The bad luk of dices!

[Timminz]

yesterday a guy in my game said he lost 700vs200ish! that is a big fucking streak???!

U wanna bet we could throw real dice till forever and never lose this attack?

I'd sooner bet that that guy in your game was lying to you.

[danfrank]

the game is rigged in favor of the brown nosers and the likes...

[shadistic]

yesterday a guy in my game said he lost 700vs200ish! that is a big fucking streak???!

U wanna bet we could throw real dice till forever and never lose this attack?

Well, if you round it down it's basically a 7vs2, which might be unlikely but by no means impossible (how did they get that may troops anyways?)

It doesn't quite work that way. You have to think of it statistically, not arithmetically/algebraically.

The more events you have, the more the trend will gravitate towards the true probability of each event. Abnormalities become more pronounced in the smaller numbers.

Here's an example:
If you attack 7v2 and lose 2 on your first round, you're down to 5v2 and the probability remains the same.
If you attack 700v200 and lose 2 on your first round, you're down to 698v200 and the probability remains the same.

I agree with Timminz, though. Let's say the 700 was rounded up from 650, and the 200 was rounded down from 290. That's still rare, but closer.

The other reason people seem to remember bad rolls is that generally players don't attack "against the odds" - ie. attack 4v12, 5v18 etc when statistically there is a (admittedly vastly small) chance of these succeeding. This means they only experience bad losing streaks because they're not even giving the dice a chance to have a "bad" winning streak.

Yeah, that's the funny thing. It's more likely for a 10v20 to succeed than it is for a 20v10 to fail, but it happens a lot less because more people try 20v10 than 10v20.

[macbone]

Back in yesteryear when I played more face to face Risk, we'd have solo armies occasionally who'd outlast larger attacking stacks. We'd cheer for that brave soldier in Yakutsk who defeated the imperialist threat. =)

The last time I played the board game, I seemed to get bad dice when I needed it the most. It happens. You'll get the good dice eventually, too, when you can't seem to lose. =)

[PLAYER57832]

Do not auto assault when dice r bad.

Also if you buy premium you get the next tier of dice.

careful, I once made a sarcastic comment about paying Lack 'C notes" and a few people took it seriously... lol

[zerror]

Code: Select all

Probabilities of winning a dice roll in Risk
(various die combinations)				Attacker
										one die	two dice	three dice
Defender		one
				die	Attacker wins	41.67%	57.87%	65.97%
						Defender wins	58.33%	42.13%	34.03%
			two
				dice	Attacker wins	25.46%	22.76%	37.17%
						Defender wins	74.54%	44.83%	29.26%
						Both win one		n/a	32.41%	33.58%

simple as that... no matter if it is 700vs200 or 7vs7, same percentages apply. If you throw a coin in the air 99 times, and all 99 times it is tail, on the 100th throw you still have a 50 - 50 probability to be a tail.

[MichelSableheart]

simple as that... no matter if it is 700vs200 or 7vs7, same percentages apply. If you throw a coin in the air 99 times, and all 99 times it is tail, on the 100th throw you still have a 50 - 50 probability to be a tail.

Partially true. when it's 700vs200, you have just as much chance to lose two troops as when it's 7vs2. However, to lose the entire battle, you would need to lose 2 troops 350 times in a row in 700 vs 200, but only 3 times in a row when it's 7 vs 2.

In your coin analogy: before you start flipping, the chance that your first 3 flips are tail is much larger then the chance that your first 100 flips are tail.

[darrens99]

Dice do effect the game big time, but at the end of the day the best strategists will win more often than a bad strategist.

my advice, spend more time improving your strategy and less time dwelling on bad rolls because at the end of the day there is nothing you can do to beat chance

[RADAGA]

"Bad luck" is the common thing here when attacking.

right now: lost a 7x1 down to 1x1 (4 3x1, plus a 2x1 straight)
then moved to lose a 9x7 down to 3x3 (lost 2, won 2 won 2, lost 2 lost 2)
then moved to 16x13, down to 5x0 (won, but still, lost more than the "average")
then moved to a 9x10, down to 3x8

every single attack was a disaster. what are the odds? I bet the odds are smaller than the total dice I have ever played here.

