So I did go buy that box yesterday. I also changed a few things about the experiment to better suit this new way of seeing the dice. First, I moved the location of the box from my windowsill to a shelf next to a lamp (its one of those tall one's with three individual lamps on it). I replaced one of the bulbs with a 7.5 watt night-light bulb, which then lights up the box at night without keeping me awake.

Early this morning I once again attempted to do a few preliminary runs, which I’m glad I did because I ran into a few problems. I ended up having two lucid dreams instead of OBEs, but both involved doing this experiment. After setting up the box and everything, I set the saltcube timer and went to bed. In the first lucid dream I had, I went to the box to see what number was on the die. It looked like a 2, but it was rather unstable and the more I tried to focus on it, the more unstable it became, and as a result, the more it looked like a 3. Any who, the actual number was a 4, which didn’t surprise me because it wasn’t until I had awoken that I noticed that in the dream, the box was out in the hall way instead of in my room and on the shelf. Clearly I didn’t attain a sufficient level of lucidity to do the experiment properly.

I wrote down the results, put in another die (making it two dice in the box), shook the box up, and went back to bed. Like the first, the second lucid dream I had wasn't as clear or "whole" as I wanted it to be. In the dream I looked in the box and saw a 6. I woke up and the number was a 6, but then I got into a fight with my fiancé about some nonsense when I realized that I had just experienced a false awakening, and was still dreaming. I then woke up for real, went over to the box to find a 3 and a 6. Again, though I got one number right, I wasn’t lucid enough to remember that I put two dice in instead of one.

Conclusion, I need more practice before moving on to the actual experiment.


thafrenchman, on one had I know what you’re saying about making the picture something I can't get right by chance... but on the other hand I know that I don't need such drastic measures to produce reliable results. One of the benefits of using dice is that, unlike what you propose, the odds can be calculated because there are actual numeric values being worked with. This basically means that I can quantify my results using dice, which is very important.

And even though I could get a few numbers right by luck, in the long run (over the span of numerous trials), it will become clear whether my results can be explained by chance alone, or if they suggest something more. 1/6 is the probability of getting a number right by chance, meaning that for every 6 trials (where only one die is being used per trial), odds are that I should get roughly 1 right by chance alone. Of course, I could get more or less than that right in only six trials, but if I do many more trials, my results should average out to be about 1 success out of every 6 trials...unless there is something else going on.

This is why I want to use as many dice as I can manage per trial. If I used two dice instead of one, Ill have twice the data to work with. Or, if I could use 6 dice at a time, I would have the same amount of data to work with in only 2 trials, as I would with 12 single-die trials. Hell, the box can hold 9 dice, it would be ideal if I could take advantage of that so that I don't have to have hundreds of OBEs to get credible results.

I’m aiming for 120 individual sets of data, which could either be 120 single-die trials, 60 2-dice trials, 40 3-dice trials, 30 4-dice trials, 20 6-dice trials, etc. But first I need to know how many dice I can manage at a time, which is why I’m doing the preliminary trials.