Streaky Dice Question

[RADAGA]

Game 9239154

17 armies, to fighr 1; 1; 1; 1 = I ended with 6, leaving 1 behind each conquerings, 9 armies left, lost 8 to kill 4 - typical.

dice stats

opponent dice, total
last 5 rolls - 7 / 5 /6 / 10 / 16 / 11 (18 below 3 - 37 above)

Ridiculous!

[natty dread]

Overall battle outcomes
RADAGA 4.6 vs all opponents
3 vs 2 3 vs 1 2 vs 2 2 vs 1 1 vs 2 1 vs 1
Luck Kills Losses Luck Kills Losses Luck Kills Losses Luck Kills Losses Luck Kills Losses Luck Kills Losses
Assault 0% 1734 1500 0% 987 516 +5% 15 19 -3% 46 38 -1% 1 3 -5% 19 33
Defend 0% 1667 1995 -1% 474 977 0% 66 42 -10% 31 65 +10% 21 4 0% 39 28
Total 0% 3401 3495 -1% 1461 1493 +1% 81 61 -7% 77 103 +8% 22 7 -2% 58 61

[RADAGA]

Again, I know it AVERAGES out.

but what is the point, when you win 10 times in a row a 3x1, and then lose six in a row the next 3x1 you attempt

the average results are fine, but that game in particular is ruined because of a ridiculous streak that is totally surreal and should not be seen.

again, today, I won a bunch of attacks, but I did lost five in a row for a 3x1.

"oh, but it averaged out, you see"

yeah, of course... but say that toi that stack of 8 armies that got reduced to 2 trying to kill a 1

"yeah, but your overall stats show...."

I am not complaining about my overal stats, ranking or anything. I am just perceivingh something fishy, and I am trying to point it out for you, I dont expect anything to bo "better" for me if thay are fixed, in the long run. I just think that, if there is someting weird, it cant simply be dismissed by a "it´s random, it COULD happen". Heck,. anything COULD happen. you can log one day, and roll a thousand "ones" in sequence. It´s random.

But "it could happen" cannot be translated automatically to "since it could happen, everything is just fine"

[natty dread]

but what is the point, when you win 10 times in a row a 3x1, and then lose six in a row the next 3x1 you attempt

the average results are fine, but that game in particular is ruined because of a ridiculous streak that is totally surreal and should not be seen.

Winning 10 times a row is not what I would call "totally surreal". Nor is losing 6 times a row.

Streaks are naturally occuring artifacts in any stream of random numbers. The sooner you understand and accept this the happier you will be with your dice.

viewtopic.php?f=6&t=138002

[Metsfanmax]

Again, I know it AVERAGES out.

but what is the point, when you win 10 times in a row a 3x1, and then lose six in a row the next 3x1 you attempt

the average results are fine, but that game in particular is ruined because of a ridiculous streak that is totally surreal and should not be seen.

again, today, I won a bunch of attacks, but I did lost five in a row for a 3x1.

"oh, but it averaged out, you see"

yeah, of course... but say that toi that stack of 8 armies that got reduced to 2 trying to kill a 1

"yeah, but your overall stats show...."

I am not complaining about my overal stats, ranking or anything. I am just perceivingh something fishy, and I am trying to point it out for you, I dont expect anything to bo "better" for me if thay are fixed, in the long run. I just think that, if there is someting weird, it cant simply be dismissed by a "it´s random, it COULD happen". Heck,. anything COULD happen. you can log one day, and roll a thousand "ones" in sequence. It´s random.

But "it could happen" cannot be translated automatically to "since it could happen, everything is just fine"

Losing or winning streaks are not unexpected given the way the intensity cubes are compared. That is, I would find it odd if you were consistently rolling 1s in long streaks; you would be constantly losing because a 1 can never win. However, you could be getting perfectly random dice and being stymied by the fact that your opponent is getting random dice too, and his just happen to be collectively equal to or better than yours. If you see long streaks of a particular number or a particular set of low or high numbers, you can definitely report that as strange; but streaks of winning or losing are actually not as strange because they don't correspond to streaks in the sequence of numbers (but actually patterns in sets of five numbers, say, and you have to think about this much differently than how you think of streaks in a sequence of numbers if you're comparing one at a time).