[zerror]

"Bad luck" is the common thing here when attacking.

right now: lost a 7x1 down to 1x1 (4 3x1, plus a 2x1 straight)
then moved to lose a 9x7 down to 3x3 (lost 2, won 2 won 2, lost 2 lost 2)
then moved to 16x13, down to 5x0 (won, but still, lost more than the "average")
then moved to a 9x10, down to 3x8

every single attack was a disaster. what are the odds? I bet the odds are smaller than the total dice I have ever played here.

as I said before, every throw of dices is a separate case... someone said this dices are generated trough random.org and if that is true... you were just way unlucky (same here, I was 11x3, just to lose with 1x2)... but, that is the way world goes around... gl nest time

[Doctor Spin]

Shadistic wrote: "Strudel, it might not come immediately, and it might not even come by the time you finish a battle or two, but after you've played a few simple maps all the way through, the dice will, in all likelihood, be closer to 52% or so win rate, which is the probability. You can get Dice Analyzer, a browser add-on, and see for yourself."

Shadistic may, or may not, be correct. But he is very wrong to assume the dice are random. They are not...in fact they are so utterly non-random that I can't believe CC has been able to get away with it. Just because the rolls might even out over time, it does not make them random. It is very, very frustrating. I strongly believe that each player is assigned a pre-programmed roll sequence - and I do think that some players are given better sequences than others. It needs sorting out, as I for one am losing patience with this game, having played and enjoyed the board game version for over 30 years.

DS

[shadistic]

Shadistic wrote: "Strudel, it might not come immediately, and it might not even come by the time you finish a battle or two, but after you've played a few simple maps all the way through, the dice will, in all likelihood, be closer to 52% or so win rate, which is the probability. You can get Dice Analyzer, a browser add-on, and see for yourself."

Shadistic may, or may not, be correct. But he is very wrong to assume the dice are random. They are not...in fact they are so utterly non-random that I can't believe CC has been able to get away with it. Just because the rolls might even out over time, it does not make them random. It is very, very frustrating. I strongly believe that each player is assigned a pre-programmed roll sequence - and I do think that some players are given better sequences than others. It needs sorting out, as I for one am losing patience with this game, having played and enjoyed the board game version for over 30 years.

DS

Okay, wise guy, what's your theory?
How's it all work out? You tell us.

[macbone]

A few weeks ago, I was rolling some really sweet dice. Any idea why I was blessed with such favor from the powers-that-be?

[Revillo]

I just started and during the 6 games I finished so far everything happened. From a 13x3 to a 3x3 loose to a 10x3 win and I didn't loose a single unite. It can happen that you loose badly (even with real dice) or that you win totally luckily.

The best advice I got from this conversation is not to use auto assault and if you see that you loose badly stop.

[Echospree]

Shadistic may, or may not, be correct. But he is very wrong to assume the dice are random. They are not...in fact they are so utterly non-random that I can't believe CC has been able to get away with it. Just because the rolls might even out over time, it does not make them random. It is very, very frustrating. I strongly believe that each player is assigned a pre-programmed roll sequence - and I do think that some players are given better sequences than others. It needs sorting out, as I for one am losing patience with this game, having played and enjoyed the board game version for over 30 years.

DS

That's one of the dumber things I've ever heard in relation to how random numbers work. It always amazes me how poor people are at understanding such a (fairly simple) thing.

If the "pre-programmed roll sequence" is generated randomly, THEN ITS STILL RANDOM. As long as there's no way to know what your "list" is before you roll the die, it's still random.

Now, if you would like to show that they are non-random, you probably should be recording your rolls. Not just the total, but the order as well.

We know that the dice tend towards the true probabilities over enough dice. Enough people have tested it, we can assume this to be true. What's under discussion is whether or not the dice are "overly streaky". How do we test this? Record your dice for a large attack (700 v 200 would do), and record not only the total wins/losses, but also what order everything came in. Obviously, streaks are going to occur in this massive amount of data. What we then have to do would be to count the average steak length, and compare it to the expected average streak length for true randomness. That shouldn't be that hard, we just need someone collect the data. I'll gladly run the analysis if it'll quiet this rabble down.

[Muerto]

I have to say that this game is far from statistically sensible.

Anyone ever play a real game? What kind of streaks exist?
Reality doesn't show for one reason: the randomizer is flawed.

I suggest the game reinvent their randomizer to be a little more realistic. It's very unlikely for someone to outroll someone, with 1 dice, 10 times (it's happened to me) who is using 3 dice.

I'm agree with this! Logically, we can't loose more than 80% of the defender's troops when the attack have more troops than the defender. like 10 v 3 = only 2 lost and like 42 v 10 only 8 lost.

