The Improbable Always Wins....WTF?

[natty dread]

Well that's funny. Because In the 20 turns I took today, I managed to roll 111 twice. So what you are saying is that the chance of that happening is 1 in 2,500,000,000 yet it manages to happen all the time....odd. It seems that the title of this thread is strangely appropriate then.

Each set of off/def gets picked from a different location. Not each dice or roll of the dice.

How is the chance of rolling 111 1 in 2,500,000,000 or are you saying the chance of that happening 2 times is 1 in 2,500,000,000 either way that is not even close to the odds that i would think.

you still havent answered this, could you please do so? i really want to know how you think rolling 111 twice should only happen 1 in 2,500,000,000...

In fairness to MNDuke, I believe his quote there was in response to one by natty_dread, who initiated the "1 in 2,500,000,000" number. I think MNDuke just went with that assertion, under the assumption natty had actual numbers behind it.

I did, but not the way MNDuke understood it to mean.

The location where the dice are read in the file of 50000 is chosen randomly for each assault. We know this. MNDuke posited that the dice are more streaky because the location is chosen randomly, and provided the reasoning, that if the same location is chosen many times in a row, then that creates a streak, which adds to the streakiness.

Not even going to how this is illogical in other ways, I presented this counterargument: for the same location to be chosen just twice in a row from among the 50000, the odds are that 1 in 2,500,000,000 I mentioned (1:50000 ^ 2).

MNDuke then tried to counter that with "but I rolled 111 twice in the 20 turns I took today, how is that possible if the chances are that low" and then I tried multiple times to explain to him in various ways, that even if he rolls the same numbers, it doesn't necessarily mean that the random stream has chosen the same location, using the reasoning that the random file contains multiple instances of each possible combination of 5 dice.

After this, MNDuke started accusing me of trolling. Go figure.

[Woodruff]

this is realy frustrate how set up works. in game 8453944 I had 6 in Cairo and 6 in Kerac. I assaulted with 6 from Cairo 3 in Ascalon, result 3:2. I assaulted with 6 from Kerac 3 in Jaffa, result 3:2. then I assaulted with 3 from Cairo 2 in Ascalon, result 2:1. my next attack with 3 from Kerac 1 in Ascalon, result 2:1. then with 2 from Cairo assaulted 1 in Ascalon, result 1:1. my next attack with 2 in Kerac against 1 in Ascalon, result 1:1.

so how stupid is this set up, I ask? in the same round, in the same game 6 units can not beat 3. and then all these attacks against Ascalon and I lost?

btw, I remeber that one time I assaulted 6 agaisnt 6 and I won 4:1. so is this realy all right? I think no.

is any chance that this will be solve in the future? and I´m not alone who is thinking that set up/dice of attacks works so stupid too often.

You do realize that dice don't always behave the way you want them to, right? Particularly given the very limited number of games that would quantify your experience.

[theBastard]

You do realize that dice don't always behave the way you want them to, right?

no, it has nothing with "way as I want". and it is nothenig about only me. it is about abnormal set up. where is any logic that one time I win with 5 against 5 and one time I lost 6 against 3? what can say guy who was beated with my 5 against his 5?

Particularly given the very limited number of games that would quantify your experience.

again no, there were enough these things...

[SirSebstar]

You are absolutely correct. We fudge the dice so when the mods play, it does not look as if we are this bad players..
Also, you are correct a statistical analyzes of 3 particulary handpicked dicerolls is statisically significant..
oh right, fail..!

3dice vs 2 dice fail roughly (very roughtly) 1/3th of the time, they kill 1 and loose 1 1/3th of the time and they will win 1/3th of the time. Meaning that if you thow a 6-3 you have only to loose 1 sets of rolls totally and middle the other. This (or worse) WILL happen about 44% OF THE TIMES YOU ATTACK 6-3..

[theBastard]

You are absolutely correct. We fudge the dice so when the mods play, it does not look as if we are this bad players..

about what you are speaking here? where I mentioned mods or fudge?