[Hannibał]

Not whining, because risk is a game of chance. But is it just me that weekend morning rolls seem the worst? Ill wake up and play all my games, and a lot of times stop midway with terrible streaks of dice and come back in the afternoon much better. Anyone else?

Maybe its just I'm tired and make more mistakes after waking up? (Ex: Auto attacking)

[QoH]

Not whining, because risk is a game of chance. But is it just me that weekend morning rolls seem the worst? Ill wake up and play all my games, and a lot of times stop midway with terrible streaks of dice and come back in the afternoon much better. Anyone else?

Maybe its just I'm tired and make more mistakes after waking up? (Ex: Auto attacking)

This happened this morning to me. Couldn't win a roll for the life of me

[natty dread]

It's a coincidence. Also, confirmation bias.

[kachiun]

I usually have a hard time explain what I mean but Radaga, say that you were to throw a gigantic dice with a thousend sides. Each side reprecents a series of battle outcomes, a lot of them very unlikely to occour (making the dice sides uneven so that some sides are likelier to come up). But realy sence there are a lot of unlikely series it is not that unlikely that one of them comes up. Do you see the point that I'm trying to make?

[RADAGA]

Yes, I do see. The problem is something called laws of probability. You should see some weird results, if you roll enough times. The problem is that there is a percentage of chance you should see them. For example, the odds of losing six 3x1 in a row are 1 to 750

So you should see one of those occourences, or something even worse, once every 750 rolls, more or less.

So, the question is "when was the last time you lost more than FIVE armies in a row, on a 3x1 roll?" if the answer is once every seven hundred and fifity rolls, you are on the average. If you sey "once every 4 hundred rolls" that means you are very unlucky, and find 3 bad rolls when the averages say you should find 2 - not a big deal.

BUT if you answer "once every hundred rolls" it means you find them seven times more than what the probability should be. So,forget about biased perceptions and such. Look at your overal dice. You have two thousand 3x1 battles fought. How many times you remember hitting a 3x1 more than five times, losing every time? More than thrice in your whole CC life? That means either you are unlucky or the 1-in-750 event is happening more frequently than 1-every-750 times for everyone. If the first option is the real one, I suggest sacrificing a virgin to some god or goddess. If you dont like blood, sacrifice a virgin CD.

If the second option is the truth, meaning everyone see 1-in-750 events more often than once every seven hundred and fifty attacks they make, then something is wrong with the dice values accquiral method.

That is different than saying there is something wrong with the AVERAGE results, mind you. Example> losing three 3x1 in a row is 1-in-27 If you make 27 attacks, the odds for individual attacks is more or less 33% for the defender. So every 30 attacks you make, you should expect to lose 10 armies. BUT you should expect to lose more then 3 in a roll only ONCE.

Imagine the two scenarios below

Win, Lose, win, lose, win, win, win, lose, win, win, lose, lose, lose, win, win, win, lose, win, win, lose, win, win, win, win, lose, win, win, lose, win, win = 20 wins, 10 loses, 1 streak

Win, win, win, lose, lose, lose, win, win, win, win, win, lose, lose, lose, win, win, win, win, lose, win, win, win, win, lose, lose, lose, win, win, win, win = 20 wins, 10 loses, 3 streaks

BOTH have perfect 66,6%/33,3% win/lose distribuition, but the second string fail to comply with the streak chance odds.

[AndyDufresne]

RADAGA, I think you are maybe on of my favorite dice posters, just because you post about them so much! I think more than half of your posts are in Dice Topics.


--Andy

[RADAGA]

Yes, WAY more than half. I give you someone to toll with (me) every week. You should thank me. I am saving the world one sigh a time.

[AndyDufresne]

Sometime, I want to see a breakdown of your posting stats, in regards to: seeing how many times you mention the word 'dice' over the course of your time here on CC, in addition to how many Posts you've created in Dice Topics, and how many unique Dice Topics you've posted in total.


