I plan to order a set of chessex dices, am I being memed ? I heard they are rigged or something

I plan to order a set of chessex dices, am I being memed ? I heard they are rigged or something.

Rigged implies there's a plan. They're poorly balanced, which means plenty of them roll crappy, or just mediocre. I have seen Chessex dice manage this by screwing up the order dice sides go in. Some of them are literally recycled old dice.

This guy is an idiot.

99.44% of dice are injection moulded plastic that have their sprue blemishes polished off in a rock tumbler to make then look shiny and smooth. Sperglords flip the fuck out at this though since it means that no die made this way can be absolutely perfect. People with the right number of chromosomes already know that nothing's perfect, and that Chessex's quality control seems to keep their dice consistently at the high end of Good Enough.

That said, the worst offenders for inconsistent rolls are the opaques that look like they're made of compressed scraps - like the ones in your pic - they're more likely to have voids and bubbles at the center than ones made of a single resin. tl;dr:

Stick with high-contrast transparents - from any manufacturer - and you'll be fine.

They are dice, no need to over complicate your life with retarded questions.

Top red bars are Chessex d20

Bottom green bars are Gamescience with a sprue nub still opposite the 14 which is fucking with it's ability to roll 14's specifically.

> the high end of Good Enough

You shitting me?

Quite literally, Chessex dice are nearly as bad as though they left a sprue injection nub on every face, and dice only even have one injection nub anyway. All Chessex does is smear the problem around the dice faces by giving it an uneven polish. This is the paradox of manufacturing dice: You can't mass manufacture them without an imperfection somewhere.

If you buy "precision" dice and just shave the nub off yourself with a hobby knife, you can solve the manufacturing paradox without fucking up the rest of the entire die. This is what people used to do before Chessex came along and made everyone lazy. Precision dice, shave off nub, color in numbers with crayon. Gamescience will last longer too, they're made of a pretty decent high impact polymer, a little less dense than chessex's Chinese plastic, but the edges won't wear as quickly.

>rigged

Look up videos of people floating them in salt water. This is a way of testing golf balls actually. Then look at videos of people breaking open their opaque dice. The plastic pellets they melt down are still evident, in addition to bubbles.

Translucent are OK typically if you can inspect for bubbles before buying.

Or just don't buy cheap dice.

...

Here
Just use these instead

Unless you're rolling the dice 10,000 times in a single session, it doesn't really matter unless it's clearly skewed towards certain results.

This. The difference between a gamescience dice and a chessex scrap dice would probably never be evident through a 10 year D&D campaign except in extreme cases.

Most dice are perfectly fine unless you're actually rolling them to create a rdistro.

>doesn't really matter that the dice aren't fair
OK

>it doesn't really matter unless--
--you plan to use it for 20+ years.

Chessex d20s start to roll like a marble in about 5 or less depending on how often you play. Some of them (The swirls in particular I've noticed) already do that from the factory on about half of them. My old speckled sets of Chessex from my early days of gaming I've already tossed and replaced.

You buy Chessex because they look nice and they're available at every LGS. You don't buy them for life.

>milled metal dice
>they cut numale soft baby hands if they roll the girls abusively
>callused hands only
>my groups are always afraid of them after they realize how sharp they are
>"Oh, where did my d4 go, did I drop it?"
>mfw

Metal dice are higher maintenance bitches, but I love the girls; they're fun as a fuck to roll and roll fairly.

3d printed lost-wax cast dice are
GOD TIER

Sterling Silver is wicked cheap too if you're casting them yourself.

that looks both difficult to read and poorly balanced. while silver may be cheap, an oxyacetylene rig, investment, a degasser, a kiln, and a centrifuge are not. You could just show off your pretty dice without making stupid claims

>oxyacetylene rig
>degasser
>etc

That's professional casting and injection. You can make even steel in your backyard with primitive techniques, and get away with sand molds. Most of it is just understanding how to design a good mold.

If it bothers you, just use pewter.

>unbalanced

You can literally balance it with computer precision in CAD.

Wizdice are cheaper and with less visible defects, but are often uglier.

You literally rolled one die a billion times and behave as though that's statistically significant or in any way reliably representative of the manufacturer as a whole, and you've been riding that for months (a year?). I remember back when it was only 2,000 times, and you were convinced you'd cracked some code then, too.

What if that particular die was more well-balanced than most Chessex dice? What if it was worse? What about the Gamescience die? Your findings were made with a test group of one (at least it was back when you first started posting this, hopefully you've picked up another die to roll at some point in the last eight-thousand rolls) and yet you home in on every die-related thread to post them. I appreciate you doing science, user, in the limited way that you have. But it's clear that you've just put so much damn time into it that you've become convinced it was important or eye-opening, and so you cram those extremely limited findings into every thread you can as if they mean anything.

Even if you were completely correct in extrapolating an entire line based on your statistically subatomic test group, it *still* wouldn't matter, for a number of reasons. The first and most obvious being that none of this is going to be noticeable at a table anyway, but the second--and far more damning--is that if anyone actually cared about approximating randomness to the degree that you appear to, they'd be using a fucking random number generator, not a different brand of flawed polyhedral.

