Dice Complaints and Various Suggestions to Fix Them

[mgconstruction]

Meh, Im not going to sit here & say I know random dice or Im a math person to say the least but the biggest problem I see is that the reason you can get rolls like we are seeing is that all CC members essentially share the same set of dice. Let me explain myself.

Your at home having a Risk night & maybe you set up 5 tables of friends with 5 Risk board games & 5 sets of dice. Each set of dice may get 200 rolls for the night so your chances of rolling 6's at each table numerous times are very low, however if you decided to only use one set of dice & pass them around from table to table you would now be rolling that one set maybe 1000 times and your chances of rolling 6's goes up. Maybe its not the randomness we should be looking at, maybe its the sharing of dice scripts with everyone else that truly takes the reality out our of rolls. Maybe there is a way to run scripts from game to game instead of across the board for everyone? I have no idea but everyone else's 10,000 rolls today should not be able to effect my 10 rolls tonight.

I do know that I will not be renewing my premium or giving away anymore with CC's current use of dice or scripts. Dice can & should play a part in a game where 2 players of the same strategy level are going at it and then the dice may make the difference for one or the other. But when you have a player who's strategy is much higher then another & cant even compete in the game due to dice, well that's a problem. Dice could make up for 1 or 2 mistakes by a player during a game but if they make up for continual mistakes through out the entire game, that's a problem & not very realistic.


MG

[Metsfanmax]

Your at home having a Risk night & maybe you set up 5 tables of friends with 5 Risk board games & 5 sets of dice. Each set of dice may get 200 rolls for the night so your chances of rolling 6's at each table numerous times are very low, however if you decided to only use one set of dice & pass them around from table to table you would now be rolling that one set maybe 1000 times and your chances of rolling 6's goes up. Maybe its not the randomness we should be looking at, maybe its the sharing of dice scripts with everyone else that truly takes the reality out our of rolls. Maybe there is a way to run scripts from game to game instead of across the board for everyone? I have no idea but everyone else's 10,000 rolls today should not be able to effect my 10 rolls tonight.

Simply put, this makes no sense. If the one set of dice is a fair set of dice, then everybody sharing the dice doesn't make rolling all sixes more (or less) likely, because fair dice give purely random results. It doesn't matter how often they're rolled. If a fixed number of rolls occur, then the amount of sixes you would expect is independent of how many (fair) dice you used to get that number of rolls.

[mgconstruction]

Your at home having a Risk night & maybe you set up 5 tables of friends with 5 Risk board games & 5 sets of dice. Each set of dice may get 200 rolls for the night so your chances of rolling 6's at each table numerous times are very low, however if you decided to only use one set of dice & pass them around from table to table you would now be rolling that one set maybe 1000 times and your chances of rolling 6's goes up. Maybe its not the randomness we should be looking at, maybe its the sharing of dice scripts with everyone else that truly takes the reality out our of rolls. Maybe there is a way to run scripts from game to game instead of across the board for everyone? I have no idea but everyone else's 10,000 rolls today should not be able to effect my 10 rolls tonight.

Simply put, this makes no sense. If the one set of dice is a fair set of dice, then everybody sharing the dice doesn't make rolling all sixes more (or less) likely, because fair dice give purely random results. It doesn't matter how often they're rolled. If a fixed number of rolls occur, then the amount of sixes you would expect is independent of how many (fair) dice you used to get that number of rolls.

Your correct but if the one set of dice being shared by all 5 tables just happens to come up 6,s only on say table 2 then it would appear to be unrealistic to the players at table 2. while in whole, pertaining to all 5 tables the number of times 6's that came up would seem normal.
Example: Using a different set of dice at each table and 6's came up 2 times in a game, no one would even notice. but if you used one set of dice for all 5 tables & 6's came up 10 times on table 2, everyone would notice.