--Andy

[RADAGA]

Joined: Sun Jan 06, 2008 1:23 pm
Last visited: Mon Jun 27, 2011 4:09 pm
Total posts: 316 | Search user’s posts
(0.01% of all posts / 0.25 posts per day)

Most active forum: General Discussion(242 Posts / 76.58% of your posts)

Most active topic: One more thing about the dice.(159 Posts / 50.32% of your posts)

[RADAGA]

WTF! WHAT IS THIS??!!!!

Combat
DD3: 1 4 3
- vs -
Sdt. Gerd: 6 6

Combat
DD3: 2 1 6
- vs -
Sdt. Gerd: 6 4

Combat
DD3: 3 2 3
- vs -
Sdt. Gerd: 1 1

Combat
DD3: 5 5 5
- vs -
Sdt. Gerd: 1 6

Combat
DD3: 3 5 6
- vs -
Sdt. Gerd: 6 6

Combat
DD3: 2 4 6
- vs -
Sdt. Gerd: 6 4

Combat
DD3: 2 1 3
- vs -
Sdt. Gerd: 6 5

DEFENCE GOT EIGHT SIXES!!!! IN 14 DICE.
6,6,6,4,1,1,1,6,6,6,6,4,6,5

14x15 end 3x12
Average defence> 4,57
Percentage of sixes> 57,14%

With a single exception, every single defence roll either got TWO sixes or at the very least, one six. What happened? did the system rolled only 1 die for defence, and added another one au automatic six? If that was what happened, it missed one roll.

this is what I was talking about yesterday > you should not expect to see such things in a million rolls. Yet we do. Game lost because of that ridiculous result.

[liamrisksmart]

Yes, I do see. The problem is something called laws of probability. You should see some weird results, if you roll enough times. The problem is that there is a percentage of chance you should see them. For example, the odds of losing six 3x1 in a row are 1 to 750

I agree, but just to get your numbers right, its only one in every 643.914651

[RADAGA]

Sometime, I want to see a breakdown of your posting stats, in regards to: seeing how many times you mention the word 'dice' over the course of your time here on CC, in addition to how many Posts you've created in Dice Topics, and how many unique Dice Topics you've posted in total.


--Andy

No, the chance for me to post on any given thread is randomly decided and have equal chances of happening.

The fact you only see me posting complaining about dice is because the odds are decided by a true random system, it is only your perception (in this case, your sight) it have nothing to do with my will, it´s random, get over it.

[kachiun]

Yes, I do see. The problem is something called laws of probability. You should see some weird results, if you roll enough times. The problem is that there is a percentage of chance you should see them. For example, the odds of losing six 3x1 in a row are 1 to 750

So you should see one of those occourences, or something even worse, once every 750 rolls, more or less.

So, the question is "when was the last time you lost more than FIVE armies in a row, on a 3x1 roll?" if the answer is once every seven hundred and fifity rolls, you are on the average. If you sey "once every 4 hundred rolls" that means you are very unlucky, and find 3 bad rolls when the averages say you should find 2 - not a big deal.

BUT if you answer "once every hundred rolls" it means you find them seven times more than what the probability should be. So,forget about biased perceptions and such. Look at your overal dice. You have two thousand 3x1 battles fought. How many times you remember hitting a 3x1 more than five times, losing every time? More than thrice in your whole CC life? That means either you are unlucky or the 1-in-750 event is happening more frequently than 1-every-750 times for everyone. If the first option is the real one, I suggest sacrificing a virgin to some god or goddess. If you dont like blood, sacrifice a virgin CD.

If the second option is the truth, meaning everyone see 1-in-750 events more often than once every seven hundred and fifty attacks they make, then something is wrong with the dice values accquiral method.

That is different than saying there is something wrong with the AVERAGE results, mind you. Example> losing three 3x1 in a row is 1-in-27 If you make 27 attacks, the odds for individual attacks is more or less 33% for the defender. So every 30 attacks you make, you should expect to lose 10 armies. BUT you should expect to lose more then 3 in a roll only ONCE.

