41 rolls are far not enough to be considered viable sample showing that the rolls are skewed.
Good. Now repeat that ten times and calculate standard deviation...
I should check how to calculate standard deviations - I remember almost nothing of statistics.
Oh, right, its 12 or 13 years since I learned basic statistics...
| 7 people marked this as a favorite. |
My impression (read: limited understanding) of this conversation:
A: "Math math math?"
B: "Psh, math math math."
C: "CHALLENGE ACCEPTED"
Did the moderators of the site change the coding of the system to make it more challenging to roll higher on a d20 than it has been in the past?
Nope. No modifications have been made to dice rolling recently.
| 1 person marked this as a favorite. |
Alright, I'll try to put this as simply as possible and I'll cut out math, if that a problem:
Question:
Did the moderators of the site change the coding of the system to make it more challenging to roll higher on a d20 than it has been in the past?
That's a simple black and white question. I just wanted a straight answer and I don't know how to contact the site administrator directly, or I would have done that instead.
Your comment didn't add anything to the post, it was just negative and wasn't helpful.
It's my opinion that if you don't have anything nice to say about a topic or other people, please just don't post anything.
Your post wasn't nice, and it wasn't very polite.
Thank you.
Okay, that clearly went over your head. Let me try it another way.
Your original post brought up the idea that there was an issue with randomization, and presented a limited sample size. The first response was someone telling you your sample size was too limited. The immediate response from a third party was to offer a larger sample size. Subsequent interactions been "B" and "C" were "B" upping the stakes and "C" gleefully rising to the challenge.
My "limited understanding" sidebar was a self-deprecating throwaway to make it clear that I was generalizing the math itself--which was irrelevant to my observation--and not asking for clarification.
Instead, I was pointing out the overall structure of the exchange, because I found it amusing--and pleasantly so, as HangarFlying seemed to get quite into the idea of testing this with a massive sample size, which as a nerd, I can appreciate on a big-picture level without having any interest in the task itself.
There was nothing in my post that was negative, and frankly I'm not even sure what you might have misread so badly as to think I was being negative. The condescension and high-handed moralization you utilized to fight back against this perceived slight, however, adequately portray your willingness to ignore your own advice.
Chief Technical Officer
|
Rolls that I will put into my calculator.
** spoiler omitted **...
Most "random number generators" take in account outside variables which change during the day so you can't find patterns.
Variables such as User Accounts; time of postings; date of postings; and so on; all of those variables which change from day to day and minute to minute make a huge difference on the outcome of most software "Randomizers" so that people can't find patterns, because as external information changes, so too will the outcome of the randomizer.
So rolling 100 d20's in one post; or even 1,000 – that doesn't matter to the base calculator of the random outcomes.
You need to get a wide variety of subjects from a wide variety of times and dates to see the averages.
But none of it matters, though.
Our random number seed doesn't use techniques like that. Whether you roll 1000 dice in a single post, or roll 1 die in each of 1000 posts in one day, or 1 die in 1 post per day for 1000 days, you should see a similar amount of randomness.
Most "random number generators" take in account outside variables which change during the day so you can't find patterns.
Variables such as User Accounts; time of postings; date of postings; and so on; all of those variables which change from day to day and minute to minute make a huge difference on the outcome of most software "Randomizers" so that people can't find patterns, because as external information changes, so too will the outcome of the randomizer.
A properly designed random number generator will appear equally random regardless of the seed. Adding more 'noise' to the seed does not make the outcome more random.
A properly designed random number generator will appear equally random regardless of the seed. Adding more 'noise' to the seed does not make the outcome more random.Most "random number generators" take in account outside variables which change during the day so you can't find patterns.
Variables such as User Accounts; time of postings; date of postings; and so on; all of those variables which change from day to day and minute to minute make a huge difference on the outcome of most software "Randomizers" so that people can't find patterns, because as external information changes, so too will the outcome of the randomizer.
I'd actually be surprised of any websites using a method such as number of user accounts, or total post counts, etc as a seed for random number generation. There is simply no need to use such odd sources for a seed when there are plenty of more 'reliable' methods for setting an initial seed number.
I wouldn't doubt there is some site out there where the programmer didn't know enough. But even then, it shouldn't change the 'randomness' of the outcome unless the underlying random number generator is also poorly designed, or the programmer is mucking with the values after the randomizer spits them out.
However, a much more likely explanation for the perception that some sites do that is the human tendency to see patterns where there are none.
Good. Now repeat that ten times and calculate standard deviation...
Gah! I'm getting flashbacks of Statistics 301 in collage...(brrr)
Good. Now repeat that ten times and calculate standard deviation...Gah! I'm getting flashbacks of Statistics 301 in collage...(brrr)
You have at least that. I can't even recall the formula beyond the fact it used square root somewhere...
| 2 people marked this as a favorite. |
Whether you roll 1000 dice in a single post, or roll 1 die in each of 1000 posts in one day, or 1 die in 1 post per day for 1000 days, you should see a similar amount of randomness.
Now there's a challenge to the endurance of HangarFlying's nerdiness.
I have a PBP character who 40-50% of the time, last I checked, rolled a 4 or less.
However I fully accept that is just my luck, and in fact proves the die roller is functioning correctly. :)
For what it's worth, I've had two game sessions in a row where players couldn't roll above a 15 (and usually not above a 10), and the GM couldn't roll less than a 12. Regardless of which dice were being rolled.
And these were RL sessions. Statistical flukes happen.
For what it's worth, I've had two game sessions in a row where players couldn't roll above a 15 (and usually not above a 10), and the GM couldn't roll less than a 12. Regardless of which dice were being rolled.
Your GM sure has some wonky 8-siders.