Also, you are correct a statistical analyzes of 3 particulary handpicked dicerolls is statisically significant..
oh right, fail..!

these 3 was only what become to my mind quickly. there were more similar things, as it is written "again no, there were enough these things..."

3dice vs 2 dice fail roughly (very roughtly) 1/3th of the time, they kill 1 and loose 1 1/3th of the time and they will win 1/3th of the time. Meaning that if you thow a 6-3 you have only to loose 1 sets of rolls totally and middle the other. This (or worse) WILL happen about 44% OF THE TIMES YOU ATTACK 6-3..

but two times in the same round, in the same game? hm, ok and what you could tell me about others attacks? what about victory 5:5? is this also normal?

[JelleR]

1) MNDuke and theBastard never finished any sort of education. Well, maybe highschool?

2) They are trolling.

One of these is correct.

[theBastard]

1) MNDuke and theBastard never finished any sort of education. Well, maybe highschool?

2) They are trolling.

One of these is correct.

and what are you any scientist?

[JelleR]

Does it really matter?

I just refuse to believe that anyone that finished some sort of education has these elementary questions about probabilities.

[WestWind]

He cannot answer this... All the people who complain about the dice, never seem to come up with any hard evidence to show they are "broken". They only remember the bad times or times where dice "go nuts" for lack of a better word. They do not remember the many other times where dice stay within what they think is normal. The placebo effect at work here.

Well in all fairness evidence has been presented. I posted a topic a while ago giving a statistical breakdown of just how skewed my dice results were. The response was, and always will be, "Well you just need to roll more dice!". Of course those that responded that way failed to see that the # of dice I had accounted for was about 50 games worth of dice, or about 4-5 months of playing.

Also, Andy came in (per usual) and said that, even though the math was over his head, I should search for other posters' topics about the dice and just agree with those that found nothing wrong with the dice.

[theBastard]

Does it really matter?

...just only for you...

I just refuse to believe that anyone that finished some sort of education has these elementary questions about probabilities.

what is understable that two times 6:3 I lost. and what is understable that 5:5, 5:4, 4:3 I win?

[JelleR]

Just to illustrate: the long run is a very long time. Here I picked a pokerplayer that has a certain EV (Expected Value) when he is playing poker. In this case, his EV is positive and he expects to earn a certain amount for every hand he puts in at the table. The dotted line shows these expected results as a relation to the number of has played. I've simulated 10 scenarios, and you can see that the outcomes are not exactly on the dotted line -- although he "should" be winning that amount based on his skilllevel. Waht you are seeing is variance, and you can also see that it take a very long time to "even out". You can run under expectation for a very large number of trial, pokerhands, or....dice rolls. I mean, I picked 150.000 hands here and there are clearly some running way under expectation and some very much over.

[Woodruff]

You do realize that dice don't always behave the way you want them to, right?

no, it has nothing with "way as I want". and it is nothenig about only me. it is about abnormal set up. where is any logic that one time I win with 5 against 5 and one time I lost 6 against 3? what can say guy who was beated with my 5 against his 5?

That is the nature of "random". It's actually perfectly logical.

Particularly given the very limited number of games that would quantify your experience.

again no, there were enough these things...

Incorrect. The sum total of your experience is far too small of a sample size for you to make those statements.

[Woodruff]

He cannot answer this... All the people who complain about the dice, never seem to come up with any hard evidence to show they are "broken". They only remember the bad times or times where dice "go nuts" for lack of a better word. They do not remember the many other times where dice stay within what they think is normal. The placebo effect at work here.

Well in all fairness evidence has been presented. I posted a topic a while ago giving a statistical breakdown of just how skewed my dice results were. The response was, and always will be, "Well you just need to roll more dice!". Of course those that responded that way failed to see that the # of dice I had accounted for was about 50 games worth of dice, or about 4-5 months of playing.

50 games really isn't a significant enough data set...not by a large margin. I would say that even if your dice results agreed with my position.

