$18.00
GeekGold Bonus for All Supporters: 128.15

7,583 Supporters

$15 min for supporter badge & GeekGold bonus
47.8% of Goal | left

Support:

Recommend
4 
 Thumb up
 Hide
12 Posts

Wargames» Forums » General

Subject: Vassal Dieroller rss

Your Tags: Add tags
Popular Tags: luck [+] [View All]
Thumis Dalidalisa
United States
Falls Church
Virginia
flag msg tools
mbmbmbmbmb
Is it really random? Honestly, I am not one to cry over unlucky rolls, but sometimes the rolls in Vassal seem skewed for a long series of rolls.

Some modules seem worse than others, but I don't know enough about programming to know if that would make any sense.
4 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
James Lowry
United States
Sunnyvale
California
flag msg tools
mbmbmbmbmb
Yes, it's random. Or as random as any computer algorithm's going to be.

http://www.vassalengine.org/wiki/Using_VASSAL#Why_are_the_di...
http://vasl.info/vassal/rng.htm
6 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Scott Muldoon (silentdibs)
United States
Astoria
New York
flag msg tools
designer
mbmbmbmbmb
Kolumel wrote:
sometimes the rolls in Vassal seem skewed for a long series of rolls.

Randomness is not some kind of smooth homogenous output.
8 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Douglas Brunton
Canada
Kitchener
Ontario
flag msg tools
mbmbmbmbmb
Kolumel wrote:
Is it really random? Honestly, I am not one to cry over unlucky rolls, but sometimes the rolls in Vassal seem skewed for a long series of rolls.

Some modules seem worse than others, but I don't know enough about programming to know if that would make any sense.


There are quite a number of Vassal users who can recount times in which they have been plagued by an unusually long series of poor dice rolls and times in which they have benefited from either an unusually long series of good rolls or the other players unusually long series of poor rolls (Hi Steve!)

As the poster above noted there are challenges when trying to create truly random number generation using a computer. I imagine as well that players quite often are more likely to see an issue when the virtual dice work against them. I've never actually seen a comprehensive test of the Vassal dice roller to get an objective idea of how well the randomiser in Vassal actually works. I don't think it actually has anything to do with the individual modules as they all use the built in vassal routine as far as I know.

Doug

6 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Wendell
United States
Yellow Springs
Ohio
flag msg tools
All the little chicks with crimson lips, go...
badge
Hey, get your stinking cursor off my face! I got nukes, you know.
mbmbmbmbmb
When the rolls are going my way, it's clearly random.

When the rolls are going AGAINST me, it's not random and those rolls should be ignored.

whistle
11 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Bart de Groot
Belgium
Vosselaar
flag msg tools
mbmbmbmbmb
If you don't like random you can take a deck of cards with the distribution of values you feel more happy with, and draw cards instead of rolling dice. Be aware though that the deck of cards will also be randomised by VASSAL. devil
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Thumis Dalidalisa
United States
Falls Church
Virginia
flag msg tools
mbmbmbmbmb
It's kind of a feature of BGG that everyone thinks they are the only ones who know anything about math, so condescending replies were probably predictable.

Yes, I know that randomness does not guarantee homogenous results. However, let's say you had a physical six-sided dice that you rolled twenty times and 14 times it came up a one. You could piously say that it's unreasonable to expect a homogenous result, but personally, I'd look for a different die.

Similar things have happened to me on Vassal. More than once in a lifetime/decade/etc. That's why I wonder if the randomizer is always truly (or approximately) random.
1 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Robert Stuart
United States
Los Alamos
New Mexico
flag msg tools
mbmbmbmbmb
dougb1 wrote:
As the poster above noted there are challenges when trying to create truly random number generation using a computer.