[owenshooter]

if you have complaints go to www.random.org at that site you will find graphs, stats, etc, showing you the randomness of the dice...-the black jesus

[jefjef]

if you have complaints go to http://www.random.org at that site you will find graphs, stats, etc, showing you the randomness of the dice...-the black jesus

Do you mean go to the site that is selling their product/service for profit and using graphs and stats as a sales tool?

What really matters is what CC does with it...

[Night Strike]

I do know that I will not be renewing my premium or giving away anymore with CC's current use of dice or scripts. Dice can & should play a part in a game where 2 players of the same strategy level are going at it and then the dice may make the difference for one or the other. But when you have a player who's strategy is much higher then another & cant even compete in the game due to dice, well that's a problem. Dice could make up for 1 or 2 mistakes by a player during a game but if they make up for continual mistakes through out the entire game, that's a problem & not very realistic.

So the better strategy player (aka higher rank) should always win against a lower player in your situation. That is complete fixing of the game. This is a game where chance is involved, and you can't just remove all the chance simply because "better" players don't win every game.

[PLAYER57832]

Mods dont you dare delete this as you always do. Because it is a serious suggestion.

If the owner of the forum hires someone (a professional) to fix the dice,

To cut to the chase, this was done and is periodically done again... and again.

Truth is that most dice complaints come from people who really do not understand random or random distributions. If you are an exception, and know of a better method, then there is an open invitation to share that information. What gets deleted are the 1001 "these dice suck".. "I just got 10 rolls that were the same, so they cannot possibly be random..." or "I lost 20 games, so the dice are just wrong.." etc type threads. Those types of complaints indicate a clear failure to understand distributions and cannot be "fixed" unless the complainer wishes to be educated.

And, per the "education" bit, there is a link to explain, in detail how the formulas work. Reading that through is a good place to start if you wish to make a constructive dice complaint. Try it.

P.S., no I don't have a "mod" title, but I have been around longer than many of the mods who have answered and likely am more aware of some of these efforts than they. I am not, however authorized to get into much detail, just to refer to to where the answers lie. (note, my lack of authorization has to do with my often long-winded responses, etc., not lack of understanding of the issues )

[blakebowling]

if you have complaints go to http://www.random.org at that site you will find graphs, stats, etc, showing you the randomness of the dice...-the black jesus

Do you mean go to the site that is selling their product/service for profit and using graphs and stats as a sales tool?

What really matters is what CC does with it...

random.org is free. They aren't selling anything, and research done on the numbers was by third parties and is displayed by random.org.

The way dice are set up now, over all of the dice rolled (a possibility of five), the numbers are random and statistically come out to be about equal, however this statistic equality is based on all five dice, (or sometimes 2 or 3 or 4), and over the entire community.
My suggestion, is to have five seperate number generation files, (eg. Attacker 1, Attacker 2, Attacker 3, Defender 1, and Defender 2). This should make each die (attacker or defender) more correlated with the probability. However the correlation is based in terms over the whole site, and not each user.

[nebsmith]

you know I'm really tired or these threads, just once I'd like to see someone who complains post some dice analyser results to back them up, but they never do.

Could this be because anyone who can be bothered to check finds there is NOTHING WRONG WITH THE DICE.

does anyone realise just how many dice rolls they take when playing 50+ games and just how long it would take to play a similar number of real world games.

I have, i think, played 1000's of real world games of risk with real dice and have seen the real dice do extreme things. But because of how much slower sitting round a table and playing at a board is, these extreme events are separated by hours or days and so make a less lasting impression than when playing on a site like this.

people are particularly unsuited to assess randomness intuitively. One of the main features of our brains is to see and impose patterns often where no pattern exists. this is why there are so many "foolproof" gambling systems out there, none of which actually work.

If you don't think the dice used here are random use maths to show it.

[stahrgazer]

My suggestion, is to have five seperate number generation files, (eg. Attacker 1, Attacker 2, Attacker 3, Defender 1, and Defender 2). This should make each die (attacker or defender) more correlated with the probability. However the correlation is based in terms over the whole site, and not each user.

