| |
BGonline.org Forums
SHG Dice Revisited - Fixed?
Posted By: NJ In Response To: SHG Dice Revisited - Fixed? (Michael Petch)
Date: Tuesday, 6 October 2009, at 5:21 p.m.
I'm going to take a wild guess at what was wrong with the dice. Here's a reasonable way to generate the dice:
- Generate a random number between 1 and 36.
- For each number, have a preassigned 2-dice combo: 1-1, 1-2, ..., 6-5, 6-6
Here's a similar but wrong way to generate the dice:
- There are 30 non doubles rolls and 6 doubles rolls.
- I know that non doubles are twice as likely to be rolled as doubles.
- Given #2, let's weigh the nondoubles by 2 and doubles by 1.
- With the weights, we have 60 nondoubles chances (30 * 2) and 6 doubles chances.
- Generate a random number from 1 to 66.
- If the number is between 1 and 60, generate one of 30 non doubles rolls.
- If the number is between 61 and 66, generate one of 6 doubles rolls.
Given that erroneous algorithm, the chance of generating a double would be 6 out of 66, or 9.09%. That would be close to the observed frequency of 9.1-9.2%. Of course, this is just speculation but it's the only reasonable explanation I can come up with for dice that are fair for nondoubles but with a 9.1% doubles rate.
| |
BGonline.org Forums is maintained by Stick with WebBBS 5.12.