Randomisation

[Fruitcake]

So what is true randomisation?

Before I start, I should make it clear that I am a ‘determinist’, and so hold the belief that nothing is ultimately random.

Machines by their very nature and the humans who control them cannot be 100% random. It is a known scientific fact that humans find pure random behaviour very difficult. Machines based on mathematics are also subject to influences that ensure true randomisation does not occur.

The argument could be taken universally. It can be argued that even sub atomic particles which seem to act in a random fashion could be adhering to, as yet uncovered, laws of physics so are not random…but this has still to be proven and is my opinion only.

So the next level of argument, should you agree with above, must be, can there really be true randomisation? If not, how close can we get to this Holy Grail?

Random.org uses atmospheric noise levels to determine a random sequence. One could say that this could be forecast, however…to quote random.org “it would require knowledge of the position and velocity of every single molecule in the planet's weather systems”

Something beyond all of us, but still not random!

If we assume for this argument that pure randomisation is not possible (as I do believe) then we have to attempt to get as close as possible.

Random.org, as I have mentioned before, uses very complex systems to try to ensure randomisation. This is the source of the dice used on cc.

The ‘stickiness’ seen by many, including myself has an oddity to it. However, on closer inspection, the records show some very abnormal runs. To try to chase these runs would be akin to writing all the numbers down that a roulette wheel spins, then only back those that haven’t appeared. Now this is great if you are a believer in averages. But any mathematician will tell you, that however unlikely, you could back a particular number that never appears! You are backing this number simply because you are confusing the laws of probability with the laws of averages. I stand guilty of this myself, after all we are only human so have a tendency to lean towards averages even when we KNOW this cannot be the case…another example of how humans find randomisation so difficult.

Probability maths is actually simpler than you think. Each time you roll the dice, it should have zero memory; otherwise randomisation goes out of the window. As with the roulette wheel not producing a number, conversely it is possible, but highly unlikely, that you roll a triple 1 for ever.

This could not happen here as CC uses a 600,000 dice roll file, originated from random.org. If we take 'n' as the number of perms on a 3 roll, including the different ways the perm can present itself (so 123 could also be 132, 213, 231, 312, 321 but still give the same net result) there are some 216, or 'n'.

So 600,000/n = 2777.7∞ average

By deleting lines after a roll, it concerns me that CC are introducing a memory and, by default, an influence, albeit very small.

Twill and I have exchanged communications about this.

The question I am going to investigate and study as a project is how we, as a community, can ensure the very nearest randomisation that we can.

If anyone could help, I would welcome this.

The first part of this project must be to gather evidence and statistics.

Even if maths, probabilities, averages, etc is not your bag, you can help by keeping a list of your 3 dice rolls. I need a random sample of a couple of hundred at least. You can list them all and pm me with them, I would appreciate it.

If you have an issue with any of my mathematical belief systems, please feel free to pm me. I always enjoy such debates.
_____________________________________________________________

I am the wisest man alive, for I know one thing, and that is that I know nothing.

[the_fatty]

they use 12000 somrthing sets of dice, and when they run out, they just reset them.

[Senfive]

they use 12000 somrthing sets of dice, and when they run out, they just reset them.

Thats his point entirely. Say you use 8000 rolls of those 12000. and 5000 of them Were in your favor. That leaves only 1000 good rolls left of the 4000 rolls remaining before being reset. Therefore it isn't Random. I hope that made as much sense to you, and everyone else, as it did to me.

[MrBenn]

I was thinking about the numbers from random.org the other day, and the following process is my understanding of how the numbers/dice rolls are compiled:

1. Generate a list of 600,000 (12,000 x 5) random numbers :
ie. {1,4,5,2,2,6,4,2,1,1,1,5,6,3...}

2. Format this into a 5-column table:
{1,4,5,2,2}
{6,4,2,1,1}
{1,5,6,3...}

3. Read the relevant numbers when dice are rolled:
{A1,A2,A3,D1,D2}
{1,4,5, -- 2,2}
{6,4,2, -- 1,1}
{1,5,6, -- 3...}

To my mind, I think it would be a better approximation of real-life random dice to start with 5 seperate lists of random numbers (The numbers in this list are not truly random, as I was arbitarily typing numbers):
A1: {5,3,2,4,5,5,6,3,6,2...}
A2: {1,1,5,5,2,4,5,2,4,3...}
A3: {5,2,4,6,3,1,5,2,3,5...}
D1: {1,4,2,3,6,5,6,1,5,5...}
D2: {4,6,1,2,4,2,5,3,6,6...}