You're right, a separate generator for each cube, with each roll choosing only between one and six and resetting each time, would make the "random rolls" more consistent with what happens when you have dice in your hands - and be "more random" than selecting from a random set of previously-rolled outcomes.

My guess is, this change would cause every game to slog, and be devilishly expensive to program.

So, as I've stated in other threads, yeah, it's true the intensity cubes are not "as random" as dice would be...which explains "more trends of highs; more trends of lows; than generally would be seen in dice".

But. No one can predict which part of the file they'll be in, so people who say "the dice are already random" are correct. Still, those who say the dice have freaky trends that wouldn't happen as often with real dice, are correct, too.

I don't think it's worth the reprogramming and testing costs and hassle, and I wouldn't like the game to slog to fix it.

So, let the complaints continue: they have a legitimate basis and everyone needs to blow off steam sometimes. Stop telling the complainers they're wrong; they're not wrong, the cubes program isn't the same as dice in the hand because it picks from a subset of previous results rather than offer brand new "rolls" potential each time, for each die being rolled each time they are being rolled.

Instead of saying, "You're wrong," say, "Okay, you're right. The cubes aren't quite as random as dice would be. But since your opponent is under the same conditions, it's fair. Good luck on your next series of rolls."

It's rather like a three-legged race, you know. Each team has its legs tied, the only question is, is the knot tight enough this time or not? You don't know till you try to run, and when you do, you must adjust your stride (your strategy) according to how loose or tight the knot is at the time... how far and how fast you can run with what you get for the race.

So, OP enjoy blowing off some steam, then settle down and realize that EVERYONE on the site has the same potential handicap, it's not just you. So you can either choose to play or choose not to play based on that 'handicap.'

Meanwhile, my guess is, other online gaming sites use random.org or similar beasts, and just aren't as open about where the results come from.

p.s. at some point, if you stick around, you'll be the one hooting over having taken a 9 with a 2

[army of nobunaga]

I put my money were my mouth is... kick this up and get it approved, Ill help hire the math geek.

And if the math geek analyzed the current system and concluded that it was the best one can get with a computer, would you be happy and let it go? For some reason I'm thinking not.

then im thinking you dont know me.

[army of nobunaga]

oh I forgot
Night Strike

you are a mod here... it is your fucking job to be rude and add nothing to threads.


good work buddy.

[Woodruff]

You're right, a separate generator for each cube, with each roll choosing only between one and six and resetting each time, would make the "random rolls" more consistent with what happens when you have dice in your hands - and be "more random" than selecting from a random set of previously-rolled outcomes.

My guess is, this change would cause every game to slog, and be devilishly expensive to program.

So, as I've stated in other threads, yeah, it's true the intensity cubes are not "as random" as dice would be...which explains "more trends of highs; more trends of lows; than generally would be seen in dice".

I disagree. Making the "dice rolls" more spontaneous or "more recently created" does not affect their randomness of the result and would not be "more random" than the current method being used from random.org using atmospheric noise.

I personally favor this suggestion simply from the perspective of ending the controversy...but I fear the cost may be prohibitive.

[mgconstruction]

I do know that I will not be renewing my premium or giving away anymore with CC's current use of dice or scripts. Dice can & should play a part in a game where 2 players of the same strategy level are going at it and then the dice may make the difference for one or the other. But when you have a player who's strategy is much higher then another & cant even compete in the game due to dice, well that's a problem. Dice could make up for 1 or 2 mistakes by a player during a game but if they make up for continual mistakes through out the entire game, that's a problem & not very realistic.

So the better strategy player (aka higher rank) should always win against a lower player in your situation. That is complete fixing of the game. This is a game where chance is involved, and you can't just remove all the chance simply because "better" players don't win every game.

