Question about dice...

[KLOBBER]

As for Klobber, I'll do an example with the man bites dog approach. Someone rolls a 6 sided die hoping to roll a 6. He does this 10 times and doesn't roll a 6. No reason to panic right? Rolling a 6 isn't expected, so going a long time without one is fine. Now lets say he rolls one million times and still doesn't roll a 6. Can we at this point assume there's something wrong with the die that he's rolling? According to you there isn't a problem because he shouldn't expect to roll a 6. Hopefully this clears things up.

There are at least two major flaws in your example:

1. It does not apply to your above-posted situation because you were referring to a handful of rolls, not a million.

2. I never said that a person should not expect to roll a 6 with a million rolls of a six-sided die, and yet inexplicably you claim that I did say that. Since I did not say it, your example neither applies to your situation, nor to mine.

I can't even console you by saying "close, but no cigar," because you are not even close, in either case.

Your example is just plain "no cigar," sadly.

[maniacmath17]

I'll do this more systematically. Answer this: If someone rolls 1 million dice should they expect to roll at least one 6. Yes or No?

[Velvecarrots]

In order for your example to hold any water, you would first have to predict RIGHT three times in a row, which you have not done. Your position is obviously more similar to the guy who repeatedly loses the lottery, which is, as I said in my original post, yet another example of "dog bites man," and is boring, quite expected, and by no means newsworthy.

I played the lottery 3 times. I expected to lose all three times. Guess what happened? I lost all 3 times!

I predicted correctly. Is that newsworthy?

[KLOBBER]

In order for your example to hold any water, you would first have to predict RIGHT three times in a row, which you have not done. Your position is obviously more similar to the guy who repeatedly loses the lottery, which is, as I said in my original post, yet another example of "dog bites man," and is boring, quite expected, and by no means newsworthy.

I played the lottery 3 times. I expected to lose all three times. Guess what happened? I lost all 3 times!

I predicted correctly. Is that newsworthy?

No, but if you predict the lottery numbers right and win even once, get back to us. Same with CC dice.

[maniacmath17]

In order for your example to hold any water, you would first have to predict RIGHT three times in a row, which you have not done. Your position is obviously more similar to the guy who repeatedly loses the lottery, which is, as I said in my original post, yet another example of "dog bites man," and is boring, quite expected, and by no means newsworthy.

I played the lottery 3 times. I expected to lose all three times. Guess what happened? I lost all 3 times!

I predicted correctly. Is that newsworthy?

No, but if you predict the lottery numbers right and win even once, get back to us. Same with CC dice.

Klobber, answer this question please. If you rolled a die one million times hoping for a 6, would you expect to roll at least one 6? And if you didn't roll a 6, would you suspect something is wrong with that die? Yes or no will do, although you can explain your reasoning if you wish.

[john9blue]

Why do people respond to KLOB's posts? He is a troll.

[cicero]

Why do people respond to KLOB's posts? He is a troll.

This discussion could still remain interesting if we respond only constructively and only to the constructive posts.

A bit like the whole of forum life I guess ...
Now what are the odds of this happening in this thread ?

[Thezzaruz]

Klobber, answer this question please. If you rolled a die one million times hoping for a 6, would you expect to roll at least one 6? And if you didn't roll a 6, would you suspect something is wrong with that die? Yes or no will do, although you can explain your reasoning if you wish.

But you are still mixin and mashing with situations that don't work together. With statistical analyses you can check out dice rolls that have been made to see if the distribution is off (or check for trends, streaks effect of the lunar gravity and what ells you want). And this has been done with CC dice without any anomalies showing up.

However, what ever statistical tool you use its expected value of a future roll will never be relevant as every dice is a new roll with the exact same probabilities as every single roll has and no two future rolls ever influence each other.

I'll give you to same advice I've given others, go to the plug-in forum and get yourself the "Dice Analyzer" plug-in and then roll a few thousand dies and if your result is way off the expected then come back to us.

