Conquer Club

Soooooooo
Im sure this has been posted many times......but
How in the hell can everyone's average roll be 3.50 or 3.51.
If the dice are really random..... I would tend to believe these averages
would very much more than the are. I may have looked at a small
sample size (100 players dice stats) but this seems unrealistic.
Maybe you need to look at other random dice generators to check statistics.
Hog wash IMO

in before: quit crying about your dice!
BaaaaHumbug!

Think of a probability distribution (those lines on the following graph). Then, let's think about the concept called "variance."





For dice, the variance, I imagine, is very small (i.e. very tight, e.g. tighter than the pink curve), such that you'll get nearly all players with 3.5 or 3.51 on average (i.e. right on the mode/median part. In the graph, the average is 0, but pretend it's 3.5). Of course, there's some chance that one player would average 3.6, and this probability would be the same for a player with an average of 3.4. But, how many players would fit this description? It depends on the variance.

With high variance, you'd have higher chances of getting individual averages beyond the average of all players. The probability distribution gets fatter (e.g. blue). With low variance, you'd get very low chances of observing a player's average deviating far from the total average. The probability distribution gets thinner (e.g. pink).

I'm assuming that the dice generate some probability distribution of very low variance, but I can't explain why. Someone more knowledgeable of statistics (and who gives a shit to spend the time explaining) will have to step up.

fiends rolling dice...


keepin' CC real...-Jésus noir

willedtowin1 wrote:How in the hell can everyone's average roll be 3.50 or 3.51.

The more rolls someone has made, the closer their average should be to exactly 3.5. (The explanation is here.) The fact that many people had an average of 3.51 was one clue that there was something wrong with the dice generation last year.*


* From what I last heard, the dice are still not truly random, but at least each number now has an equal probability of showing up, and no one has an unfair advantage.

willedtowin1 wrote:Soooooooo
Im sure this has been posted many times......but
How in the hell can everyone's average roll be 3.50 or 3.51.
If the dice are really random..... I would tend to believe these averages
would very much more than the are. I may have looked at a small
sample size (100 players dice stats) but this seems unrealistic.
Maybe you need to look at other random dice generators to check statistics.
Hog wash IMO

in before: quit crying about your dice!
BaaaaHumbug!

3.5 is the expected average for rolling a D6.

(1+2+3+4+5+6)/6=3.5

The more times you roll, the closer you get to the average, and most people on CC have rolled very many times indeed.

BigBallinStalin wrote:Think of a probability distribution (those lines on the following graph). Then, let's think about the concept called "variance."





For dice, the variance, I imagine, is very small (i.e. very tight, e.g. tighter than the pink curve), such that you'll get nearly all players with 3.5 or 3.51 on average (i.e. right on the mode/median part. In the graph, the average is 0, but pretend it's 3.5). Of course, there's some chance that one player would average 3.6, and this probability would be the same for a player with an average of 3.4. But, how many players would fit this description? It depends on the variance.

With high variance, you'd have higher chances of getting individual averages beyond the average of all players. The probability distribution gets fatter (e.g. blue). With low variance, you'd get very low chances of observing a player's average deviating far from the total average. The probability distribution gets thinner (e.g. pink).

I'm assuming that the dice generate some probability distribution of very low variance, but I can't explain why. Someone more knowledgeable of statistics (and who gives a shit to spend the time explaining) will have to step up.

I would suspect the larger the sample size the smaller the variance. Also, the number of possible outcomes would clearly affect the variance.