.

There seems to be a lot of emphasis on the "randomness" of the dice. I have to take a step back and ask whether the underlying assumption has been tested. The underlying assumption is of course that if the dice rolls are random then they would be fair. Now everyone's definition of "fair" probably differs and therefore i suggest that the best measure of fair is correspondence to actual dice rolls - yes the ones with six-sided blocks. Generated randomness is a tricky business and seems to have taken the focus off the real question. Does the online experience correspond with actual risk games in real life?

At this point i feel more inclined to say no. I play a reasonable amount of Risk IRL and i don't get the same "feel" from the dice i get online and i don't mean just an occasional streak either. A lot of people have made the point that the feel done in by the dice and considering all aspects i must agree. There are some pretty complex mathematics involved in the outcomes of dice rolls IRL (the amount of dice being rolled, tipping points of equal and over combinations). The same is true for generators, but the issues are different, such as the type of generator (pseudo, quasi, etc) and how it is being modulated to a range of 1 to 6. Interesting the issues on each differ and the question becomes obvious - are the results comparable?

So what we probably need is a mathematician to study the outcomes and issues or RL dice rolls and then compare them with the results of generated dice rolls on this site. Someone could probably do a thesis on this...

[natty dread]

There are a lot of discussions about dice randomness and that by itself should be an indicator that something is wrong.

No it shouldn't. That's a classic argumentum ad populum.

There seems to be a lot of emphasis on the "randomness" of the dice. I have to take a step back and ask whether the underlying assumption has been tested. The underlying assumption is of course that if the dice rolls are random then they would be fair.

Of course they would. Real life dice are very much random (unless you use weighted dice or something). The dice must be random. They don't need to be "fair", and if you try to artificially make the dice rolls "fair", you open a whole new can of worms... this has been suggested before, lots of times, by the dice complainers - get random numbers, but "normalize" them so that long streaks won't occur... the only problem is, then the dice would no longer be random, they would be predictable - at a certain point, with enough losses, you would know with 100% certainty that you would get a winning roll. Predictable dice outcomes would make the game predictable and boring.

There's a reason risk uses dice, there's a reason for the random element. It makes games fun. Except for those few complainers.

The same is true for generators, but the issues are different, such as the type of generator (pseudo, quasi, etc) and how it is being modulated to a range of 1 to 6

CC dice are not generated with RNG:s, they come from Random.org which uses atmospheric noise for random numbers.

[donkeymile]

Another day of the same PATTERN - yes not random but PATTERN of my dice sucking while my opponents win dice defying odds .... PATTERN - CONSISTENT PATTERN ...

I lose 8v2 (happens) opponent next turn wins 7v8 losing 0 .. lol

I have 3 piles of 8 vs a 8 ... end result is all my piles are 1 my opponent has 7 left .... AND this is right after he took a 6v8 perfect .....

I could go on but its pointless - that was just two of the games I played today ...

AGAIN - not complaining that my dice suck - it happens in random ... but when my dice suck my opponents win dice defying odds at the same time CONSISTENTLY ... which is indicative of nonrandom ....even my opponents comments are funny as tehy can't explain how their dice are so good lol ..... just laughing to keep from crying as I close out my remaining 2 games here ... sure hope that new add-on comes soon so I can prove what I've been saying.

had to add 0- another game opponent wins 4v8 then I lose 8v3 and 6v3 and 6v3 ... ho hum

[Metsfanmax]

There already exist add-ons which could "prove" what you've been saying. They're in the available tools forum.

[donkeymile]

I haven't been using Firefoxm, and last time I checked it out I needed Firefox.

I might just start using it but I'd need to play a long time before the averages would be accepted here.

so sick of the consistent bullshit .............. the pattern that isn't supposed to exist in random ... lol

[SirSebstar]

There already exist add-ons which could "prove" what you've been saying. They're in the available tools forum.

soonish there is going to be a new one, that does not require firefox, but it integrated into the site.
soonish is an unknown timeperiod, sorry

[natty dread]

Donkeymile... I took a look at your recent games, and one the first page of your games (the most recent ones) there's only 3 non-1v1 games. All others are 1v1 and most looked to be speed games.

Furthermore, you seem to have won a fair amount of them.