[nikola_milicki]

Not looking to stir up trouble. Mainly just trying to find reason and resolution. Wanting to have a rational meaningful discussion because I do feel that there is something inherently wrong with the dice. I do believe that the streakiness has increased significantly and reduced the game play, strategy and overall enjoyment. Before my membership expires, I would like to see something done or I will not be renewing like so many others. Yes, just because a lot of people perceive something doesn't make it absolute. But the actions of many of the older, more respected members of the cc community, and not just them, would have me to believe something differently. Now, unless they are just focusing on the bad as well, something is wrong. People who have been here for years and years have given up because they don't like the behavior of the new dice. I would think their experience would account for something. That they have something to draw from. So based on my experiences and the experience of others, I would say new dice bad.

I offered a reasonable explanation for this. Previous strategy used by some old-timers that worked very well with the dice that were previously used "hasn't changed with the times enough" for it to be successful with the way the new dice behave. That doesn't make the new dice bad, and it doesn't make the old strategy bad necessarily, but it does make the combination of the two...bad. What I'm saying is that the change in dice SHOULD have engendered a change in strategy, but too many folks who have had great success here for years want "their strategy" to work, regardless of the change in dice. They're unwilling (or possibly unable in some cases) to move with the change.

what is this mumbo jumbo??? sounds like something a politician would say..

1. please define "previous strategy used by old-timers"
2. Im especially interested to hear about changes that have to be made so that the 'old strategy' would work better with the new dice, make a short list of these changes please
3. explain how can these changes help me in games where I have to attack, there's no other option, like in 1on1s, but I roll 0-12..

[Woodruff]

Not looking to stir up trouble. Mainly just trying to find reason and resolution. Wanting to have a rational meaningful discussion because I do feel that there is something inherently wrong with the dice. I do believe that the streakiness has increased significantly and reduced the game play, strategy and overall enjoyment. Before my membership expires, I would like to see something done or I will not be renewing like so many others. Yes, just because a lot of people perceive something doesn't make it absolute. But the actions of many of the older, more respected members of the cc community, and not just them, would have me to believe something differently. Now, unless they are just focusing on the bad as well, something is wrong. People who have been here for years and years have given up because they don't like the behavior of the new dice. I would think their experience would account for something. That they have something to draw from. So based on my experiences and the experience of others, I would say new dice bad.

I offered a reasonable explanation for this. Previous strategy used by some old-timers that worked very well with the dice that were previously used "hasn't changed with the times enough" for it to be successful with the way the new dice behave. That doesn't make the new dice bad, and it doesn't make the old strategy bad necessarily, but it does make the combination of the two...bad. What I'm saying is that the change in dice SHOULD have engendered a change in strategy, but too many folks who have had great success here for years want "their strategy" to work, regardless of the change in dice. They're unwilling (or possibly unable in some cases) to move with the change.

what is this mumbo jumbo??? sounds like something a politician would say..

1. please define "previous strategy used by old-timers"

Self-explanatory. This is whatever strategy the old-timer found particularly successful prior to the dice change.

2. Im especially interested to hear about changes that have to be made so that the 'old strategy' would work better with the new dice, make a short list of these changes please

You appear to be looking for specifics regarding a very general statement. I'm afraid that won't work.

3. explain how can these changes help me in games where I have to attack, there's no other option, like in 1on1s, but I roll 0-12..

You DON'T always have to attack in 1-vs-1 games.

[nikola_milicki]

well then can u shortly describe at least 2 or 3 strategies that I as an old-timer use? cuz I cant..

maybe u shud talk about something specific for a change.. something that u can explain or give examples for.. point is u wrote a post but cant explain parts of it.. u keep talking about strategy but cant explain what the heck is it..

oh u dont? really? in which situations would that be then?

[WestWind]

He cannot answer this... All the people who complain about the dice, never seem to come up with any hard evidence to show they are "broken". They only remember the bad times or times where dice "go nuts" for lack of a better word. They do not remember the many other times where dice stay within what they think is normal. The placebo effect at work here.

