One more thing about the dice.

[Timminz]

Yeah, for instance, I should expect things likewhat just happened>

No. You should stop expecting the dice to behave in any predictable fashion.

[RADAGA]

And you say it is nothing wrong with dice>>that was an autoassault on an already lost game

Game 6382466

the results? TEN double losses / FIVE ties and meager THREE double wins - thats more double losses (expected to be 29%) than the sum of the other results (which should be a little more than 79%) - not to mention it starts with a streak of five double losses in a row...

I lost 25 armies and killed 11

Combat - Double lossx1

Roma: * 3 * 4 * 3
- vs -
Sardegna: * 4 * 5

Combat - double lossx2 (1 six - 10 dice)

Roma: * 3 * 5 * 3
- vs -
Sardegna: * 3 * 6

Combat - double lossx3 (2 sixes - 15 dice)

Roma: * 3 * 3 * 4
- vs -
Sardegna: * 6 * 4

Combat - double lossx4 (4 sixes - 20 dice)

Roma: * 3 * 4 * 5
- vs -
Sardegna: * 6 * 6

Combat - double lossx5 (6 sixes - 25 dice, second 66 in a row)

Roma: * 4 * 1 * 3
- vs -
Sardegna: * 6 * 6

Combat - double lossx5 / double winx1

Roma: * 5 * 2 * 3
- vs -
Sardegna: * 1 * 1

Combat - double lossx5 / double winx1 / tie x1

Roma: * 1 * 3 * 1
- vs -
Sardegna: * 2 * 2

Combat - double lossx5 / double winx2 / tie x1

Roma: * 2 * 5 * 4
- vs -
Sardegna: * 1 * 3

Combat - double lossx6 / double winx1 / tie x1 (ANOTHER 66 for defence)

Roma: * 4 * 5 * 2
- vs -
Sardegna: * 6 * 6

Combat - double lossx6 / double winx2 / tie x1

Roma: * 6 * 2 * 6
- vs -
Sardegna: * 1 * 4

Combat - double lossx6 / double winx2 / tie x2

Roma: * 4 * 2 * 3
- vs -
Sardegna: * 5 * 2

Combat - double lossx7 / double winx2 / tie x2

Roma: * 3 * 6 * 3
- vs -
Sardegna: * 5 * 6

Combat - - double lossx8 / double winx2 / tie x2

Roma: * 2 * 3 * 2
- vs -
Sardegna: * 4 * 5

Combat - double lossx8 / double winx2 / tie x3

Roma: * 3 * 2 * 6
- vs -
Sardegna: * 4 * 5

Combat - double lossx8 / double winx2 / tie x4

Roma: * 3 * 1 * 4
- vs -
Sardegna: * 3 * 1

Combat - double lossx9 / double winx2 / tie x4

Roma: * 2 * 3 * 2
- vs -
Sardegna: * 4 * 6

Combat - double lossx9 / double winx2 / tie x5

Roma: * 6 * 1 * 1
- vs -
Sardegna: * 2 * 1

Combat - double lossx10 / double winx2 / tie x5

Roma: * 3 * 2 * 3
- vs -
Sardegna: * 5 * 5

Combat - double lossx10 / double winx3 / tie x5

Roma: * 2 * 6 * 5
- vs -
Sardegna: * 1 * 5

[Chuuuuck]

Rolling 5 doubles losses in a row should statistically happen 1 time in every 487 sets of rolls.

Now think about how many rolls probably happen in a given day or even in a given hour on this site? I have rolled 1,000 armies vs 1,000 armies multiple times in 1 autoattack. It happens, the statistics don't lie. Quit complaining.

[Streaker]

The dice are just unfair today.

I lost my entire stack on taking someone's 1's all the way to his castle, and I had 22 against his 49. I won with 19 left and won easily.
That's just unfair

[RADAGA]

The dice are just unfair today.

I lost my entire stack on taking someone's 1's all the way to his castle, and I had 22 against his 49. I won with 19 left and won easily.
That's just unfair

It is. kill 49 and lose 2 is just as surreal as the other way around. You just think it is normal because your perceptions are weird.

[Timminz]

You just think it is normal because your perceptions are weird.

Hey now! You're not allowed to use that argument!

[RADAGA]

You just think it is normal because your perceptions are weird.

Hey now! You're not allowed to use that argument!

I know, it is just evil because it cannot be contested

[RADAGA]

Lame, as usual.