[Muerto]

Code: Select all

Probabilities of winning a dice roll in Risk
(various die combinations)				Attacker
										one die	two dice	three dice
Defender		one
				die	Attacker wins	41.67%	57.87%	65.97%
						Defender wins	58.33%	42.13%	34.03%
			two
				dice	Attacker wins	25.46%	22.76%	37.17%
						Defender wins	74.54%	44.83%	29.26%
						Both win one		n/a	32.41%	33.58%

simple as that... no matter if it is 700vs200 or 7vs7, same percentages apply. If you throw a coin in the air 99 times, and all 99 times it is tail, on the 100th throw you still have a 50 - 50 probability to be a tail.

but its not a 50 - 50 (100%) its a 700 - 200 (for 100% 78 - 22)

[Muerto]

Dice do effect the game big time, but at the end of the day the best strategists will win more often than a bad strategist.

my advice, spend more time improving your strategy and less time dwelling on bad rolls because at the end of the day there is nothing you can do to beat chance

If i win a game in 3 round, and the opponent can't win because of his dices, he loose 5 v 1 in each of the 3 turns... Don't tell he's a bad strategist That's the fucking random progr.

[nippersean]

To quote you

I'm agree with this! Logically, we can't loose more than 80% of the defender's troops when the attack have more troops than the defender. like 10 v 3 = only 2 lost and like 42 v 10 only 8 lost.


Why is that logical?
Please explain using random principles and your interpretation of chaos theory.

Answers in full please.

[Muerto]

To quote you

I'm agree with this! Logically, we can't loose more than 80% of the defender's troops when the attack have more troops than the defender. like 10 v 3 = only 2 lost and like 42 v 10 only 8 lost.


Why is that logical?
Please explain using random principles and your interpretation of chaos theory.

Answers in full please.

If you consider that are in 1v1 troops, its a 50% for each player. But if we cumulate the 50% for more troops, we have more than 50% to win! 2v1 its 75% v 25%.
But before i had suggest only 80% of the defender and only if you have more troops than the defender: you loose 80% of the defender troops. Thats for the autoRoll !

If 2 men attack 1 man (3v1); the end is at 30%: 1 man stand for the attackers, at 20%: 1 man stand for the defender or at 50%: 2 men stand for the attackers. Thats the logic!

[SirSebstar]

Dear muerto, please ask the money back from your school. They did a lousy job with your math skills.

a 2 vs-1 attack (essencially 1 dice vs 1 dice) really does behave like in the following text

Code: Select all

Probabilities of winning a dice roll in Risk (various die combinations)            
Attacker       
one die 	two dice 	three dice
Defender one die   
Attacker wins          			41.67%   	57.87%   	65.97%
Defender wins           			58.33%   	42.13%   	34.03%

Defender two dice   
Attacker wins         			25.46%   	22.76%   	37.17%
Defender wins          		 74.54%   	44.83%   	29.26%
Both win one         	 		n/a      	32.41%   	33.58%

It is NOT 50% chance. just write it out if you dont believe me. there are 36 (= 6 times 6) combinations. write it out and see it is correct.
e.g. say the attacker throws a 1, then no matter what the defender throws, its always a win. equals dice is always a win for the defender, therefor the defender has the advantage when defending 1 vs 1. therefor you should not attack 1 vs 1... simple as that

[Muerto]

Dear muerto, please ask the money back from your school. They did a lousy job with your math skills.

a 2 vs-1 attack (essencially 1 dice vs 1 dice) really does behave like in the following text

Code: Select all

Probabilities of winning a dice roll in Risk (various die combinations)            
Attacker       
one die 	two dice 	three dice
Defender one die   
Attacker wins          			41.67%   	57.87%   	65.97%
Defender wins           			58.33%   	42.13%   	34.03%

Defender two dice   
Attacker wins         			25.46%   	22.76%   	37.17%
Defender wins          		 74.54%   	44.83%   	29.26%
Both win one         	 		n/a      	32.41%   	33.58%

It is NOT 50% chance. just write it out if you dont believe me. there are 36 (= 6 times 6) combinations. write it out and see it is correct.
e.g. say the attacker throws a 1, then no matter what the defender throws, its always a win. equals dice is always a win for the defender, therefor the defender has the advantage when defending 1 vs 1. therefor you should not attack 1 vs 1... simple as that

But why when I play for real, my dices don't loose -2 army more than 2 consecutively and here, thats happens more and more and more? that's why i suggest by men and not only by dices. The random program of this dices is not like real dices. Never, in real, I have loose when I attack with 20 or more against 5 or less. Here... thats suck to many time!

[SirSebstar]

trust me, unless you use crooked dice or cheat, in real life it will also happen.
the alternative would be for you to install software to count ALL your dices. The good and the bad. And you will see your 4-20 win more often then a 20-2 will loose.

[flexmaster33]

Why don't the dice behave just as I expect...it seems they always favor my opponent. This has nothing to do with my poor play or punching Auto Assault on a whim. The fates hate me

lol

[The Bison King]

Never look directly at the dice, that is bad luck. It is much better luck to keep your eye on the territories.