Well in all fairness evidence has been presented. I posted a topic a while ago giving a statistical breakdown of just how skewed my dice results were. The response was, and always will be, "Well you just need to roll more dice!". Of course those that responded that way failed to see that the # of dice I had accounted for was about 50 games worth of dice, or about 4-5 months of playing.

50 games really isn't a significant enough data set...not by a large margin. I would say that even if your dice results agreed with my position.

1. The type of analysis I did takes sample size into account.

2. We've been talking about statistical significance so far, which is fine. But, let's switch definitions to player significance. 50 games is about a fifth of the games I've played over the course of a year. That means for 1/5th the time I had premium, my dice were essentially broken. They were removed from the expected results by huge degrees. Could the overall population of numbers have been random? Sure. But mine were so far removed from expectations that it was no longer enjoyable to play. You can hold up the Holy Grail of lists of 50,000 random numbers all you want, but at my scale it was breaking down. As a result, I'm not paying for this site any more. The administrators on this site can determine if that's "significant" or not, though their responses have already indicated how they feel about their player base.

[theBastard]

ok guys, all this talking is about what?
I understand that it is about possibilities, but nobody answer on question: where is any logic when one time I win 6:3, 6:4, 3:1. 3:2, 5:5 and sometimes I lost witht the same numbers?
how can I think about strategy when all is build on lucky only? then it is better to play cubes...

[SuicidalSnowman]

ok guys, all this talking is about what?
I understand that it is about possibilities, but nobody answer on question: where is any logic when one time I win 6:3, 6:4, 3:1. 3:2, 5:5 and sometimes I lost witht the same numbers?
how can I think about strategy when all is build on lucky only? then it is better to play cubes...

I hear what you are saying. The dice are sometimes so difficult to predict that they seem more important than any good strategy. At times, you could play the perfect game and the dice would just prevent you from ever having a chance.

I would respond that 1) this is the very nature of randomness and 2) that it happens to everyone at some time, you need to take the good with the bad.

However, I have asked this of many other posters, and now I specifically ask you, theBastard: What would your suggestion be to improve the dice, specifically on ConquerClub? Are you advocating that we abandon dice? Are you saying that dice are fine, we just need a new system for generating them?
As I have repeatedly stated, I do think the dice are fine, but if someone can show me not a way that they are wrong, but instead a way that they can be better, I would devote my time to getting that implemented.

1) MNDuke and theBastard never finished any sort of education. Well, maybe highschool?

I think theBastard may write poor English as it may be his/her second language. On profile, it says Slovakia and in the signature "sorry my english."

[theBastard]

I hear what you are saying. The dice are sometimes so difficult to predict that they seem more important than any good strategy. At times, you could play the perfect game and the dice would just prevent you from ever having a chance.

I would respond that 1) this is the very nature of randomness and 2) that it happens to everyone at some time, you need to take the good with the bad.[/quote]

I understand. and also I understand that dice brings new dimension to game. and I meant nothing bad, just when I wrote my frist suggestion here I was very angry it is still the game

However, I have asked this of many other posters, and now I specifically ask you, theBastard: What would your suggestion be to improve the dice, specifically on ConquerClub? Are you advocating that we abandon dice? Are you saying that dice are fine, we just need a new system for generating them?
As I have repeatedly stated, I do think the dice are fine, but if someone can show me not a way that they are wrong, but instead a way that they can be better, I would devote my time to getting that implemented.

only what I can say: dice is fine (as terrain or weather in real battles for example), just when I assault with 6 against 3 there could be 80% chance for victory. in my turn hour back I assaulted with 3 against 1 and I lost. when I assaulted 2 against 1 I won... so yes, I say it needs new system of generating dice (6:3 is big predominance).

1) MNDuke and theBastard never finished any sort of education. Well, maybe highschool?

