My impression is that GameScience were so classed _by GameScience_ - in particular, they tended to have flash left between two faces you had to cut off yourself, and if you didn't get that right, you've got two definitely-bogus faces.
I think "several hundred rolls" very definitely sounds less feasible than drilling some holes in an old drive caddy.
The GameScience dice I have owned all had a visible imperfection from where the sprue was on the mould, but nothing I would characterise as "noticeably requiring any cutting or filing". But, I don't think I have seen a GS die produced after, um, 1992-or-so (possibly not after 1988-or-so), so I can only generalise from them. I'd say that they seemed fair. But, I have not at ALL taken the time to do a proper investigation (and these days, most of my RPG needs are fulfilled by D6).
I guess what constitutes "least resistance, at least for a quick check" really depends. With an estimated 15-20 minutes knocking up a "up to 20 clicky-buttons and maybe 25 rolls per minute", getting a single die to 300 rolls looks like about 35 minutes. Making a more compicated mechanical system is most probably worth it if (1) you want to do it as an exercise in hw/sw integration and/or (2) you want to test more than a single die or test it into thousands of rolls.
I'd personally knock up a small app and test a single one, first. If writing the analysis code was interesting enough, I'd then look at a Raspberry Pi, a camera of some sort, and a way of shaking/rolling the die.
![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png){kind=small}
(no subject)
I think "several hundred rolls" very definitely sounds less feasible than drilling some holes in an old drive caddy.
(no subject)
I guess what constitutes "least resistance, at least for a quick check" really depends. With an estimated 15-20 minutes knocking up a "up to 20 clicky-buttons and maybe 25 rolls per minute", getting a single die to 300 rolls looks like about 35 minutes. Making a more compicated mechanical system is most probably worth it if (1) you want to do it as an exercise in hw/sw integration and/or (2) you want to test more than a single die or test it into thousands of rolls.
I'd personally knock up a small app and test a single one, first. If writing the analysis code was interesting enough, I'd then look at a Raspberry Pi, a camera of some sort, and a way of shaking/rolling the die.