New d20

2000 times is too small a sample size. Try 50,000 rolls next time.

What where the modal values and standard deviation?

>Try 50,000 rolls next time.
That sounds like a helluva game

Also, I'm a civil engineer, 2000 sample size is pretty much what we use in my company

Will look at my stuff once I come home, what I posted here was from memory, I think I have my stuff under some folder in my PC

Use for what?
You need different sample sizes for different predicted distributions. You'd know this if you weren't a retard. Consider quitting your job to retake the first grade, I'm sure you'll pass within two years.

Return periods in rivers floodings for example. Is not a matter of being exact, is a matter of being economically efficient.

>Cuck doesn't roll out his 1s.

>Rolls out his 1s
>Also rolls out his 18s, 19s and 20s

>Rolls out before transforming

Not for this, fuck off retard. Don't pretend to know anything about statistics when you don't.

Obviously, my "waste your life doing nothing productive" joke seems to have gone over people's heads. Perhaps it was my fault for not being clear.

It was not written like a joke

More importantly, you're not testing dice in a vacuum. The life of a casino dice only is 2000 throws. 50,000 throws is more than sufficient to introduce - or exacerbate - bias in a 'fair' dice.

Rolled 4, 4, 20, 8, 11, 6, 6, 3, 1, 10, 19, 19, 13, 16, 17, 20, 3, 20, 18, 1 = 219 (20d20)

ok, lets see if we have to throw this board into the trash

You know what to do.

Fun fact:
The chance of getting every number on a d20 once over the course of 20 rolls is .5%.

Show the maths, or you are lying.

may le dice gods grant ye many NAT FUCKING TWUNNYs my fine fellow

TRASH THE FUCKING BOARD!!

You have a 100% chance of getting a number you haven't rolled on the first roll.
Then a 95% chance (19 numbers left). Then a 90%, 85%, etc.
Multiply through, and you've got .005, or 5%.

I fucking suck at probability and statistics but let me give this a shot.
Chance first roll is unique: 100%
Chance second roll is unique:95%
...
Chance 20th roll is unique: 5%
so .95*.90*.85...*.05
Which is approximately 0.0000000232019615953125
or .0000023%

So question OP, have you ever kept a d20 before?

So back in school I did a statistics final on the odds for various D20s. I tested six Chessex D20s, three of those "Spindown" D20s that Wizards of The Coast gives out as life counters, one GameScience one, Wizards of the Coast's online dice roller, and Roll20's dice roller. I did 2000 rolls each.

Interestingly, all ten physical dice had essentially the same standard deviation. Despite common claims, the spindown d20s did not show trends towards a particular group of numbers, but none of the D20s showed strong trends. The GameScience Die was no better than the Chessex ones. The Wizards of The Coast dice roller had a slightly higher standard deviation, but the highest by far was Roll20's roller, which was extremely biased towards 1, 3, and 17.

Thought that would be interesting to people in here.

That's not how it works either.

20^20=104857600000000000000000000
Meaning basically there are that many possible results to rolling 20d20, including different orderings of the same numbers. Now, the amount of sequences that result in a satisfiable outcome are 20! (Or 20*19*18*,,,*1).

20!/104857600000000000000000000=0.0000000232019615953125

Same number as I got before. Pretty sure I'm right bud.

What were the test statistics for your chi-squared tests?

WRONG!

Only by dousing your dice in the finest of alcohols, a whiskey no younger than 30 years of age, burning an effigy of your enemies and making truly devout prayers to Lord Kraken, He Who Divides the Oceans and Armies can you even think about rolling said die. If it fails to roll it's highest number then forsake it and repeat the ritual anew!

It is

My statistics course was introductory and didn't cover those.

Wrong.

Right.

Any roll 1
Not the previous 0.95
etc 0.9
0.85
0.8
0.75
0.7
0.65
0.6
0.55
0.5
0.45
0.4
0.35
0.3
0.25
0.2
0.15
0.1
0.05

Product of them 2.3202E-08

That's not how it works because the fact is that you have a 5% chance to roll a twenty every time you roll the die. Reality doesn't care what your math claims, as is proven by the fact that people can roll several natural 20's in a row.or nat 1's in a row.

Yep, and a few lucky or unlucky streaks do not disprove overall statistics. If you take a suitably large sample size, in the order of millions or even tens of millions of dice rolls, you will find that these lucky streaks of 2 20s or 2 1s in a row only happen about 5%*5% of the time.

You seem to be missing the point.

This isn't about how likely you are to roll a 20, it's about how likely (not guaranteed, HOW LIKELY) you are to roll each number on the d20 within 20 rolls.

The results of the math people have done aren't guarantees, if you roll X number of times, you aren't guaranteed Y result, but you can reasonably expect that Y will occur once within X rolls.

>I'm gonna post this because pointlessly posting something that makes people angry makes me SO clever, right?
Says the attention-whoring namefag.

You have the exact same chance of rolling 3 20s in a row as you do of rolling 20, 3, and 17 in a row. You just assign more value in it because it's an easily recognized pattern and there is positive social connotations with that particular number on that particular kind of die.

>not making everyone anonymous with Veeky Forums x

What are the odds that I can roll a d20 three times in a row!?

That depends on if you have some kind of palsy, have at least two fingers, and if you are rolling in such a way or location that the d20 might be irrecoverably lost or damaged during a roll.
Find your roll factor on each relevant table and get back to us.

I actually think this would make for an interesting game. It uses dice that have an internal weighting mechanism that actually reduces the probability of a particular face being rolled multiple times, not without its opposite face being rolled to "zero" the die.