This way, I suspect that any "streakiness" would be more confined to individual dice, and I'm certainly happier with the matrix above than one that is effecitvely {A1,A2,A3,D1,D2,A1,A2,A3,D1,D2,A1,A2,A3,D1,D2...}

[DiM]

i have the best solution. get a guy put him in chains and each time somebody clicks the attack button a small electric shock forces him to throw the dice.
now that would be truly random with real dice.
of course the guy may have to be changed every few hours because of the electric shocks but we can live with that for true randomness.

[jiminski]

factory-farmed aardvarks, fed on dice (and ants of course) When they poop, an elf records each roll, sending the results by motorcycle courier-mouse to Sean Connery.
Sean Connery is fed exclusively on hagis, neeps, tatties and whiskey! This makes Sean Connery very windy; he then farts the number of times represented on the Aardvark-poop-dice. The pungent parps are then recorded onto a compact magnetic tape and in turn sent (using the same courier-mouse from earlier) to Lack the love-turtle, who manually inputs the result in each case... (except on the 3rd full moon after summer-solstice, when Wicked takes over, due to Lacks mating season.)

I think it is the only truly random solution!

[alex_white101]

factory-farmed aardvarks, fed on dice (and ants of course) When they poop, an elf records each roll, sending the results by motorcycle courier-mouse to Sean Connery.
Sean Connery is fed exclusively on hagis, neeps, tatties and whiskey! This makes Sean Connery very windy; he then farts the number of times represented on the Aardvark-poop-dice. The pungent parps are then recorded onto a compact magnetic tape and in turn sent (using the same courier-mouse from earlier) to Lack the love-turtle, who manually inputs the result in each case... (except on the 3rd full moon after summer-solstice, when Wicked takes over, due to Lacks mating season.)

I think it is the only truly random solution!

its hagGis

but otherwise very interesting

[Iron Butterfly]

Perfect Randomness possible?

Lightening strikes are random acts enfluenced by many variables. You have a greater chance of getting hit while under a tree. There is a tree outside my house, which is why I roll triple ones alot.

I think its a terrific project you are carrying out.

Honestly I think that alot of the hub bub is in peoples imaginations as they are on the losing end of bad dice.

How close to perfect randomness would one need to get to to recognize a difference? *grin* nice philosophicle question to ponder.

[BaldAdonis]

how we, as a community, can ensure the very nearest randomisation that we can.

Maybe we can use the best available random number generators and an efficient system for making specific rolls. Or, we can ramble on about how nothing is ever random and conduct surveys. We, as a community, need to decide which course is best.

the following process is my understanding of how the numbers/dice rolls are compiled

Go here, number 17, "How do the dice work?" Also, go see random.org for how they make the files. It's more like the way you describe wanting it to be, except without the problem of skipping over columns when certain dice are not used.

[jiminski]

factory-farmed aardvarks, fed on dice (and ants of course) When they poop, an elf records each roll, sending the results by motorcycle courier-mouse to Sean Connery.
Sean Connery is fed exclusively on hagis, neeps, tatties and whiskey! This makes Sean Connery very windy; he then farts the number of times represented on the Aardvark-poop-dice. The pungent parps are then recorded onto a compact magnetic tape and in turn sent (using the same courier-mouse from earlier) to Lack the love-turtle, who manually inputs the result in each case... (except on the 3rd full moon after summer-solstice, when Wicked takes over, due to Lacks mating season.)