I see this point made a lot, and I've already done the numbers for it, it still doesn't add up. The problem is assuming that these few hundred thousand rolls are done in the situation in question. Of the almost 5 million games, a 32+ attacking a 2 probably only happens maybe 20 times on average for a total sample of 100 million, and so for one person to lose that's perfectly likely, but from what I've seen this isn't a common occurrence and I'm sure others have lost even though statistically anymore than 1 is highly unlikely with random dice.

You are a bit hard to follow sometimes but you can be sure that a large stack attacking a small stack is quite a common situation. And any results that people find strange with those attacks gets highlighted a lot as most people think that they should be easily won and hence any time they don't it gets remembered.

Why do people respond to KLOB's posts? He is a troll.

You might not like Klobber or his post but all I've seen him post regarding dices is correct, a bit aggressive possibly but still correct and noting close to trollworthy.

[maniacmath17]

Thezzaruz, I'll try not to throw out anymore math as it seems to be hard to follow. Just think of it this way: Big stacks attacking small stacks happen all the time right? Right. But it is sometimes hard to comprehend just how small the chances of that small stack coming out on top are.

Here are the only numbers I'm going to use in this post. The chances of losing 32 v 2 is about one out of 500,000,000. This is just one example. But there are countless other extremely improbable dice results out there on this site. I say improbable having already accounted for all of the many rolls made on the site each day (I'm not going to try to explain the math behind it again).

I'm not saying individual dice rolls are not random because we have plenty of data showing that in the long run, these numbers do level out at their expected values. But if the reported incidents of these outrageous losses are true, then there is reason to believe that the dice are capable of becoming "streaky" based on the shear unlikeliness that all these losses occurred with random dice. Again, i must emphasis the fact that I'm talking about unlikely losses once all the dice rolls in the history of the site have already been factored in.

[tyche73]

i had a 52v12 that ended 3v3

Was that an auto-attack?

yes

[darth emperor]

Thanks Sully. I'll just go back to assuming I'm just the unluckiest person in the world, lol.

Maybe the Dice gods are trying to balance out your skill.


--Andy

Or maybe Dice demonds are attacking him...

[tyche73]

dog just ate my lottery ticket and i bit him
dog is in the corner with a calculater working out the odds of it happening again

[KLOBBER]

dog just ate my lottery ticket and i bit him
dog is in the corner with a calculater working out the odds of it happening again

LOL, good one!

That is one unique dog you have there!

[john9blue]

Why are people against auto attack so much? Previous rolls have no bearing on subsequent rolls, so even if it went from 52-12 to 10-5, you'd still keep attacking, because the odds are still on your side. Unless the dice really are streaky...

[Sperpurber]

Why are people against auto attack so much? Previous rolls have no bearing on subsequent rolls, so even if it went from 52-12 to 10-5, you'd still keep attacking, because the odds are still on your side. Unless the dice really are streaky...

Sure previous rolls have a bearing. Eventually you'll want to cut your losses, especially after losing 42 troops, unless you want to leave yourself wide open.

Then again, it all depends on the circumstances.

[KLOBBER]

He means that the patterns that manifest previously have no effect on what patterns may manifest subsequently.

For example, if I happen to roll a six, 10 times in a row, that neither means that the next roll will be a six, nor that it will not be a six. It could just as easily be either a six or any other number on the dice, no matter what I happen to have rolled in the past.

[ppgangster]

He means that the patterns that manifest previously have affect on what patterns may manifest subsequently.

For example, if I happen to roll a six, 10 times in a row, that neither means that the next roll will be a six, nor that it will not be a six. It could just as easily be either a six or any other number on the dice, no matter what I happen to have rolled in the past.

I think you are old or older for your age, you seem to like to repeat yourself a lot!

[Thezzaruz]

Thezzaruz, I'll try not to throw out anymore math as it seems to be hard to follow.

Math isn't the problem, some of your posts could use a bit of punctuation and stuff though.

But if the reported incidents of these outrageous losses are true, then there is reason to believe that the dice are capable of becoming "streaky" based on the shear unlikeliness that all these losses occurred with random dice.