This again supports my theory: there's really nothing wrong about your dice. You just love to complain and victimize yourself.

Whenever you want to break the habit again, I'll be there to offer you counceling if you wish.

Best wishes to you. Addictions are hard to break.

[musicalmaven]

sorry natty - i have to throw a wet blanket on your arguments.
first of all your "argumentum ad populum" may be generally correct, it isn't always so. just because many people are positing a proposition or presenting a complaint doesn't mean that it is a false premises. you may have a problem that many people have experienced.
windvanzyt's results may be extreme, but still may be factual and the fact that numerous other are complaining may, in fact, prove that there is a problem.
secondly - you state that you went into his games and found most of them to be 1 vs. 1 and that he's "won a fair number of them" and that posits your theory that he, and bu extension, most, if not all, complainers are victimizing themselves and love to complain about it.
first of all if most of his games are 1 vs. 1, and if he is a decent player, he should have won his "fair share of them" - it is a 50/50 shot at winning 1 vs. 1, hence winning a "fair share of them" hardly proves anything. in fact, if he is a better than average player it may actually prove his point, and he should be winning more than "a fair share of his games".
secondly - whether it's windvanzyt, donkeymile, ragadan or myself, the fact that we are complaining about lousy dice results, doesn't mean that we just love to complain and make ourselves out to be victims. there may actually be substance to our position.
it's too bad that the dice tracker upgrade is not working. it may actually help those of us who experience lousy dice far too frequently prove our point. i am anxiously awaiting its implementation to see how many others like me there are.
your side loves to say that all of us who complain are whiners and have a slanted perception of how the dice are performing (or not performing). a common response is to tell us to show actual numbers, not just an occasional lousy result - to add the dice analyzer to our program, and if we do so, it would prove that overall, the dice are close to normal (and it's not just metsfanmax who has said this).
however, when we do - your side says: a)the numbers are within the normal variance; b)the numbers aren't large enough to be a statistical significance; or c) oh well, that just randomness at work.
finally - a throw out to woodruff. i'm still awaiting your answer to my question from the previous page - what the hell is wrong with the dice analyzer? (sirsebstar answered my question about the dice tracker.)

[Metsfanmax]

your side loves to say that all of us who complain are whiners and have a slanted perception of how the dice are performing (or not performing). a common response is to tell us to show actual numbers, not just an occasional lousy result - to add the dice analyzer to our program, and if we do so, it would prove that overall, the dice are close to normal (and it's not just metsfanmax who has said this).
however, when we do - your side says: a)the numbers are within the normal variance; b)the numbers aren't large enough to be a statistical significance; or c) oh well, that just randomness at work.

Shouldn't (a) be an indication that the null hypothesis is correct; that is, that you really are getting normal results and just aren't perceiving it this way? The "dice complainers" make it sound like they get results several standard deviations outside the mean on a regular basis. If your aggregate results are within one SD or less (about where your results were, MM) then I simply wouldn't expect constantly bad results, I would just expect slightly below average results, that I don't think most humans would generally perceive. The only remaining conclusion is that you are magnifying the effects of the bad results in your head without taking a look at the bigger picture.

[wmdvanzyl]

sorry natty - i have to throw a wet blanket on your arguments.
first of all your "argumentum ad populum" may be generally correct, it isn't always so. just because many people are positing a proposition or presenting a complaint doesn't mean that it is a false premises. you may have a problem that many people have experienced.
windvanzyt's results may be extreme, but still may be factual and the fact that numerous other are complaining may, in fact, prove that there is a problem.
secondly - you state that you went into his games and found most of them to be 1 vs. 1 and that he's "won a fair number of them" and that posits your theory that he, and bu extension, most, if not all, complainers are victimizing themselves and love to complain about it.
first of all if most of his games are 1 vs. 1, and if he is a decent player, he should have won his "fair share of them" - it is a 50/50 shot at winning 1 vs. 1, hence winning a "fair share of them" hardly proves anything. in fact, if he is a better than average player it may actually prove his point, and he should be winning more than "a fair share of his games".
secondly - whether it's windvanzyt, donkeymile, ragadan or myself, the fact that we are complaining about lousy dice results, doesn't mean that we just love to complain and make ourselves out to be victims. there may actually be substance to our position.
it's too bad that the dice tracker upgrade is not working. it may actually help those of us who experience lousy dice far too frequently prove our point. i am anxiously awaiting its implementation to see how many others like me there are.
your side loves to say that all of us who complain are whiners and have a slanted perception of how the dice are performing (or not performing). a common response is to tell us to show actual numbers, not just an occasional lousy result - to add the dice analyzer to our program, and if we do so, it would prove that overall, the dice are close to normal (and it's not just metsfanmax who has said this).
however, when we do - your side says: a)the numbers are within the normal variance; b)the numbers aren't large enough to be a statistical significance; or c) oh well, that just randomness at work.
finally - a throw out to woodruff. i'm still awaiting your answer to my question from the previous page - what the hell is wrong with the dice analyzer? (sirsebstar answered my question about the dice tracker.)