But then, I imagine there are also issues with trying to create random numbers using dice. I would think that no die would be truly random, and certainly not after very long use. But there would be a way (theoretically) to test it: set up a casino where all the craps rolls are done by computer generation. If (a) people keep coming to the casino rather than switching to other, dice-based, casinos, and (b) the casino doesn't go bust, that would be a reliable long-term test of the randomness of the numbers.
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Douglas Brunton
Canada
Kitchener
Ontario
flag msg tools
mbmbmbmbmb
bob_santafe wrote:
dougb1 wrote:
As the poster above noted there are challenges when trying to create truly random number generation using a computer.


But then, I imagine there are also issues with trying to create random numbers using dice. I would think that no die would be truly random, and certainly not after very long use. But there would be a way (theoretically) to test it: set up a casino where all the craps rolls are done by computer generation. If (a) people keep coming to the casino rather than switching to other, dice-based, casinos, and (b) the casino doesn't go bust, that would be a reliable long-term test of the randomness of the numbers.


I would expect that would be the case for dice as well. I've heard of lots of cases where dice were poorly balanced - either through poor quality control or deliberately.

Doug
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Jeb
United States
San Jose
California
flag msg tools
mbmbmbmbmb
Kolumel wrote:
It's kind of a feature of BGG that everyone thinks they are the only ones who know anything about math, so condescending replies were probably predictable.

Yes, I know that randomness does not guarantee homogenous results. However, let's say you had a physical six-sided dice that you rolled twenty times and 14 times it came up a one. You could piously say that it's unreasonable to expect a homogenous result, but personally, I'd look for a different die.

Similar things have happened to me on Vassal. More than once in a lifetime/decade/etc. That's why I wonder if the randomizer is always truly (or approximately) random.


Two things ...

1) Can a number show up multiple times in a row and be random ... Yes, it happens all the time. No math snarkyness intended.

2) Can vassal be used by a vassal email competitor to cheat? ... Yes, an opponent who restarts the game whenever they don't get an outcome they like or uses the playback feature could cheat. If this is a concern, you should use the ACTS service.

To be blunt, I have a hard enough time staying on top of my PBEM that I don't worry about this much.

(Edit to clarify vassal by email can be gamed)
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Joel Uckelman
United Kingdom
Durham
flag msg tools
mbmbmbmbmb
VASSAL's die roller is a wrapper around Java's java.security.SecureRandom class.

Here's a test program which writes 40MB of output from the random number generator, equivalent to 327m coin flips:

import java.security.SecureRandom;
import java.util.Random;

public class Test {
public static void main(String[] args) {
final Random prng = new SecureRandom();
byte[] b = new byte[4096];
for (int i = 0; i < 10000; ++i) {
prng.nextBytes(b);
System.out.write(b, 0, 4096);
}
}
}


When I feed the output of this program to ent, a randomness tester, I get the following result:

Entropy = 7.999996 bits per byte.

Optimum compression would reduce the size
of this 40960000 byte file by 0 percent.

Chi square distribution for 40960000 samples is 225.29, and randomly
would exceed this value 91.00 percent of the times.

Arithmetic mean value of data bytes is 127.4903 (127.5 = random).
Monte Carlo value for Pi is 3.141935463 (error 0.01 percent).
Serial correlation coefficient is -0.000208 (totally uncorrelated = 0.0).


The upshot is that the output of the random number generator which I tested is indistinguishable from a true random sequence.

I've had similar results every time, regardless of whether I've tested output from the RNG I generated myself or output I was given by others. In more than a decade of asking people for evidence that there's something wrong with VASSAL's die roller, I've yet to be presented with any. If you have such evidence, I'd like to see it. If not, please stop suggesting there's something wrong with VASSAL's die roller, as it's getting tiresome.
16 
 Thumb up
1.27
 tip
 Hide
  • [+] Dice rolls
James Lowry
United States
Sunnyvale
California
flag msg tools
mbmbmbmbmb
18 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Front Page | Welcome | Contact | Privacy Policy | Terms of Service | Advertise | Support BGG | Feeds RSS
Geekdo, BoardGameGeek, the Geekdo logo, and the BoardGameGeek logo are trademarks of BoardGameGeek, LLC.