Seriously your making up words now quit it!
Dice Roller 3000
by
I use the dice app and it's usually pretty fair but yesh tonight the reds were just not cooperating at all I'd roll four dice with a re-roll and only get one hit.
I didn't loose but it looked dicey for a while fortunately the decimator soaked up the damage long enough for my luck to turn, it died but vessery finished off the rebels.
Sad part is I can't perform an execution like I would with physical dice.
Honestly when I started this, I was almost expecting the dice to favor blanks by a very wide margin. The results so far show otherwise.
You guys and your big words ha. I think I'll just ensure the data collection is accurate then someone trained in stats can make a final determination.
Long day at work, so my tired little machine racked up almost 12,000 rolls on green06. Just finished image processing, but now my script is sorting the results (darn free software limitations!).
The general process is this:
Step 1
------------------
Roll Die
Take Picture
Repeat
Step 2
------------------
Read Picture
If Find Evade, then evade
If Find Focus, then focus
If find both, then error
if find none, then blank (blank by process of elimination, so error prone)
If shape is too rectangle, then error (when the die doesn't land flat)
Repeat
-export a text file with [focus/evade/blank/error]_picfilename
Step 3
---------------
Use a script to read the text file and sort each file into evade/focus/blank/error folders
Visually check the entire blank folder. This is actually very easy as they stick out like sore thumbs.
Toss out the false blanks
sum up how many pics are in each folder
results
geez I thought by the time I finished writing this post my sorting script would have finished. Time to kill omnigul. I'll be loading green04 tonight as I haven't calibrated the recognition for red dice yet.
Green06
Observed / Expected / Difference
Evade: 4341 4446 -2%
Focus: 2960 2964 0%
Blank: 4555 4446 2%
Total: 11856
P-Value: 0.08
Fair? Maybe not. Is it meaningful? Depends on if you think 2% less evades and 2% more blanks is significant.
I think one thing worth doing might be acquiring a couple of specimen dice (like the trued ones used in casinos) and testing them on your rig. Since those are required to be as fair as possible would be interesting to see how many pass/fail
Very cool. Would love to see this done with the app as well. Then compare.
Hmmmmm interesting. I think it can be done fairly easy, would just need to research how to fake the touch screen. Could screen cast it direct to the pc and skip the webcam.
All you need to fake the touch screen would be your button pushing rig with a stylus attached. I haven't used the app, so I don't know how many "touches" it would take to make the rolls happen, though.
How did I miss this post? So cool! Love the YouTube video.
KTreu42 was kind enough to send me the obviously broken die and I fired it up in the Dice Roller 3000 and ran it for a couple of days. 15,000 rolls later, this is what I got:
Observed / Expected / Difference
-----------------------
Evade : 5556 / 5553 / 0.05%
Focus : 3715 / 3702 / 0.34%
Blank : 5538 / 5553 / -0.28%
Total: 14809
P value: 0.957 - fair
I actually split this into 3 5kish sessions. The p-values were .377 .916 and .31 respectively. That value is roughly the chance the variance happened by shear chance.
So what does this mean? It means even with a giant defect, the die will roll pretty darn fair..... in a dice tower at least. What it doesn't say is whether or not the rolling method (mostly on a table) has any impact.
Sorry about pulling this from the graveyard but any chance the video still out there? The original post link does not work anymore.
Sorry about pulling this from the graveyard but any chance the video still out there? The original post link does not work anymore.
+ 1
I have reuploaded the vid and updated the first post.
Thanks for updating the video.
And, on behalf of humanity, thanks for only using your powers for good.
I have reuploaded the vid and updated the first post.
How awesome!! You are a genius
Thanks!!
@ OP. Can I just ask if there was 2999 prototypes before you came up with this one?
Do the dice wear out after being rolled thousands of times?