Rerolls - Recycled Dice

$9.95

No Compromises
In Quality
100% Satisfaction
Guarantee
Free Shipping
Over $40 in USA

The druid knelt on the ashy loam, taking a handful of soil and letting it slip through their fingers. While the area looked desolate and destroyed, their natural magic told them a multitude of seeds were scattered throughout the area. All they needed was some work, care, and a little bit of magic, and this area would be as lush and verdant as it had once been.

Rerolls are recycled dice with a purpose! A lot of plastic is left behind in the acrylic dice making process, and that has never sat well with us. It took time and research, but we’ve finally found a means to recycle that plastic into something new.

We aim to take these extra bits of plastic, and turn them into something that makes our world a better place. The proceeds from these dice will go toward a new charity every fiscal quarter.

This quarter, we have decided to support the Palestine Children's Relief Fund.

Additionally, Die Hard Dice will donate a full 7-piece set of Rerolls to MagiKids for every purchase made through our store, no matter what dollar amount you spend.

To Learn more about Rerolls, visit our Rerolls Page.

FEATURES

Our popular Elessia mold
Muted red with silver flecks and white inking
Traditional 7-piece RPG set
Made of recycled acrylic polymer
Each run of rerolls will be completely unique, based on the available range of colors of left over plastics
Comes with a Watermelon d20 sticker!

Shipping Policy:

Our About Shipping page has many useful details about international shipping and other in depth info.
Free Shipping within the USA with just $40 in your cart!
Orders are typically processed and shipped within 1-3 business days.
We aren't shipping on Fridays because we've been testing out a 4-day work week here at Castle Die Hard.
Shipping rates will vary based on the size and weight of the item(s) and the destination.

Return Policy:

Didn't like what you got? Just email us and find out why the bards sing songs about our customer service.
We'll make it better, whether you want a refund, a replacement, or a swap with some other artifact from our selection.

Excellent Dice

I had to toss two sets of these dice into my cart during my last purchase because not only is the colour of these dice absolutely stunning, I also truly appreciate the idea behind the Rerolls. They roll really well for me and they match my paladin's colour scheme so what else is there to say except that I really recommend these dice and will gift the second set to a friend?

Rerolls - Recycled Dice

Love!

I love these beautiful dice, and for such a great cause!

Great idea, great dice

I admit I was skeptical, I expected ugly dice that didn't roll well.

I was wrong. They roll great so far, and I love the depth of color and variation in the dice. And I love the values behind DHD that they would seek to repurpose what would otherwise be garbage.

Great dice - and really, really random (in a good way!)

Short version: these dice roll great (love the wobbly effect non-'precision' dice have for d12s and d20s), roll fair (more on that below), and to borrow from the sultan in Indiana Jones and the Last Crusade - I even [really] like the color. And the font. And the little dice bag they came with. And the nice card wishing good karma for these (woo! recycled resin!).
----------
Longer, stats-nerdy version: I'm a statistician, and while we use programs for serious simulation work, I love the idea of 'random number generators' you can play with - much more satisfying (if slower) than a mouse click. When I get new dice, I roll them many times (more sides on the die = more rolls needed to have the same statistical power when assessing 'fairness' of the die - I roll d20s 1851 times, down to 197 times for d4s. [Caveat: I'm hand-rolling these in a dice tray, not using a dice tower or some other more scientifically-rigid setting. That's because I like to test 'em how I rolls 'em, and frankly it's more fun to toss the little buggers.]

After rolling and logging the results, I run some stat tests checking how likely the dice are to be random. To save anyone not -really- interested in the details some reading: these dice passed with flying colors. D20s are the hardest in the 7-die set to have as truly random since it has the most (and smallest) sides and edges, and this one blew my other dice out of the water.

Since I'm assuming anyone still reading this is either a statistician or really interested in this level of nerdosity, I'll start throwing in some stats terms to save space (I know, I know - bit late for that...)

A useful test here is the chi square goodness of fit test, comparing the actual counts (how many 1s? ... how many 20s?) against the expected results (for a fair d20, 1851 / 20 per side in my case).

A generally-used threshold for 'maybe not so random' in this case might by a p-value of 0.05 ("If you rolled a truly fair d20 1851 times, and did that a whole lot of times, there's a 5% chance of getting a chi-square test result as or more extreme compared to what you actually got") - it means you might be fooled 1 time in 20 (not because of using a d20 - it's just an 'eh, seems alright' benchmark we tend to use...might wanna be stricter if you're sending a rocket to space, say!). Anyway, with this d20 I had a p-value of 0.09; that is far and away the largest result I've gotten among the 8 d20s I've tested so far, two of which are 'precision' metal dice (neither passed that '0.05' threshold; it's largely a function of such a large number of rolls. I did bootstrap sampling to estimate how the chi-square p-values are distributed for each d20 for a given sample size, and they'd very likely have passed if I'd rolled 'only' 500 or 1000 times). Every single other dice in this set passed that test; the 'worst' was the d12 with a p-value of 0.079, which again is really not that surprising a result coming from a truly random die.

I also did some Bayesian modelling (multinomial model using the Dirichlet distribution) to estimate the rough probability of rolling a specific number (for a d20, perfect randomness means 5% per side, 1/20), and the d20 once again did great - marginally less likely to roll a 4 than a 12 or 13, but really nothing eyebrow-raising in terms of departure from randomness.

Anyways, thanks (or maybe "I'm sorry"?) to anyone who actually read this far!