Eredan's World

buso deL cueo · 08-11-2010 19:42:01

is the random factor of aLL cards reaLLy random or are there vaLues which occur more often when certain cards are pLayed? i'm asking because i've pLayed "theft" countLess times now and it very rareLy discarded 3 cards, most often it just discarded 1, so i was wondering if it was just me being unLucky or if the random is not reaLLy random
thank you in advance

Rathedan · 08-11-2010 19:48:36

No computer generated number is truly random, but it's pretty close to it.

Streaks happen. Same as flipping a coin. Flip it ten times and will you have 5 heads and five tails every time?

Alakan · 08-11-2010 19:51:12

I feel ***bip*** pain... my [card]Heads or Tails[/card] proc'd -4 at least 75% of the time

buso deL cueo · 08-11-2010 19:56:16

by countLess i mean more Like 200 times, rather than just 10 and at Least 70% of the times onLy 1 card was discarded

pekpekkypek · 09-11-2010 02:20:53

Well statistically to truly determine a probability you'd have to do it a number of times approaching infinity. It is true what Rathedan said: no random number generator is truly random although it's pretty close and streak do happen.

Dernière modification par pekpekkypek (09-11-2010 02:23:32)

steven_allen · 09-11-2010 02:27:49

There is also the possibility that God hates you. I'd go with that.

buso deL cueo · 09-11-2010 09:40:14

LoL i think so too

i was just trying to understand if the distribution was supposed to be 33%*3 or if it was something Like 60/30/10

Rathedan · 09-11-2010 14:17:37

steven_allen a écrit :

There is also the possibility that God hates you. I'd go with that.

Lol true. I keep forgetting to include that as an explination

AxeMurder · 10-11-2010 22:00:42

buso deL cueo · 11-11-2010 12:27:00

http://en.wikipedia.org/wiki/Discrete_p … stribution

apparentLy i need to invoke wiki powers to get the point of my question through.

Dernière modification par buso deL cueo (11-11-2010 12:27:43)

pekpekkypek · 11-11-2010 13:26:29

buso deL cueo a écrit :

http://en.wikipedia.org/wiki/Discrete_p … stribution
apparentLy i need to invoke wiki powers to get the point of my question through.

you can't do discrete distribution with only 200 numbers. you need an infinite number of them.

buso deL cueo · 11-11-2010 17:16:34

d'uh, reaLLy?

it seemed Like it was a sufficient sampLe, and it was just a question.

pekpekkypek · 12-11-2010 02:02:57

buso deL cueo a écrit :

d'uh, reaLLy?
it seemed Like it was a sufficient sampLe, and it was just a question.

no it's not. statistical distribution curves are based on the notion that an infinite number of samples is being analyzed. you will also always have outliers and deviations and it's up to the statistician to set how many standard deviations are acceptable as random.

and i'm also only answering to the best of my ability, we did some statistical analyses on my group's thesis back in college.

Dernière modification par pekpekkypek (12-11-2010 02:03:55)

steven_allen · 12-11-2010 05:05:15

Pwn'd?

Rathedan · 12-11-2010 05:15:38

N=200 is a maybe a medium sample size... Something easily observable like this you could go 500 for good results....

With only 3 different observations 200 should be enough with a normal distribution....but again, theres always outliers. Remember too though, every sample is a dependant group. 18 or 19 times out of 20, the distribution should be 33/33/33%, +/- 2 standard deviations in a sample of 200. The other 1 or 2 can be crazy....

My Metrics is a bit rusty or I'd go through the equation to actually do up proper sample sizes so that your size would be sufficient with a 95% confidence and make sure your sample would be within X of the actual mean. It's been a while since I used stats now

buso deL cueo · 12-11-2010 09:34:08

what you're missing is that i considered 200 to be a sufficient sampLe -for me to ask-, not for me to say without doubt that the randomizer isn't working...

pekpekkypek · 13-11-2010 00:36:23

Maybe, but for 200 samples in something easily observable as Rathedan said, you have to take into account random scatter across the samples. If you take the random scatter and consider it a representative sample it's going to appear like it favors one end of the random range than the other, but to truly see "randomness" you have to add more observations.

bleachman · 13-11-2010 06:01:33

This is a card game forum. Where I will not deny that some of this game does deal with mathamatical issues, I will, however, enlighten you on the fact that it doesn't really matter. Random numbers are random, if you feel the game hates you, it probably does. Otherwise just accept you may have a bit of bad luck and look forward to the other billion times where you had and will have good luck. Do not forget those when randomness bites you in the ass.

For further headache material please check www.mathgeeks.com

Eredan's World

Annonce

#1 08-11-2010 19:42:01

randomness of some cards

#2 08-11-2010 19:48:36

Re : randomness of some cards

#3 08-11-2010 19:51:12

Re : randomness of some cards

#4 08-11-2010 19:56:16

Re : randomness of some cards

#5 09-11-2010 02:20:53

Re : randomness of some cards

#6 09-11-2010 02:27:49

Re : randomness of some cards

#7 09-11-2010 09:40:14

Re : randomness of some cards

#8 09-11-2010 14:17:37

Re : randomness of some cards

#9 10-11-2010 22:00:42

Re : randomness of some cards

#10 11-11-2010 12:27:00

Re : randomness of some cards

#11 11-11-2010 13:26:29

Re : randomness of some cards

#12 11-11-2010 17:16:34

Re : randomness of some cards

#13 12-11-2010 02:02:57

Re : randomness of some cards

#14 12-11-2010 05:05:15

Re : randomness of some cards

#15 12-11-2010 05:15:38

Re : randomness of some cards

#16 12-11-2010 09:34:08

Re : randomness of some cards

#17 13-11-2010 00:36:23

Re : randomness of some cards

#18 13-11-2010 06:01:33

Re : randomness of some cards

Pied de page des forums