21x12 - end 2x7

14x10 - end 2x9

16x7 - end 1x1

10x10 - end 2x4

All in a single round, all against the same (higher ranking, paying) opponent.

Is he Random.org owner? Or he bought a premium set of dice for his defence rolls?

After those failed attacks, another one came and stole my kill. Once again....

[The Neon Peon]

I'm going to make a radical assumption here, it's in several parts, so please bare with me:
1. When you attack with a 4, you roll 3 dice
2. When you attack with a 3, you roll 2 dice
3. When you attack with a 2, you roll 1 die
4. When a defender has 1 troop, he/she rolls 1 die
5. When a defender has more than 1 troop, he/she rolls 2 dice

Assuming all that to be correct, this:

14x10 - end 2x9

Is impossible.

Other than that, if you want anyone to care, you should post a dice analyzer of hundreds of thousands of rolls. On my old laptop, I had it get up to somewhere around 700,000-800,000 rolls with me winning 0.2% more than probable. As far as I'm concerned, you're getting average dice in your other games, seeing as you come here to complain about every bad roll, yet only do so once a week. I can only assume that all the other rolls you had that week were good.

[RADAGA]

Other than that, if you want anyone to care, you should post a dice analyzer of hundreds of thousands of rolls. On my old laptop, I had it get up to somewhere around 700,000-800,000 rolls with me winning 0.2% more than probable. As far as I'm concerned, you're getting average dice in your other games, seeing as you come here to complain about every bad roll, yet only do so once a week. I can only assume that all the other rolls you had that week were good.

Yes, bu tthe bad rolls come is sequence, ALWAYS.

I simply cannot accept it is perfectly normal to lose 11x3 up to 2x3 (yesterday) and today, attacking 3x1 I got a 6,5,1 x 6, another 6,6,2 x 6 another 4,2,1 x 6 and another 1,1,6 x 6 defending, and then moving toward another attack, manage to get a 3x2 rolling 6,5,2 and the defender get a 6,6 - just too many sixes to be believable.

21 dice - should wield 3 or 4 of each number, in the average. - and it got twice as expected sixes. not one more or one less, but TWICE

last week I posted an autoattack where defence got TWICE the expected wins as well, pooling over 70% of the results.

Autoattack and observe - it ALWAYS come like that - a row of 3-7 double wins, or a roll of 3 - 7 double losses, with negligible exceptions.

In large numbers, you get a lot of sixes in a row but a lot of others in a row as well, and get a good average, compatible with the statisticals expectancies.

Again, if the numbers tend to come in streaks and ties always favor defence, it gets complicated.

[RADAGA]

Also, look at those, from Random.org itself>



This hardly looks random to me, it looks like a pattern that tends toward the lower numbers in this particular case. Tomorrow they may move to higher numbers, and then the average will be beautiful. But today is a good day to expect lots os 0.1s and 0,2s - a disaster for the tie-defence system...

[The Neon Peon]

I simply cannot accept it is perfectly normal to lose 11x3 up to 2x3 (yesterday) and today, attacking 3x1 I got a 6,5,1 x 6, another 6,6,2 x 6 another 4,2,1 x 6 and another 1,1,6 x 6 defending, and then moving toward another attack, manage to get a 3x2 rolling 6,5,2 and the defender get a 6,6 - just too many sixes to be believable.

You're right. I wouldn't accept something that is completely impossible such as magically losing 9 troops when your opponent rolls 2 at a time and wins them all. Seriously, stop making up bs dice stories, or at least use outcomes that are possible to obtain by rolling those dice.

21 dice - should wield 3 or 4 of each number, in the average. - and it got twice as expected sixes. not one more or one less, but TWICE

Dang, twice as many sixes as expected in a couple of dice. That's pretty random. Luckily, after a thousand dice, you get less than 1% difference so it won't matter at all.

last week I posted an autoattack where defence got TWICE the expected wins as well, pooling over 70% of the results.

And of course, you have no evidence to prove it, the 30% you did not pool were perfect dice by you, and you'll come around here with yet another hypothetical case where you lose 9 troops 2 at a time.

Autoattack and observe - it ALWAYS come like that - a row of 3-7 double wins, or a roll of 3 - 7 double losses, with negligible exceptions.

In large numbers, you get a lot of sixes in a row but a lot of others in a row as well, and get a good average, compatible with the statisticals expectancies.