I think it is the only truly random solution!

its hagGis

but otherwise very interesting

noo noo! the ommision of a 'G' maintains the random-schmandom-matrix!!
(I appologise if i am geting over-technical)

[yeti_c]

I was thinking about the numbers from random.org the other day, and the following process is my understanding of how the numbers/dice rolls are compiled:

1. Generate a list of 600,000 (12,000 x 5) random numbers :
ie. {1,4,5,2,2,6,4,2,1,1,1,5,6,3...}

2. Format this into a 5-column table:
{1,4,5,2,2}
{6,4,2,1,1}
{1,5,6,3...}

3. Read the relevant numbers when dice are rolled:
{A1,A2,A3,D1,D2}
{1,4,5, -- 2,2}
{6,4,2, -- 1,1}
{1,5,6, -- 3...}

To my mind, I think it would be a better approximation of real-life random dice to start with 5 seperate lists of random numbers (The numbers in this list are not truly random, as I was arbitarily typing numbers):
A1: {5,3,2,4,5,5,6,3,6,2...}
A2: {1,1,5,5,2,4,5,2,4,3...}
A3: {5,2,4,6,3,1,5,2,3,5...}
D1: {1,4,2,3,6,5,6,1,5,5...}
D2: {4,6,1,2,4,2,5,3,6,6...}

This way, I suspect that any "streakiness" would be more confined to individual dice, and I'm certainly happier with the matrix above than one that is effecitvely {A1,A2,A3,D1,D2,A1,A2,A3,D1,D2,A1,A2,A3,D1,D2...}

A query for you MrBenn...

How would you read from your new matrix?

Would you randomly read a column from the row for each dice?

If so - Where do you get the random read from?

My point - which I'm sure people will have missed - is that the earlier example (Where data is parsed into preassigned rolls) is more random as you are only assigning the random data once...

In your instance - you would need to read from your new matrix "randomly" using the PHP random number generator... which is distinctly less random than random.org... Thus making that method more constrained and less random...

C.

PS

How close to perfect randomness would one need to get to to recognize a difference? *grin* nice philosophicle question to ponder.

Great point - shame you spelt philosophical incorrectly!!

[Fruitcake]

I am concerned more that the 3 dice rolls by the attacker seem to have some very odd anomolies.

As I have mentioned previously, triple 1 appears with a monotonous regularity. This naturally makes one begin to lose faith in the way it is structured, QED I would ask you to keep a record of the dice rolls you receive.

Only once we start garnering this information can we even begin to see if there is a problem.

Thanks.

[owenshooter]

i have the best solution. get a guy put him in chains and each time somebody clicks the attack button a small electric shock forces him to throw the dice. .

no thanks... someone in my family pretty much, already had that job... but thanks for bringing up the cruel reality of slavery yet again, DiM. DiM, you rant and rave against a member with the name "hitler", and campaigned to make CC the most PC place on the web. yet you turn around and joke about slavery... i am appalled and disgusted by your post... i hope a forum ban is forthcoming... ahem...-0

p.s.-see? where does it stop?

[DiM]

i have the best solution. get a guy put him in chains and each time somebody clicks the attack button a small electric shock forces him to throw the dice. .

no thanks... someone in my family pretty much, already had that job... but thanks for bringing up the cruel reality of slavery yet again, DiM. DiM, you rant and rave against a member with the name "hitler", and campaigned to make CC the most PC place on the web. yet you turn around and joke about slavery... i am appalled and disgusted by your post... i hope a forum ban is forthcoming... ahem...-0

p.s.-see? where does it stop?

don't make me come and spank you

[owenshooter]

don't make me come and spank you

and now with the jungle fever.-0

p.s.-can we keep this in PM's, so this
thread can continue without our warped humor
and Twill doesn't come lock us out for a bit, and
wicked doesn't threaten me with banishment in
a PM? thank you.

[Audax]

Firstly I must declare an personal interest as I know Fruitcake in the real world.

We have taken copies of some truly wierd sequences.

You already have my files Fruitcake, but I will continue to record and pass them over.

[MeDeFe]

To my knowledge this is how the dice work: there's a file consisting of 600000 lines with 5 numbers in each which in turn are based on athmospheric noise, being used by people from all over the world all the time, and when the 600k lines have been used up the file gets reshuffled and sometimes a new file is obtained from random.org


You know, to me that sounds about as random as you can reasonably get it.

[Vyk]

How about a program that shuffles the 6 numbers on the dice around, 1,2,3,4,5,6 for each dice. Would that be so hard?

The program should only know the %s of each number appearing. I'm not a programmer or a script kiddie, but it seems to me this shouldn't be so hard to do. Why bother with atmospheric whatever... I mean honestly do my rolls have to be ass raped because it's snowing in china? I mean honestly these dice suck.

