I put this over on Warseer to get a wider group of people that want to do stats.

http://warseer.com/forums/showthread.php?p=2962908

We will see what the final result is.

I rolled 200 dice after the weekend when people said I should have been called a cheat but for using dice straight out of spacks sack and came up with a similarly odd curve- but my raw data (inc pics of the 5 bar gates) seem to have vanished.. they were in the scores on the doors thread- keyword "dentist" because I said I was waiting for a dentists appointment so I had the time.
good luck with your research

EDIT found it
http://www.ageofstrife.com/modules.php? ... 5557#15557 (only available to those currently logged in)
3 turned up more times than normall 5 turned up less than normal in mine

when i get a chance i will roll some of my dice to see what sort of results i get. i think i will use just 1 die of each colour, and that way i can track individual results.

~ Tim

Gah PC crashage, I'm sure I just posted this
"Use 2 of each colour tim, 1 can be an anomoly (bad bubble etc) 2 will give you A) a mean average B) A comparison.
I use 4 blue dice 50 times and still most people would pick whole in my method of selection size of volunteers and usage 9also not a double blind- my force influenced the power within the super dice"

Would rolling several dice together also unduly affect the outcome since sometimed the dice might hit one another, and other times not? It seems like a fair evaluation would require that each die be put through the same simulation (i.e. shaken, rolled, and bounced off of the endwall in a similar manner).

Culven wrote:Would rolling several dice together also unduly affect the outcome since sometimed the dice might hit one another,.

thats what these guys took into account
http://warseer.com/forums/showthread.php?t=65531
(the thread died slowly as all things do on warseer but is more easily retrieved from the dakka articles section)

Mine were rolled 5 at a time. It was always the same 5 but not necessarily the same place in my hand. I presume there would be some benefit in rolling them separately but that is a lot of time to devote.

More data is up.

Good think I don't play against this guy.
http://www.youtube.com/watch?v=fJbUwnnlIFQ

OK. I was intrigued so I did my own little test. Only 300 rolls. But my dice appear to be almost DEAD ON what they should be based on an actual average hit of 16.67%.

Code: Select all

Value:      1         2          3          4          5         6
Times:      50        46         54         57         48        45
Percentage: 16.67%    15.33%     18.00%     19.00%     16.00%    15.00%      


Actually, I am a bit disappointed that mine don't roll more 6s. I really could have used that edge.

Culven wrote:Would rolling several dice together also unduly affect the outcome since sometimed the dice might hit one another, and other times not? It seems like a fair evaluation would require that each die be put through the same simulation (i.e. shaken, rolled, and bounced off of the endwall in a similar manner).

I don't think it makes any difference unless you roll each die in the same place and hope that things like air currents and minor imperfections on the table surface don't affect the roll (which they could do). Dice knocking into each other is what you would do in a game...otherwise the conditions you are testing in are artificial. Who rolls each die individually in game, in the same place at the same pace?

Whether a die hits another or not is all part of the randomness of rolling dice and you cannot eliminate all of the factors.

Here's an idea... why doesn't everyone conduct the same experiment and then we can pool all the results. That way we get a much bigger result set which will be more accurate.

i would have thought that without being able to track results back to an individual die, one very unbalanced one could make a whole sample look a bit unbalanced.

im going to clear some space on the games table (to recreate normal conditions) and see if i can gather some data of my own...

~ Tim

I wonder if that is the case with mine. If I had say two dice in particular that were rolling 5 or 6 more often it would be lost in the 25.

How about also splitting data based on colour? After all, I think Ben only has the Scarab Red and Scarab Blue dice, and the Speckleds are definitely heavier

ive got my firs sets of results.

i rolled 3 different D6 and one of my Scatter dice all together, on my normal gaming table. sometimes they rolled until they stopped by themselves, sometimes they hit things and bounced before stopping (just like in a normal game). i rolled this handful of dice 200 times (giving 200 results for each die).

the D6 i used will be refered to as "LB1" (Light Blue 1), "DB1" (Dark Blue 1), "R(Gold)" (Red with Gold pips).

here are the totals:

LB1
1 = 45
2 = 32
3 = 26
4 = 39
5 = 24
6 = 34
(Total 200)

DB1
1 = 33
2 = 37
3 = 28
4 = 49
5 = 20
6 = 33
(Total 200)

R(Gold)
1 = 30
2 = 37
3 = 28
4 = 34
5 = 31
6 = 40
(Total 200)

Scatter
Hits = 71
Arrows = 129
(Total 200)

feel free to analyze them.

next i will be rolling 4 different D6 (3 are AoS, 1 is a custom dice given to me by a guy at club, from an event he attended). i will roll this batch 200 times as well. i will post again when i have these results.

cheers

~ Tim
p.s. when i am done i will post a labelled pic of all the dice i used for these experiments, so you can compare the colours to the ones you all have.