Again, if the numbers tend to come in streaks and ties always favor defence, it gets complicated.

Seriously, take screenshots. I don't care about bs examples that you're posting that come up to outcomes that are impossible. And in case you haven't noticed, no one gives a sh*t about any dice story you have because you're not posting 80% of the rolls you make.

You come here to rant every time you have bad dice, and any time that you have good dice, you're still to mad about the bad ones you just had to notice. Seriously, stop posting every 3v1 you've ever lost. No one will care until you post ALL of your dice, with not one good roll missing.

Btw, that will never happen for one of several reasons:
1. You'll continue saying the same crap over and over again providing impossible examples. or...
2. You'll continue posting every bad roll you ever get and leave out any good ones. or...
3. You'll leave the site. or...
4. You'll actually try to prove something using every roll that you make and won't be able to say anything because you won't find a streak of more than 100 rolls that proves your point.

[The Neon Peon]

Btw, I found it rather funny that you chose to ignore my pointing out of how it is impossible to lose an odd number of troops while losing 2 troops at a time in the previous post and then came up with another case that did the same thing this time.

Thanks for the laugh. I appreciated it.

[RADAGA]

Btw, I found it rather funny that you chose to ignore my pointing out of how it is impossible to lose an odd number of troops while losing 2 troops at a time in the previous post and then came up with another case that did the same thing this time.

Thanks for the laugh. I appreciated it.

Yes, so I made one or two mistakes, just because I dont have the bad roll and instantly come to complain, but keep making my turns.

As for the "stop posting made up bs posts" I´d answer you what you deserve to hear, but I dont want to be banned.

As for the "no evidence" of the 70% loss - I DID cut-paste the autoattack in question, you had no trouble checking and accused me of making them up once more.

As for ignoring and appreciating - I appreciated you ignored data from your GOD (Random.org) itself, just to make personal attacks towards me.

As for those personall attacks: you are the one rolling each dice yourself? If so, apologies for making personal attacks myself towards you.

BUT IF YOU DONT ROLL THEM YOURSELF, shut up!- you know why? Becuase I complain - on a GIVEN space to complain - about something abstract and about the mechanics of the site.

I dont accuse ANYONE here of acting dishonestly or anything similar. You, on the other hand, did not hesitated in doiing so, right? You dont like my rants about the dice - you dont come and read them - i dont like anchovies - I dont go where they are the main dish. If I were you, however, I think I would have to GO to those places and tell their owners anchovies suck because they are a bs fish. And tell them they should stop selling those bs fish.

This is a topic to complain and rant about the dice - you dont like it, you go elsewhere. You dont like the arguments, STICK TO COMPLAIN ABOUT THE ARGUMENTS

you dont like the person posting them - complain IN PRIVATE

in other words - try to be polite, I know it is VERY HARD, but try to remember what your mother (hopefully) taught you. If she didn´t, then I am truly sorry for your misery...

Have a nice day.

oh, by the way, I did post a while ago several dice analyzer results, so, for your made up bs argument of>

...because you won't find a streak of more than 100 rolls that proves your point.

try this - viewtopic.php?f=4&t=69749&p=1885871#p1885871

from 775 / 417 win/loss 3x1 up to 997 win x 549 , in six steps, loss showing a steady INCREASE of discrepancy between expected results and actuall rolls.

ANY statistic book would say that you should DECREASE the discrepancy between real and expected when you increase the universe of observations.

[The Neon Peon]

ANY statistic book would say that you should DECREASE the discrepancy between real and expected when you increase the universe of observations.

And again you display a fundamental misunderstanding of what "random means" ...

I'll just point something out again from previously in this thread:

sigh... what did I say about using logic in this thread?

Now he's gonna come back, say something to the effect of "you're an idiot," ignore most or all of what you said, and repeat something he's said before.

2 out of 3... not bad.

Still 2 out of 3, although a different two. I should really listen to my own advice as the same thing happened again. Had I just said "Shut the F*** up." in my first post, your last post wouldn't have been any different, and we'd have saved some time.

I just thought that maybe someone managed to explain to you what something being random means, but clearly that has not happened yet. I'll be back in another 4-5 months to see if someone manages to do so by then.

[The Neon Peon]

As for ignoring and appreciating - I appreciated you ignored data from your GOD (Random.org) itself, just to make personal attacks towards me.

Off topic: Maybe you're not a religious person, but that is blasphemy and is "taking the Lord's name in vain."