Seriously? Where in my post did I even remotely mention rank or taking out all of the chance of dice? I meant that dice should not be able to lop side a game so far one way or the other yet they seem to do so quite a bit. Read the part of my quote in bold because you obviously didn't read it before spewing this nonsense.

[Woodruff]

I do know that I will not be renewing my premium or giving away anymore with CC's current use of dice or scripts. Dice can & should play a part in a game where 2 players of the same strategy level are going at it and then the dice may make the difference for one or the other. But when you have a player who's strategy is much higher then another & cant even compete in the game due to dice, well that's a problem. Dice could make up for 1 or 2 mistakes by a player during a game but if they make up for continual mistakes through out the entire game, that's a problem & not very realistic.

So the better strategy player (aka higher rank) should always win against a lower player in your situation. That is complete fixing of the game. This is a game where chance is involved, and you can't just remove all the chance simply because "better" players don't win every game.

Seriously? Where in my post did I even remotely mention rank or taking out all of the chance of dice? I meant that dice should not be able to lop side a game so far one way or the other yet they seem to do so quite a bit. Read the part of my quote in bold because you obviously didn't read it before spewing this nonsense.

I wouldn't say this is the case in general, but I think it's widely understood that specifically in a 1-vs-1 game, it really comes down to the combination of going first and the dice. So in that particular instance, the dice really do control the game almost entirely, even to the point of making one player or the other "unable to compete"...it's really the nature of that type of game.

[Metsfanmax]

You're right, a separate generator for each cube, with each roll choosing only between one and six and resetting each time, would make the "random rolls" more consistent with what happens when you have dice in your hands - and be "more random" than selecting from a random set of previously-rolled outcomes.

My guess is, this change would cause every game to slog, and be devilishly expensive to program.

So, as I've stated in other threads, yeah, it's true the intensity cubes are not "as random" as dice would be...which explains "more trends of highs; more trends of lows; than generally would be seen in dice".

Yes, you've stated this opinion in other threads too, and it was wrong there as well. There's no logical reason why the current system is any different (in any meaningful way) than having a separate file to run through for each die, and there's no logical reason why that system is any different from actually generating a random number each time you call a die roll. If the list of previous outcomes was generated in a random manner, then the rolls we get are theoretically random. There's no way around that mathematical certainty. This follows because each time you read the next number on the list, it's functionally equivalent to rolling a die and getting a number between 1 and 6. If the list were only pseudo-random, then that would be false. But it's truly random, and so the act of reading the next number on the list is equivalent, for all purposes here, to rolling a die.

So, let the complaints continue: they have a legitimate basis and everyone needs to blow off steam sometimes. Stop telling the complainers they're wrong; they're not wrong, the cubes program isn't the same as dice in the hand because it picks from a subset of previous results rather than offer brand new "rolls" potential each time, for each die being rolled each time they are being rolled.

Instead of saying, "You're wrong," say, "Okay, you're right. The cubes aren't quite as random as dice would be. But since your opponent is under the same conditions, it's fair. Good luck on your next series of rolls."

If your assertion about the randomness of the dice were correct, then I'd agree with your description of the situation. But the assertion is wrong, and therefore so is your claim that the complainers have a leg to stand on.

[PLAYER57832]

you know I'm really tired or these threads, just once I'd like to see someone who complains post some dice analyser results to back them up, but they never do.

Could this be because anyone who can be bothered to check finds there is NOTHING WRONG WITH THE DICE.

does anyone realise just how many dice rolls they take when playing 50+ games and just how long it would take to play a similar number of real world games.

I have, i think, played 1000's of real world games of risk with real dice and have seen the real dice do extreme things. But because of how much slower sitting round a table and playing at a board is, these extreme events are separated by hours or days and so make a less lasting impression than when playing on a site like this.

people are particularly unsuited to assess randomness intuitively. One of the main features of our brains is to see and impose patterns often where no pattern exists. this is why there are so many "foolproof" gambling systems out there, none of which actually work.

If you don't think the dice used here are random use maths to show it.