Dice "streakyness" has been checked too and found non existing.

[KLOBBER]

He means that the patterns that manifest previously have no effect on what patterns may manifest subsequently.

For example, if I happen to roll a six, 10 times in a row, that neither means that the next roll will be a six, nor that it will not be a six. It could just as easily be either a six or any other number on the dice, no matter what I happen to have rolled in the past.

I think you are old or older for your age, you seem to like to repeat yourself a lot!

Sorry about that. It's just that the very intelligent learn the first time, the moderately intelligent on the second or third, and the less intelligent sometimes require a lot of repetition before they learn even the simplest of facts.

Since it's a mixed bag in this forum, saying it once doesn't always suffice, sadly.

[Sperpurber]

He means that the patterns that manifest previously have no effect on what patterns may manifest subsequently.

For example, if I happen to roll a six, 10 times in a row, that neither means that the next roll will be a six, nor that it will not be a six. It could just as easily be either a six or any other number on the dice, no matter what I happen to have rolled in the past.

Right, but I was referring to reasons not to use auto-attack, nothing about patterns of probability.

[KLOBBER]

I know what you were referring to.

I was referring to the fact that the previous commenter was correct on his point, and that your comment did not change the fact that he was correct.

[maniacmath17]

Dice "streakyness" has been checked too and found non existing.

That's the whole point of this thread. HOW did they determine this? More specifically, I would like some sort of mathematical explanation for how there can be so many of these nearly impossible losses when factoring in both the odds of the attack and # of dice rolled on the site.

And KLOBBER, you still didn't answer the question. I never asked what were the chances of rolling a 6 AFTER the 1 million rolls (which as KLOBBER keeps mentioning is 1/6 with random dice). But rather, if you were to look back on the 1 million rolls would you or would you not think that the die you rolled was random given the fact that no 6's were rolled.

[KLOBBER]

Dice "streakyness" has been checked too and found non existing.

...nearly impossible losses when factoring in both the odds....

You just don't get it, do you? The dice are designed to be unpredictable, and the design is successful.

When dealing with UNPREDICTABLE dice:

1. No loss is "impossible," or even "nearly impossible," and

2. there cannot possibly be any predictable "odds" to factor in.

Your error boils down to assuming predictability in the dice, and becoming frightened or angry when your irrational assumption is proven wrong. No matter what arbitrary, unscientific, imaginary "odds" you have in mind before rolling, they will necessarily be incorrect -- that is the inherent nature of unpredictability, and this is a fact that you seem incapable of apprehending, sadly.

The dice are definitely unpredictable, just as their designers intended, and just as internet gaming dice should be. The dice are perfect as they are -- it is only your view of them that is erratic. It's time for you to stop posting such unscientific nonsense based on ignorance of the nature of these dice.

[KoE_Sirius]

I never have problems with my dice .They are pretty much perfect .I do feel for you though.

[xelabale]

Dice "streakyness" has been checked too and found non existing.

...nearly impossible losses when factoring in both the odds....

You just don't get it, do you? The dice are designed to be unpredictable, and the design is successful.

When dealing with UNPREDICTABLE dice:

1. No loss is "impossible," or even "nearly impossible," and

2. there cannot possibly be any predictable "odds" to factor in.

Your error boils down to assuming predictability in the dice, and becoming frightened or angry when your irrational assumption is proven wrong. No matter what arbitrary, unscientific, imaginary "odds" you have in mind before rolling, they will necessarily be incorrect -- that is the inherent nature of unpredictability, and this is a fact that you seem incapable of apprehending, sadly.

The dice are definitely unpredictable, just as their designers intended, and just as internet gaming dice should be. The dice are perfect as they are -- it is only your view of them that is erratic. It's time for you to stop posting such unscientific nonsense based on ignorance of the nature of these dice.

You still aren't answering maniac's question. It is relevant, as the dice aren't random, they are generated in order to appear random. Therefore there is a definite possibility of a problem in the generation. This has nothing to do with odds so don't start on that one.

Personally I don't believe there is a problem with the generation, but it's still a very good question...