I think theBastard may write poor English as it may be his/her second language. On profile, it says Slovakia and in the signature "sorry my english."

thanks Suicidal. I have sometimes problems with understanding or to explain fine my thoughts in english

[natty dread]

only what I can say: dice is fine (as terrain or weather in real battles for example), just when I assault with 6 against 3 there could be 80% chance for victory. in my turn hour back I assaulted with 3 against 1 and I lost. when I assaulted 2 against 1 I won... so yes, I say it needs new system of generating dice (6:3 is big predominance).

Ok, so you want the dice to act more "realistic", ie. representing real life combat more accurately? Is that what you're saying?

But, CC is not about simulating combat situations accurately. If you want that, you should play some other type of strategy game. CC is a simulation of a board game. In board games, sometimes you get funky dice results that make no sense in real-life context (ie. 5 armies defeating 15 or something like that).

Although: it is not uncommon even in real life for sometimes smaller armies to defeat many times larger ones. I should know, being from Finland... (google WWII and Winter War).

In CC you should always keep in mind that the dice are an unpredictable element. You should not expect anything from the dice. If you have a 20v5 battle, there is a chance that those 5 will defeat your 20. A small chance, statistically, but it's still there, and when it happens, it does not mean that the dice are flawed, just that you were sort of unlucky. You shouldn't take it personally.

2. We've been talking about statistical significance so far, which is fine. But, let's switch definitions to player significance. 50 games is about a fifth of the games I've played over the course of a year. That means for 1/5th the time I had premium, my dice were essentially broken.

Look at it any way you want, 50 games is not statistically significant. I'd be more worried if a sample that small was exactly spot on to the expected results. Random means there is always some variation, which evens out in the long run.

3. explain how can these changes help me in games where I have to attack, there's no other option, like in 1on1s, but I roll 0-12..

1v1 games have always been more or less a crap shoot. I suggest playing game types where dice are in a less important role.

[theBastard]

Ok, so you want the dice to act more "realistic", ie. representing real life combat more accurately? Is that what you're saying?

not exactly, but yes.

But, CC is not about simulating combat situations accurately. If you want that, you should play some other type of strategy game. CC is a simulation of a board game. In board games, sometimes you get funky dice results that make no sense in real-life context (ie. 5 armies defeating 15 or something like that).

board games? then why we make maps so real as is possible? but ok, I could fine also with this, but it is realy stupid when 5 can beat 15. sorry me but it is realy is.

Although: it is not uncommon even in real life for sometimes smaller armies to defeat many times larger ones. I should know, being from Finland... (google WWII and Winter War).

I know about Winter War and brave Soumi but this is another storry (Soviet army had not officers and at the end Finland lost territory). I just said 6:3 "must" win, only result could be different. ok, sometimes could be draw...

In CC you should always keep in mind that the dice are an unpredictable element. You should not expect anything from the dice. If you have a 20v5 battle, there is a chance that those 5 will defeat your 20. A small chance, statistically, but it's still there, and when it happens, it does not mean that the dice are flawed, just that you were sort of unlucky. You shouldn't take it personally.

disagree. this chance is absurd. if there is this chance it is not conquer club but as any game with cubes. I think CC is about strategy more as about lucky.

I do not take it personally, if yes I will be not here .

[natty dread]

board games? then why we make maps so real as is possible? but ok, I could fine also with this, but it is realy stupid when 5 can beat 15. sorry me but it is realy is.

We don't, actually. We make maps to be as fun to play as possible, and as fair (as in drops, bonuses etc.) as possible... realism is, most of the time, a secondary priority.

I know about Winter War and brave Soumi but this is another storry (Soviet army had not officers and at the end Finland lost territory). I just said 6:3 "must" win, only result could be different. ok, sometimes could be draw...