[Twill]

Mr Benn hit it pretty close.

The file is 500,000 rolls and once each roll has been used we turn around and re-use the file.
We re-fresh the file from time to time...on a random basis...when we have time

We are looking at setting up a system which pulls a new dice file every day, but will still work on the same A1A2A3D1D2 format

the A1{1,2,3,4,5,6,1,2,3,4,5,6} is an interesting idea but difficult to implement...I'll have to talk to Lack about the technicalities of it.

[owenshooter]

We are looking at setting up a system which pulls a new dice file every day, but will still work on the same A1A2A3D1D2 format.

that is truly interesting...-0

p.s.-now tone it down

[Bean_]

Mr Benn hit it pretty close.

The file is 500,000 rolls and once each roll has been used we turn around and re-use the file.
We re-fresh the file from time to time...on a random basis...when we have time

We are looking at setting up a system which pulls a new dice file every day, but will still work on the same A1A2A3D1D2 format

the A1{1,2,3,4,5,6,1,2,3,4,5,6} is an interesting idea but difficult to implement...I'll have to talk to Lack about the technicalities of it.

Yes, if you just pull a new dice file each day, that would be much better than the current methodology. Really, the best solution is just to pull up a dice file in advance of when you use it, and use it only once, straight through. Putting things into A1A2A3D1D2 format is perfectly fine. That isn't what needs fixing.

It's the reuse of the existing file and requiring all lines be used that injects non-randomness into the system.

The quickest fix that I can suggest, with the most bang for the buck, would be to restart the full 500K-line dice file after 100K rolls or so. That won't be perfect, but should be very easy to implement and should largely eliminate the "stickiness" or "bunching" problem, because the dice lines will not be so depleted that they diverge too much from randomness.

[Bean_]

So what is true randomisation?

Before I start, I should make it clear that I am a ‘determinist’, and so hold the belief that nothing is ultimately random.

Machines by their very nature and the humans who control them cannot be 100% random. It is a known scientific fact that humans find pure random behaviour very difficult. Machines based on mathematics are also subject to influences that ensure true randomisation does not occur.

The argument could be taken universally. It can be argued that even sub atomic particles which seem to act in a random fashion could be adhering to, as yet uncovered, laws of physics so are not random…but this has still to be proven and is my opinion only.

So the next level of argument, should you agree with above, must be, can there really be true randomisation? If not, how close can we get to this Holy Grail?

Random.org uses atmospheric noise levels to determine a random sequence. One could say that this could be forecast, however…to quote random.org “it would require knowledge of the position and velocity of every single molecule in the planet's weather systems”

Something beyond all of us, but still not random!

If we assume for this argument that pure randomisation is not possible (as I do believe) then we have to attempt to get as close as possible.

Random.org, as I have mentioned before, uses very complex systems to try to ensure randomisation. This is the source of the dice used on cc.

The ‘stickiness’ seen by many, including myself has an oddity to it. However, on closer inspection, the records show some very abnormal runs. To try to chase these runs would be akin to writing all the numbers down that a roulette wheel spins, then only back those that haven’t appeared. Now this is great if you are a believer in averages. But any mathematician will tell you, that however unlikely, you could back a particular number that never appears! You are backing this number simply because you are confusing the laws of probability with the laws of averages. I stand guilty of this myself, after all we are only human so have a tendency to lean towards averages even when we KNOW this cannot be the case…another example of how humans find randomisation so difficult.

Probability maths is actually simpler than you think. Each time you roll the dice, it should have zero memory; otherwise randomisation goes out of the window. As with the roulette wheel not producing a number, conversely it is possible, but highly unlikely, that you roll a triple 1 for ever.

This could not happen here as CC uses a 600,000 dice roll file, originated from random.org. If we take 'n' as the number of perms on a 3 roll, including the different ways the perm can present itself (so 123 could also be 132, 213, 231, 312, 321 but still give the same net result) there are some 216, or 'n'.

So 600,000/n = 2777.7∞ average

By deleting lines after a roll, it concerns me that CC are introducing a memory and, by default, an influence, albeit very small.

Twill and I have exchanged communications about this.