TLDR: To everyone in the thread, none of any of this matters. Pick dice you think look pretty with the understanding that you're never going to approach true randomness regardless of the brand you pick up. To this user in particular, for the love of god stop. You've been posting that *for ever*.

>REEEE STOP BRINGING UP FACTS THAT PROVE ME WRONG
>Just use inferior dice it doesn't matter if you always wind up rolling 1s at crucial moments.

k.

There is a difference between facts and conclusions, user, and your conclusions don't follow from your facts. I could buy a car at random and drive it until something breaks, but would that finding give me a credible reason to say that all the cars from that maker break at that amount of mileage? No, because that's anecdotal, and adding another (single) car from another manufacturer and comparing the two gives me statistics that are dodgy at best and outright misleading at worst.

It was a fun experiment, user, and I'm glad you did it, but fucking treat it like one.

The ones I've tested have been pretty unbalanced. Notice especially how disinclined they are to roll 20s.

Did my statistics final on Chessex vs. Gamescience vs. various web dice rollers. Chessex and Gamescience's standard deviations were almost identical, using three of each d20(could only get three gamescience d20s from a friend, so that decided my test) over the course of 2000 rolls per die. No real outliers from any of them, gamescience or chessex. Those spindown dice that come in MTG packs were also basically the same. The only die that I found suspect was the Roll20 dice roller, which was strongly balanced to 1, 3, and 17.

Unfortunately I don't have the raw data anymore.

When I test two of my really old dice from the 80s, however, the results were even worse. I expected this from the beat up d20 I got from the Basic D&D Box, as you could see how the edges were crumbling away. I was surprised how bad the results were for the translucent purple d20 that was the first die I ever bought on its own, as it has no obvious signs of wear. Both dice are smaller than is normal now, and perhaps the process for making them has changed somewhat. I'm thinking of getting some gamescience dice, and if so, I'll probably run similar tests on them to see how they match up. Almost all of my dice (at least the ones that weren't left over from the 80s) are chessex dice.

Ignore 1st image. I'm tired and made a really stupid mistake.

They're fine. Some people (including spergs in this very thread) will say that they roll unevenly and thus are useless.
The reality is that all dice do, even game science dice, which will always have one side with a 'lump' on it that causes it to almost never land on that side.
If you're obsessed with having your dice have perfectly even odds of landing on any given face, your best option is casino dice, but they only come in D6s.
If need an absolutely 100% random method of rolling, use a random number generator.
If you're vaguely worried about bubbles and the sort making your dice roll unevenly, get translucent dice so that you can see if they have bubbles and either refund them or buy new ones.
Finally, if you're just a regular human being who isn't all that worried that the 14 on their die might show up 0.5% more times than the 13, then just pick whichever dice look pretty to you.

I'm personally okay with having my dice be a bit less than perfect, especially if nobody knows the way that they're less than perfect. But if the die gets only somewhere between 1/2 and 2/3 as many twenties as it should (as in ), I consider that problematic. There's "off" and then there's just plain "fucked".

>if you're just a regular human being who isn't all that worried that the 14 on their die might show up 0.5% more times than the 13
What if the 6 shows up 2.49 times more often than the 20?

In that case buy translucent dice. The incredibly low amount of 20s from can be resolved by making sure you have no bubbles, as the speckled dice are just stylised opaque dice, so their quality control can't throw out ones with bubbles.
What I do is use one of those huge packs of dice (specifically the wiz dice 100+ pack) and just alternate between dice. In theory by using a selection of different dice I will get something very close to even odds of landing on any given face.

>all chessex dice are shit because one study tested a grand total of 2 dice that where shit

>numbers on dice only have a .5% difference in the frequency with which they appear because 0 studies tested a grand total of 0 dice

Why the FUCK would you roll the same sample dice 2000 times? That's not how you do statistical research, what the fuck?!

>not using your computer to roll dice

Lol grandpa go to bed.

>FACTS THAT PROVE ME WRONG

Not him, but they don't. Your tiny sample size means your data is statistically insignificant. You have to be a special kind of stupid to not understand this.

The fact that the only counter-argument you could come up with was greentext memeing says it all. You're not just stupid, you're stupid but think you're smart, and that's just tragic.

>that hypothetical example isn't scientifically accurate so a study with a sample size of 2 is now relevant again

How many times would you roll it?

>Why the FUCK would you roll the same sample dice 2000 times?
I think they mean they'd roll more dice fewer times.
Like maybe roll 100 dice 1000 times each.
Maybe even roll several thousand dice 1 time each.

>using dice
Y'all niggas need to get on this radioactive decay phenomenon, for real.

>People with the right number of chromosomes already know that nothing's perfect, and that Chessex's quality control seems to keep their dice consistently at the high end of Good Enough.
I have multiple Chessex d20s that are visibly malformed and weighted. I still buy them, but there are very real issues with Chessex.

500 per sample should be more than enough. Hell, you can probably go as low as 400 without losing anything and probably even lower, but my Statistics Voodoo is rusty, so I'd leave it at 500 abd be done with it.