musicalmaven - i think you are confusing me with the thread creator. I just added my voice and i tried to be rather neutral. i haven't been here long enough to really say and i haven't made any real inquiries myself. i am just observing and commenting. i do agree that perception plays a big role. Bad rolls hurt more than good ones. My standpoint atm is just that i don't the same feel from the online experience as from actual Risk games. That's all. I am not complaining - just observing and commenting. I am also not saying that anything needs to be changed. I do hear what those who are unhappy are saying, but i also hear what those are not unhappy are saying. That's all.

The problem is that a mathmatician has already gone over it. It is a sound as a bell...

Can you perhaps publish these results? - i would love to know how he approached it and what he found. I am just curious.

[SirSebstar]

The problem is that a mathmatician has already gone over it. It is a sound as a bell...

Can you perhaps publish these results? - i would love to know how he approached it and what he found. I am just curious.

really? But you never took the opportunity to actually look at the website random.org where we got our numbers from. I could not find the post from the CC'er who went over our numbers, but this shoud do it for now.
http://www.random.org/analysis/ see near the bottom of the page, but the rest is nice too..

regards,
SirSebstar

[wmdvanzyl]

Thank you SirSebstar! I will have a look at it. Just one thing to those who oppose the complainers - i realise you guys see them as nagging and irritating (which i can't imagine as i haven't had to endure the complaining), but just be careful to not be too aggressive. I felt that the response was quite harsh to my first post in this thread. So i can imagine that some newbie comes to the site has some bad rolls and posts something in ignorance and lack of experience about dice rolls and the community (or one or two people) just trashes them. That's one less CC'er for the future. Let's try to stay as positive and uplifting as possible. If we really can't deal with the complaining any more, just completely ignore all discussions about dice rolls(?) - just a suggestion.

[SirSebstar]

all true, I think the potitiveness kinda disappears after having to reply the same happy note over and over again. It tends to make people more cynical which in turn is less nice... still. i concur

[musicalmaven]

wmdvanzyl - correct me if i am wrong, but you did post some pretty bad results. if they simply were comments of a bad result fine, however, most of the people on this thread are complaining about results like yours on a consistent basis. i doubt many of us, on either side of the issue, expect to see a post of such bad dice as a simple comment.
metsfanmax - what is a standard deviation? we have a fairly straight mathematical computation of what the dice results should be (unlike polls which often have a 2 or 3 point variant of the final figure), so what is the variant for our dice results?
and you must also remember, the problem isn't just that the results are "within a single deviant", and so should be perceived as normal, it's that the result is always below normal, and the fluctuation is between 1 sd and 2 sds (or worse - depending on what an sd is for our dice).

[SirSebstar]

wmdvanzyl - correct me if i am wrong, but you did post some pretty bad results. if they simply were comments of a bad result fine, however, most of the people on this thread are complaining about results like yours on a consistent basis. i doubt many of us, on either side of the issue, expect to see a post of such bad dice as a simple comment.
metsfanmax - what is a standard deviation? we have a fairly straight mathematical computation of what the dice results should be (unlike polls which often have a 2 or 3 point variant of the final figure), so what is the variant for our dice results?
and you must also remember, the problem isn't just that the results are "within a single deviant", and so should be perceived as normal, it's that the result is always below normal, and the fluctuation is between 1 sd and 2 sds (or worse - depending on what an sd is for our dice).