The odds are 1 because I already rolled before I asked

>Perhaps it was my fault for not being clear.

Way to hit the fucking nail on the head.

Rolled 1, 6, 19, 2, 14, 8, 17, 8, 4, 9, 3, 8, 16, 6, 18, 20, 13, 20, 4, 19 = 215 (20d20)

Its a fluke, they should all occur once

>all these replies
>"obvious bait"

>Get eight three times

Into the trash!

>New d20
>Roll it infinite times
>If all of the numbers don't appear an infinite number of times, throw the universe into the trash

Be rigorous in your testing!

>roll die infinite number of times
>die disintegrates
>die returns after an infinite time because what is entropy
>roll infinite dice infinite times for infinity

How am I supposed to have time to actually play RPGs?

The fuck? I took stats in the community college and we did chi square....

How the fuck did this many people fall for a bait thread.

I don't have anything modern to compare them to, but the stats I got on a couple of my Chessex dice showed some pretty strong bias.

>I don't have anything modern to compare them to
By which I mean that I also tested a couple of really fucking old dice I got in the eighties -- so old they were a smaller size, and one was really crumbly -- but I've not gotten around to testing anything put out since the turn of the century, other than those Chessex dice.

OP here.
I am a master baiter

I bought a metal d20. It was poorly made so that the 1 side, which is opposite the 20 side, was rounded on two sides so it would easily roll off 20. Compounding things the 20 side is nice and crisp so it easily sticks on 1.

Maybe it was bad luck, maybe the shop owner is an asshole and intentionally sent me a fail dice.

This

....but you didn't show the math! We're not stupid and aren't going to fall for your scheming numbers tricks again.

So, unless i'm retarded, this is a really simple probability check, simply going from 100% to 5% in 5% increments, resulting in roughly

0.000032019615953125% odds of occurring.

Or i'm entirely wrong and bad at math.

>0.000032019615953125% odds of occurring.
I get 0.00000232019615953112%. So one more zero after the decimal place, then a 2 inserted before the 3, and then at the end 3112 instead of 3125.

That's a less than 1 in 43 million chance.

So how is it even possible to roll three nat twenties in a row? Which happens a lot more often than 3 in 43 million times.

Most basic concept in statistics,
Something extremely unlikely is never impossible.

>I get 0.00000232019615953112%.
Scratch that. I get 0.0000023201961595312500% (so still one more zero after the decimal and then a 2 inserted before the 3, but it ends the same way yours did). Excel was doing some weird shit where it was interpreting 40% as 39.9999999999999% for no reason.

>So how is it even possible to roll three nat twenties in a row? Which happens a lot more often than 3 in 43 million times.
Maybe because it happens 1 in 8000 times?

Disregarding all of the discussion here, that's a really shitty-looking pair of dice. You'd need to squint and cock your head to the side just to see the numbers after rolling, and I thought they were blank dice before opening up the image. Who would make such a thing?

They haven't been inked.

To expound on this, when you're trying to get a different number for all of your first 20 rolls on a d20, your chance to succeed at each step goes down by 1/20 from the time before it. So the first time you roll the die, you have a 20/20 chance to succeed (because you haven't rolled before, any of the 20 numbers are fine). The next time you roll, it's only 19/20, because you have to avoid the number you rolled before. Then, assuming you're successful, you have an 18/20 chance on your next roll, because now you have 2 numbers to avoid. Then 17/20. Then 16/20 and so forth until you get to 1/20. So that's 20/20 * 19/20 * 18/20 * 17/20 ... * 1/20 = 0.00000232019615953125%

To get three twenties in a row, you have a 1/20 chance to succeed each time, so that's 1/20 * 1/20 * 1/20 = 1/8000. But your chances of following a 20 you already rolled with two more twenties is just 1/400.

...

I used to play Iron Warriors in 4th edition 40k, as you might know they were a bit broken and a 1500pt army consisted of a stacked Daemon Prince, 9 Obliterators, some artillery like a Basilisk or Vindicator, and a couple of minimum squads of Chaos Space Marines.

Since I didn't need that many dice, I held dice trials before any tournaments or trips to the LGS. I'd start with about 60 and leave the house with the 12 best rollers, and the most accurate scatter dice.

I probably just fucked up translating from notation, since I just used a calculator and multiplied .95*.9*.85 on down the line

pick this one for them sweet repeating numerals

pls

pls pls

...

Rolled 18, 5, 16, 4, 16, 4, 17, 3, 5, 16, 2, 14, 1, 14, 10, 8, 2, 8, 4, 1 = 168 (20d20)

Rolled 15, 18, 10, 6, 17, 1, 20, 18, 3, 17, 8, 18, 11, 6, 1, 6, 12, 16, 2, 20 = 225 (20d20)

Lame, let's try again

>4, 20
DUDE

you realise this isn't a problem to anyone who properly thins their paints, right?

>roll 1 d20 twenty times
>if it rolls more high numbers than low, it goes into the DM Dice Bag
>if it rolls more low numbers than high it goes into the Player Loan Dice Bag

And this is how my group of players learned the hard way to get their own damn dice sets

The problem here is people assume rolling a dice is only a question of probability when it's not
Rolling a dice can be train and master, it's a skill
It's not random

I rub my cum all over my dice so when my friends roll them they touch my cum

You're supposed to ink the numbers yourself, I have two sets like that which I've filled in. They were cheaper (I like the gamescience edges in my hands, not because they're "more fair").

Had a guy I played with that had a translucent set. I would roll and immediately pick it up so only he could see. He rolled a lot of 16+ on that d20, some really great luck with that die.

>can be train and master

not trained and mastered? are you east asian?