Bears repeating.. and repeating... and repeating.

[jaimito101]

The dice are random, that's a fact. People collecting their dice roll's will tell you this. The only problem you may find is that it takes many dice & games to achieve this normal graph leading to many easy wins/losses. The only problem is that this, is but a side effect of true randomness, each dice is on its own and is technically not affected by the previous roll, resulting in many possible reppetitions in a row.

What you are asking is less fluctiations in this randomness, so that strategy is more prevalent. This can simply be achieved by having a fixed nr of 1,2,3,4,5,6's happening in a fixed nr, say 360 dice rolls. A new random pattern would then refresh after every set of 360 sitewide would be used (as to remaining "random").

This will fix the squewness of the dice, but this itself would also bring along countless complaints, now that dice are not "trully" random but predetermined to have the same amount of each nr thrown after 360 rolls. This though may be a future option strategy purists may want to use.

[natty dread]

Hey guys, I just lost an auto-assault today. It was a 14v3, I clicked auto and lost all 6 rolls straight on, leaving me with 2v3.

So why am I not running around shouting how "CC manipulates teh dices to cheat my points" or some other such drivel?

Maybe it's because I have some measure of self respect. Maybe it's because I'm not a 12-year old crybaby whose world collapses instantly if I get some bad luck in an online boardgame. Or maybe it's simply because I have no illusions of grandiose, that I'm such a great player that I shouldn't be able to lose any games. Or maybe it's because I actually understand what random means. Or that correlation doesn't imply causation.

Ok, so everyone seems to be an expert on statistics these days. "These rolls should just not happen!" "Ok these rolls should happen but not this often!" "Ok these rolls should happen but not all in a single day!" Give me a break. One of the key parts of the definition of random is that random cannot be predicted. Let's say you're flipping a coin. There's a 50% chance for any given toss to be either heads or tails. But the moment you have tossed the coin, the result no longer has any bearing on the following tosses. So even if you toss heads for 50 times, the probability for the next toss to be heads is still exactly 50%.

But then, logic has never been the forté of the wilfully ignorant martyrs.

Furthermore I find it funny how people with ranks of colonel and up are walking around claiming how "CC manipulates the dice and fixes games". I mean, the sheer stupidity of the whole premise is an insult to the intelligence of the human race. Why on earth would CC be fixing the outcomes of games? What possible benefit could there be other than to annoy players, driving away paying customers? But that aside, if CC is truly fixing or manipulating the games against these players on a constant basis, how on earth are they able to maintain their rank? Either they are cheating and using the conspiracy theories to justify it or they are so deluded as to think that they are actually master strategians, and they should be at #1 on the scoreboard if it were not for the evil CC stealing their points. In the real world though, not everyone can be best at everything. Maybe it's time to face the truth and just admit maybe you aren't all that excellent? I know I'm not. I'm probably an average player at best. That's ok, I'm good at other things (just ask my girlfriend, if you know what I mean... wink wink)

Logic doesn't even enter the equation here.

Lastly, for the OP... AON, I'm not saying this to bash you or anything, this is just my honest opinion... the idea of hiring a professional mathematician would be a waste of money. There's nothing a mathematician can do to fix something that is in the heads of players. Let's say we do it, and he says there's some flaw in the dice and lack fixes it. Will it keep people from complaining? No. There will be just as many people threatening to leave if they don't get the "old dice" or the "old old dice" back. Or let's say we do it, and the mathematician says the dice are fine and there's no way to improve them. Same result. People will keep throwing hissy fits on the forums and threatening to leave "as soon as my premium runs out."

[Juan_Bottom]

That is, before it changed 'the dice were truly random, but streaks were broken up. Now the streaks are left alone.'

At the risk of sounding crude, whoever told you that had no idea what they were talking about.