Finland lost some territory, but you have to remember USSR was out to take the whole country. They had at least 3x the troops we had. Yet, look at the amount of kills on both sides:

Finland:
25,904 dead or missing [F 7][12]
43,557 wounded[13]
1,000 captured[F 8][14]
957 civilians in air raids[12]
20–30 tanks
62 aircraft[15]

Soviet union:
126,875 dead or missing [8][F 9]
188,671 wounded, injured or burned[8]
5,572 captured[17]
3,543 tanks[F 10][9][11][17]
261–515 aircraft[F 11][9][17]

25k dead on one side, 125k dead on the other - that's about 5 times more dead on one side. And that's the side that had 3x more troops.

Yes, things can be said about poor preparations and strategy on the side of the soviet union. But still, it goes to show that things happen even in real life that you wouldn't expect, which can drastically alter the results of a battle.

So a 5 winning a 20 doesn't seem so ridiculous now, does it?

if there is this chance it is not conquer club but as any game with cubes. I think CC is about strategy more as about lucky.

Like it or not, CC has an element of luck. If you want a game that has no luck involved at all, play chess. CC has an unpredictable element of luck, which IMO is a part of what makes it so fun.

If you want a realistic strategy simulation, then a site that simulates a board game is not for you. I mean come on: the armies are represented with numbers, with no other parameters. On some maps, you can use 1 army to hold a ridiculous size of land, and that same army is also required to hold one small city. There's only 1 type of troops. We can have maps that feature real-life wars, but they are still always based on the board game format, and need to work within that framework. Then we have maps like chinese checkers, crossword, or conquer 4 which really have nothing to do with real-life war simulations, at all.

[theBastard]

We don't, actually. We make maps to be as fun to play as possible, and as fair (as in drops, bonuses etc.) as possible... realism is, most of the time, a secondary priority.

come on natty. look how much notices have real maps for more accuracy

Finland lost some territory, but you have to remember USSR was out to take the whole country. They had at least 3x the troops we had. Yet, look at the amount of kills on both sides:

Finland:
25,904 dead or missing [F 7][12]
43,557 wounded[13]
1,000 captured[F 8][14]
957 civilians in air raids[12]
20–30 tanks
62 aircraft[15]

Soviet union:
126,875 dead or missing [8][F 9]
188,671 wounded, injured or burned[8]
5,572 captured[17]
3,543 tanks[F 10][9][11][17]
261–515 aircraft[F 11][9][17]

25k dead on one side, 125k dead on the other - that's about 5 times more dead on one side. And that's the side that had 3x more troops.

Yes, things can be said about poor preparations and strategy on the side of the soviet union. But still, it goes to show that things happen even in real life that you wouldn't expect, which can drastically alter the results of a battle.

I expect these things. as I mentioned dice is fine as weather, morale, command skill... but I still think 6:3 "must" 6 wins. more if I´m able to win 5:5...

So a 5 winning a 20 doesn't seem so ridiculous now, does it?

no, necer. realy not. and it is baseless if CC is board games or realism simulating. in percentages 6:3 - 80% victory, 15% draw, 5% loss. but not in the same round in the same game two times 6:3 end with loss...

Like it or not, CC has an element of luck. If you want a game that has no luck involved at all, play chess. CC has an unpredictable element of luck, which IMO is a part of what makes it so fun.

I take CC with luck and so on. just CC could be build more on strategy, brain as luck...

If you want a realistic strategy simulation, then a site that simulates a board game is not for you. I mean come on: the armies are represented with numbers, with no other parameters. On some maps, you can use 1 army to hold a ridiculous size of land, and that same army is also required to hold one small city. There's only 1 type of troops. We can have maps that feature real-life wars, but they are still always based on the board game format, and need to work within that framework. Then we have maps like chinese checkers, crossword, or conquer 4 which really have nothing to do with real-life war simulations, at all.

if there is only 1 type of troops - is unlogic when 5 beat 20. when we should have 2 types of troops - base/elite, then I except 5 elite could beat 20 base. but no everytime 50% to 50%...

[SirSebstar]

5a vs 5b statisticly 5A should loose the majority of the time, in cc anyways