I rolled these hand-forged metal dice hundreds of times. How fair are they?

That’s a ball of 28Si from the Avagadro project

4 Likes

Now I’m thinking of writing a really terrible time travel fantasy book, where a corporate trainer is sent back in time to Medieval England and tries to introduce Six Sigma seminars to blacksmiths. (Then I’m picturing blacksmiths sitting around rough wooden tables like Rob’s and being told to take up specific team rolls to make the tallest tower they can out of pieces of straw and cow dung…)

3 Likes

Having a look at melting points and densities, you could use a pot of molten polonium on the stove instead.

Probably want to do this outside or at a friend’s house though.

5 Likes

Cool!

1 Like

Eh. This guy had 240 lbs of mercury to flush down the toilet.

5 Likes

I think the real issue is the poor rolling + the weight of the dice.

With regular dice, that casual rolling would be fine, because the dice would bounce off the table more, introducing chaos. You could probably drop a regular die directly down, with 6 up every time, and get pretty random results.

With this die, I’ll bet if you did that you’d get much less random results.

So honestly, I think the question of whether the die are fair or not is actually a canard. Even if they were weighted, they bounce so little that it mostly depends on what face it fell on, not what side weighs more.

Since you pick it up and roll the same way each time, you’re probably getting into some kinds of cycles, such that patterns show up. This in turns leads those numbers to be over-represented in your final table.

So then it just comes down to your rolling. The final result will be as random as your roll is.

My prediction: if you did this another 300 times, with the same technique, you’d still get biased-looking numbers, but they will be different numbers.

If you roll it around properly in a cup, you’ll get random numbers.

2 Likes

This topic was automatically closed after 5 days. New replies are no longer allowed.