What I've noticed is that most people view "streaky" as "unrandom", which isn't true. Any single event does not influence any of the other events, which means the odds of hitting 6-6 defender dice are 1/36 every time. The old dice compensated for this by having a file that had equal ratios of the 6 numbers in the file, which is not random.

The result is simply streakier dice -

This is out of context, but the broader context was that the die are random.

[oldrisky44]

The topic is locked, so I couldn't do quote, but...

New "Intensity Cubes"
by lackattack on Mon Jun 21, 2010 12:20 pm

We haven't touched the dice... erm, intensity cubes since February 14, 2006 because they were pretty darn random (see here). However, thanks to some prodding and advice from Dako, sherkaner and jakewilliams I became convinced that it is worthwhile to make some alterations.

This is how the intensity cubes now work:
We have a series of 50,000 high quality random numbers from random.org
Each time the game engine generates a random intensity cube, the next number is read in sequence from the series (e.g. in a 3v1 attack 4 numbers are read sequentially)
When the last number in the series is read, we "rewind" and continue with the first number in the series

The advantages are twofold:
Each individual number in the series is used for both attacker and defender, so our intensity cubes cannot be biased for either side.
The series is stored in memory so the dice perform much faster. This makes a huge difference when auto-assault is used with large numbers of troops on both sides.

An interesting note - as of June 2010 Conquer Club processes 1,000,000 assaults each day!

lackattack
Webmaster


Posts: 6197
Joined: Sun Jan 01, 2006 11:34 pm
Location: Montreal, QC
Medals: 7


1,000,000 assaults per day. So the random number file is being recycled quite often. The "impossible" is bound to happen in that many permutations.

The hardest part about dealing with random numbers is that they are so darn unpredictable.

[Thezzaruz]

And why would these things be randomized anyways? If a true army of 1000 men attacked one of 200 would the 200 randomly defeat them?

If you don't like the basic premise that CC (and the game it is based upon) is built on then you are free to go and invent your own game that suits your own preferences.

[Tennisie]

The "streaky dice" problem has been addressed in a previous thread: http://www.conquerclub.com/forum/viewtopic.php?f=6&t=124113&hilit=dice+problem. Streaks can be "neutralized" by using many dice rolls during an attack because losing streaks tend to be balanced by winning streaks, thus evening out the overall result. More dice can be achieved by using manual Initial Troops setting or modifying the attack/defense system.

Manual Initial Troops allows you to create fewer and larger armies and thus roll many dice during each attack. The disadvantage of this approach is that it doesn't address games with automatic Initial Troops.

The current attack/defense system involves 3 (or fewer) dice attacking 2 (or fewer) dice, no matter how many armies the attacker and defender each have. I believe this is what upsets so many players. It certainly shocked me when I lost 87 to 0 during a single auto-attack against 3 defenders. I'm still waiting two years later for the corresponding winning streak LOL. Nevertheless, the attack/defense system could be modified to be similar to the Axis and Allies game's technique: one die for each and every army for both attacker and defender. Then, if you use only large armies, you get many dice rolls, thus neutralizing the streaks.

Of course, with randomness, streaks can never be eliminated, but that's okay because real life has apparent randomness also, no matter how much planning is done or how good a strategy is used.

[Night Strike]

Nevertheless, the attack/defense system could be modified to be similar to the Axis and Allies game's technique: one die for each and every army for both attacker and defender. Then, if you use only large armies, you get many dice rolls, thus neutralizing the streaks.

It could be, but then this site would no longer be what it is. It's similar to another board game. If lack wanted it to be similar to Axis and Allies, then he would have designed the site to be that way.

[Metsfanmax]

1,000,000 assaults per day. So the random number file is being recycled quite often. The "impossible" is bound to happen in that many permutations.

The hardest part about dealing with random numbers is that they are so darn unpredictable.

To get a rough estimate, 5,000,000 dice per day distributed uniformly over 24 hours is ~ 200,000 dice per hour. The dice list has 50,000 entries, so we can guess that a given list is run through about four times before it is replaced.