The only thing I said in my last post is that you've been making up your data, even if you do want to accuse me of flaming (for which there is a button right next to the quote one in the shape of a triangle), there's absolutely no reason to do so through blasphemy. I know you might not take things like that seriously, but many people do, including myself.

[RADAGA]

start 27x8 end 3x4 what are the odds.... well, here 100% - every players have had it´s share of it. ============ just another day at CC.

[soylentpurple]

Off topic: Maybe you're not a religious person, but that is blasphemy and is "taking the Lord's name in vain."

The only thing I said in my last post is that you've been making up your data, even if you do want to accuse me of flaming (for which there is a button right next to the quote one in the shape of a triangle), there's absolutely no reason to do so through blasphemy. I know you might not take things like that seriously, but many people do, including myself.

Off topic: how is blasphemy against the christian god even possible, since in the bible he impregnates his own mother, which can only be considered incest? Comparing god to a random number generator is hardly worse than that.

On topic: you still haven't addressed the random.org sample.

edit: whoops, you're american, which means you are about to get really mad

[The Neon Peon]

Off topic: Maybe you're not a religious person, but that is blasphemy and is "taking the Lord's name in vain."

The only thing I said in my last post is that you've been making up your data, even if you do want to accuse me of flaming (for which there is a button right next to the quote one in the shape of a triangle), there's absolutely no reason to do so through blasphemy. I know you might not take things like that seriously, but many people do, including myself.

Off topic: how is blasphemy against the christian god even possible, since in the bible he impregnates his own mother, which can only be considered incest? Comparing god to a random number generator is hardly worse than that.

You lost me entirely. Never heard any of that before.

On topic: you still haven't addressed the random.org sample.

It's been addressed 50 times in this thread already. His 1k rolls don't even begin to compare with the stats of rolls of hundreds of thousands of rolls others have previously posted in the forums showing almost identical stats to the ideal ones.

edit: whoops, you're american, which means you are about to get really mad

Actually, I'm Ukrainian. I just happen to live here for the time being.

[RADAGA]

The random.org sample (the graphic) was taken from random.org itself, and consists of 100.000 numbers.

As for the blasphemy - send my name to the pope to be excomunicated.

And it is STILL weird... anyway...

here is an actual auto-attack I just made:

Combat (2 Defense) (0 double A / 0 tie / 1 double D)

Denmark: * 4 * 6 * 5
- vs -
Faroe Islands: * 6 * 4

Combat (1A 1Def) (0 double A / 1 tie / 1 double D)

Denmark: * 1 * 5 * 3
- vs -
Faroe Islands: * 5 * 2

Combat (0 double A / 1 tie / 2 double D)

Denmark: * 4 * 1 * 1
- vs -
Faroe Islands: * 2 * 1

Combat (0 double A / 1 tie / 3 double D)

Denmark: * 3 * 6 * 1
- vs -
Faroe Islands: * 4 * 3

Combat (0 double A / 1 tie / 4 double D)

Denmark: * 3 * 1 * 4
- vs -
Faroe Islands: * 4 * 3

Combat (0 double A / 1 tie / 5 double D)

Denmark: * 1 * 2 * 3
- vs -
Faroe Islands: * 1 * 5

Combat (0 double A / 1 tie / 6 double D)

Denmark: * 3 * 6 * 2
- vs -
Faroe Islands: * 4 * 6

Combat (0 double A / 1 tie / 7 double D)

Denmark: * 2 * 3 * 2
- vs -
Faroe Islands: * 6 * 4

Combat (0 double A / 2 tie / 7 double D)

Denmark: * 4 * 1 * 6
- vs -
Faroe Islands: * 4 * 4

Combat (0 double A / 3 tie / 7 double D)

Denmark: * 4 * 3 * 4
- vs -
Faroe Islands: * 3 * 6

Combat (0 double A / 4 tie / 7 double D)

Denmark: * 6 * 1 * 2
- vs -
Faroe Islands: * 5 * 2

Combat (0 double A / 4 tie / 8 double D)

Denmark: * 3 * 1 * 2
- vs -
Faroe Islands: * 5 * 4

Combat (0 double A / 4 tie / 9 double D)

Denmark: * 4 * 2 * 1
- vs -
Faroe Islands: * 3 * 4

NO double attack succeded. only FOUR ties, and NINE double defences. Thirteen rounds, and the sample that shouild accour 29% of the time almost 60% - twice as much.