The question I am going to investigate and study as a project is how we, as a community, can ensure the very nearest randomisation that we can.

If anyone could help, I would welcome this.

The first part of this project must be to gather evidence and statistics.

Even if maths, probabilities, averages, etc is not your bag, you can help by keeping a list of your 3 dice rolls. I need a random sample of a couple of hundred at least. You can list them all and pm me with them, I would appreciate it.

If you have an issue with any of my mathematical belief systems, please feel free to pm me. I always enjoy such debates.
_____________________________________________________________

I am the wisest man alive, for I know one thing, and that is that I know nothing.

I concur with this. It is exactly right that a memory is injected into subsequent rolls because each of the 500,000 lines must be used. Memory is also injected through a re-use of the dice file (until it is flushed). The "memory" effect is apt to be very small on the first several hundred thousand lines, but once you are down to the last thousand lines or so, the probabilities could be significantly off (in either direction), due to the small variations from expected values that were naturally present in the first 490,000 or so lines used. What CC should do is to check the last 1000 dice lines in each cycle. Over the long run, they should have the right distribution in the Dice Analyzer sense (i.e., 1s, 2s, 3s etc. should show up about 16.67% of the time, or more accurately, reflect the underlying percentages in the dice file, which is not going to be exactly 16.67%), but I am betting that they will be much more streaky than randomness should indicate.

If you click on my post history, you'll see that I have been saying the same thing for months. The problem does not appear to be with random.org, but with CC's use of the numbers provided by random.org.

[KoE_Sirius]

Another dice thread...Well done .........by far the lengthiest....
Even if they are not random....The are not random for all of us.So all is fair.

[MrBenn]

I was thinking about the numbers from random.org the other day, and the following process is my understanding of how the numbers/dice rolls are compiled:

1. Generate a list of 600,000 (12,000 x 5) random numbers :
ie. {1,4,5,2,2,6,4,2,1,1,1,5,6,3...}

2. Format this into a 5-column table:
{1,4,5,2,2}
{6,4,2,1,1}
{1,5,6,3...}

3. Read the relevant numbers when dice are rolled:
{A1,A2,A3,D1,D2}
{1,4,5, -- 2,2}
{6,4,2, -- 1,1}
{1,5,6, -- 3...}

To my mind, I think it would be a better approximation of real-life random dice to start with 5 seperate lists of random numbers (The numbers in this list are not truly random, as I was arbitarily typing numbers):
A1: {5,3,2,4,5,5,6,3,6,2...}
A2: {1,1,5,5,2,4,5,2,4,3...}
A3: {5,2,4,6,3,1,5,2,3,5...}
D1: {1,4,2,3,6,5,6,1,5,5...}
D2: {4,6,1,2,4,2,5,3,6,6...}

This way, I suspect that any "streakiness" would be more confined to individual dice, and I'm certainly happier with the matrix above than one that is effecitvely {A1,A2,A3,D1,D2,A1,A2,A3,D1,D2,A1,A2,A3,D1,D2...}

Mr Benn hit it pretty close.

The file is 500,000 rolls and once each roll has been used we turn around and re-use the file.
We re-fresh the file from time to time...on a random basis...when we have time

We are looking at setting up a system which pulls a new dice file every day, but will still work on the same A1A2A3D1D2 format

the A1{1,2,3,4,5,6,1,2,3,4,5,6} is an interesting idea but difficult to implement...I'll have to talk to Lack about the technicalities of it.

Surely it wouldn't be much more difficult... Instead of a single incremental variable to say which line of dice rolls to read next, you would need 5 variables (one for each dice) that increment only for the dice that are actually used.

[MrBenn]

A query for you MrBenn...

How would you read from your new matrix?

Would you randomly read a column from the row for each dice?

If so - Where do you get the random read from?

My point - which I'm sure people will have missed - is that the earlier example (Where data is parsed into preassigned rolls) is more random as you are only assigning the random data once...

In your instance - you would need to read from your new matrix "randomly" using the PHP random number generator... which is distinctly less random than random.org... Thus making that method more constrained and less random...

C.

Yeti_c, my intention was to read the numbers from the matrix sequentially, but only when each dice is rolled... (rather than getting numbers and discarding them when attacking/defending with 1, for example)