Imagine the two scenarios below

Win, Lose, win, lose, win, win, win, lose, win, win, lose, lose, lose, win, win, win, lose, win, win, lose, win, win, win, win, lose, win, win, lose, win, win = 20 wins, 10 loses, 1 streak

Win, win, win, lose, lose, lose, win, win, win, win, win, lose, lose, lose, win, win, win, win, lose, win, win, win, win, lose, lose, lose, win, win, win, win = 20 wins, 10 loses, 3 streaks

BOTH have perfect 66,6%/33,3% win/lose distribuition, but the second string fail to comply with the streak chance odds.

Well you are right there, it's sort of a lot... I have'nt realy noticed any big streaks at all (a few realy good ones maybe) witch is kind of weird. Though I have'nt realy been looking for streaks... I just downloaded some streak-counter so I guess I'll know soon.

I'd like to see the algorithm on witch this special comparasing cude is built. Is it in some forum or something? It would realy show everybody how the randomness "works".

[AndyDufresne]

...

I'd like to see the algorithm on witch this special comparasing cude is built. Is it in some forum or something? It would realy show everybody how the randomness "works".

19. How does the dice work?

We have a series of 50,000 true random numbers from Random.org. Each time the game engine processes an assault or auto-assault, it select a random spot in the series to read from using a pseudo-random computer function. Each time the game engine rolls the dice, the next number is read in sequence from the series (e.g. in a 3v1 attack 4 numbers are read sequentially). The series of 50,000 true random numbers from Random.org is replaced every hour. As of June 2010 we process 1,000,000 assaults per day.

--Andy

[RADAGA]

I'd like to see the algorithm on witch this special comparasing cude is built. Is it in some forum or something? It would realy show everybody how the randomness "works".

LAST 5 BATTLES
Defend -44% 2.40 0 6 4 0 0 0 - TEN DICE, only 2 and threes rolled. Get 10 d6 and roll, repeat until you only find 2 numbers, post here how long it took. If the forum does not exist anymore due to armaggedon or the death of the sun, call me at 555-3467

OVERALL TOTAL
Total 0% 3.50 2505 2718 2559 2588 2633 2579 TWOs have 8% more results than any other. and it is not such a small sample. It is a sample two thousand and five hundred times greater than the number of possible results.

[kachiun]

...

I'd like to see the algorithm on witch this special comparasing cude is built. Is it in some forum or something? It would realy show everybody how the randomness "works".

19. How does the dice work?

We have a series of 50,000 true random numbers from Random.org. Each time the game engine processes an assault or auto-assault, it select a random spot in the series to read from using a pseudo-random computer function. Each time the game engine rolls the dice, the next number is read in sequence from the series (e.g. in a 3v1 attack 4 numbers are read sequentially). The series of 50,000 true random numbers from Random.org is replaced every hour. As of June 2010 we process 1,000,000 assaults per day.

--Andy

Yes I know but if we had the algorithm we would bee able to decide the probability of winning a 3vs1 attack when attacking with x number of troops so we can now what the chance is of winning and how streaky we can expect the dice to be. Again in my games I have'nt really noticed any problem with streakyness.

[RADAGA]

I posted last 5 battles yesterday.

I post them today

-21% -41% -58% -42% -21% -34% -21% -39% -58% -42%

cute, isn´t it?

[TeeGee]

love dice stats



hate the dice

[jleonnn]

Ouch, that had to be painful. That's the problem with risk. It's so fun, and all based on luck... like gambling(I don't do gambling tho)

[Augustus Maximus]

I have no doubts about the streakiness of the dice. I have seen it go both for me and against me. To have a 25 vs 30 battle and have one side still have 20 armies left , or to go on a territory conquering spree and take 9 territories with 20 armies without any losses is a bit much. Both of these results can be aggrevating and/or unbelievable. If it happened occasionally people could accept it, but I see it happen in more games than I don't.