[Timminz]

you still haven't addressed the random.org sample.

Did someone forget to log into their proper account before posting?



Anyway, to address the graphic from random.org. Do you have any idea what the graph represents? I don't, but I suspect that it's not as damning as you seem to think. It's listing intervals, not absolute values. If it's representing the frequency at which the different intervals between values present themselves, then it makes absolute sense that there are much more of the lower values, and fewer of the higher values. Large intervals require a low number, followed by a high one, or the opposite, whereas a low interval will present itself whenever a low number is followed by another low number, or a high followed by a high, or a mid value... and so on. As you can see, you SHOULD be seeing more lower intervals, in a truly random sequence.

Besides all that, do you really think random.org would willingly put up a graph that definitively proved their service was useless?

[RADAGA]

you still haven't addressed the random.org sample.

Did someone forget to log into their proper account before posting?

Oh, PLEASE, do make a formal multi account accusation! I am begging you, make my day. I would LOVE to see the outcome of such.

as for them posting things that disprove their service, I dont know, but those graphs are strange>

[spiesr]

http://www.random.org/statistics/frequency-monobit/copenhagen-hw1/2010-03-01.png

Thus graph is not really relevant to the CC dice in any way. CC does not go to random.org to generate dice every time a roll is made. It reads from a large dice file that was generated one time from random.org. This file has been analyzed in the past and been shown to not have any real significant bias in it.

[Doc_Brown]

As was pointed out earlier, you shouldn't complain about those graphs if you're not aware of what they actually mean. I have a BS in math, and I couldn't figure out exactly what they're plotting without trying to dig through the values. Certainly they're not just plotting a distribution of the produced values. The graphs they show are based on statistical frequency analysis of their random numbers. The following graphs all represent analysis of the exact same data:

Click image to enlarge.

Click image to enlarge.

Click image to enlarge.

Just because the last one looks flat and the first two don't doesn't say anything about how random the numbers are if you don't know what's actually being plotted. It's the mathematical equivalent of saying that f(x)=3x^2 can't be a quadratic function because the graph of its derivative (f'(x)=6x) is a straight line and its antiderivative (F(x)=x^3) is an odd function.

Some people have been pushing for an alternative random number generator. Here is a plot of how the random numbers from the PHP rand() function are distributed:

Click image to enlarge.

Here is the equivalent distribution from random.org:

Click image to enlarge.

Your other alternative is to have a deterministic dice algorithm that would ensure that all rolls fit the proper distribution. However, it wouldn't be difficult to reverse-engineer such a system and predict what the rolls should be in a given attack based on the previous rolls. In a truly random sample, a given set of rolls are completely independent of all previous rolls (the exact opposite of random numbers). The expectation values for various outcomes are not models that the random numbers are supposed to conform to. Rather, they're the statistical outcomes derived from the random samples.

Instead of posting a single bad attack sequence that went bad, why don't you record every single roll you make for a full week. Post the results in here and we can determine how bad your luck actually is.

[RADAGA]

2010-03-09 13:48:24 - RADAGA cashed in a group of Troisvierges, Wiltz, and Bourglinster worth 10 troops
2010-03-09 13:48:24 - RADAGA got bonus of 2 troops added to Troisvierges
2010-03-09 13:48:24 - RADAGA got bonus of 2 troops added to Wiltz
2010-03-09 13:48:30 - RADAGA deployed 9 troops on Echternach
2010-03-09 13:48:56 - RADAGA deployed 9 troops on Mersch
2010-03-09 13:49:05 - RADAGA assaulted Redange from Feulen and conquered it from Commander62890
2010-03-09 13:49:15 - RADAGA assaulted Capellen from Redange and conquered it from Commander62890
2010-03-09 13:49:37 - RADAGA assaulted Bourglinster from Mersch and conquered it from Commander62890
2010-03-09 13:50:44 - RADAGA assaulted Grevenmacher from Echternach and conquered it from Commander62890
2010-03-09 13:50:52 - RADAGA assaulted Remich from Grevenmacher and conquered it from Commander62890

I had already two groups of 8 armies on three territories.

So I used 18+8+8 = 34 armies. I used ALL of those to conquer those FIVE lands, each guarded by a single defender.

Funny thing this guy is a premium captain, as usual. I even made an autoattack 11x1 just to advance 4 armies, after a massacre including 4 sixes in a row for the defender, among other wins.