RTFT
www.random.org

[RADAGA]

as per RTFT, I mailed Random.org a while ago, and they said they are by no means responsible for the dice in the site conquerclub.com, and the owner said he dont even know about CC at all.

from André Guia xxxxxxxxxxxxxxxxxxxxx@gmail.com
to Mads Haahr <Mads.Haahr@cs.tcd.ie>
date Fri, Oct 3, 2008 at 11:09 AM
subject Re: random.. perfectly random.


Hello André,

So I'm finally finding some time to reply to your email. I have no idea who ConquerClub is and I have no commitment to them to reply to their users' emails. Fortunately, most emails I get are from people who appreciate the free services that are available at my site.

[natty dread]

Care to share the email you sent him?

[SirSebstar]

in the exerpt you have shown us, your comment is either a lie, wrong or incomplete..
they

have no commitment to them to reply to their users' emails

so in otherword, go away. they say nothing about their relationshop with cc other then they not getting paid to reply to your mails.
whats your point again?

as per RTFT, I mailed Random.org a while ago, and they said they are by no means responsible for the dice in the site conquerclub.org, and the owner said he dont even know about CC at all.

from André Guia xxxxxxxxxxxxxxxxxxxxx@gmail.com
to Mads Haahr <Mads.Haahr@cs.tcd.ie>
date Fri, Oct 3, 2008 at 11:09 AM
subject Re: random.. perfectly random.


Hello André,

So I'm finally finding some time to reply to your email. I have no idea who ConquerClub is and I have no commitment to them to reply to their users' emails. Fortunately, most emails I get are from people who appreciate the free services that are available at my site.

[RADAGA]

of course>

Hello, mister.

Concerning the dice, I think they might be displaying a behavior very different from what you would get with real dice.

Don´t you think your numbers wield far too many streaks in a row to be considered random? I am a ConquerClub user, they use your numbers and claim them to be perfect. But the number of times we find strange things with the results is amazing.

I mean, look at the dice roll, it is HARD to find a roll without repeated numbers.

You rolled 5 dice: a double

3 6 6 3 1

You rolled 5 dice: three in a row, 4 of them sixes

6 2 6 6 6

Timestamp: 2008-08-11 20:46:47 UTC

You rolled 5 dice: three same results

5 3 5 2 5

Timestamp: 2008-08-11 20:46:56 UTC

You rolled 5 dice: even dice

1 5 2 4 6

Timestamp: 2008-08-11 20:47:04 UTC

You rolled 5 dice: a double

5 1 6 6 2

Timestamp: 2008-08-11 20:47:17 UTC

You rolled 5 dice: a triplet

1 6 1 1 3

Timestamp: 2008-08-11 20:47:24 UTC


How can we trust it, if every time we get the die roller, it shows results that "are possible" but "should happen once every many rolls", not on "every other" roll


on those 30 dice, for instance, I´ve seen only one FOUR.. ....one four in 30 rolls ... where it should be something around five.

but I got 10 sixes .. 33,3% of the dice.


Try yourself, it happens every time.


Thanks for your time.


André

[AgentSmith88]

Just like to point out that not everyone get get 30% bad dice and only 10 % good dice. If your dice suck, then whoever you are attacking is getting good dice. While losing multiple troops while attacking is frustrating, it in no way means the dice are not random. I agree the dice do appear streaky at times, but perception does not always equal reality.

[RADAGA]

in the exerpt you have shown us, your comment is either a lie, wrong or incomplete..
they

have no commitment to them to reply to their users' emails

so in otherword, go away. they say nothing about their relationshop with cc other then they not getting paid to reply to your mails.
whats your point again?

Let me teach you to read whole sentences> it goes from the Capital letter that eiher start a paragraph or is preceeded by a symbol, indicating an end of sentence. those generally look like "." ; "!" or "?" and go until you fin one of those symbols again.

therefore, the sentence you quoted really is --> "I have no idea who ConquerClub is and I have no commitment to them to reply to their users' emails.

The important part, that yo left out, surely because you did not knew whee the setence began, and not for any other more obscue reason, is the first one.

Glad to be of service,