I don't know all that much about bitcoin, but I'm supposing that by "the caller" you mean the person who calls you with the sound that is associated with what the bitcoin value did at the end of the day. If that's what "the caller" means, then you're exactly right. In fact, I was a guest on the Gerry Ryan morning drive-time show in Dublin once and we did just that for a stock on the Irish Stock Market. Gerry challenged me to prove that ARV worked. So, his listeners were told to phone or write in (on Tuesday) what sound they thought they would hear him play as feedback on his show on Thursday morning. He would then pick a company's stock, known only to him, associate three separate sounds for rise in value, fall, or stay the same, and watch the stock's value on Wednesday to see what it would do by the end of that day. Then, on his Thursday morning show, he would play the associated sound, determined by what the stock's value had done. His staff logged in all the email and phone predictions that came in all day on Tuesday. Then, on Thursday, he played the sound that was associated with the rise, fall, or staying even (3 possible outcomes) that the stock's value had been at the end of Wednesday's trading day. Over 27,000 people phoned and emailed in their predictions. Chance for any of the 3 outcomes was 33%. The correct sound was described by (if I remember correctly) a little over 57% of the predictions. The experiment was a success.

Not necessarily. If you have the list of sounds, tastes, whatever type of feedback you have set the associations for, you can either make the sounds yourself at feedback time, or you can even download sounds from the internet and associate each with a possible outcome - then, after you know the outcome of the event, only play the sound associated with that outcome as your feedback. If you use another person to do this, then it's important that YOU don't know what the associations are, when you do the session, and that it is the other person who places the bet or takes action of the predicted outcome - then gives you the feedback later. But's if you don't want to involve another person, then it's important that you don't select the associations before you do the session. So, in that case, you would do the session, then, pick out, say, 2 different sounds, one for Team A, the other for Team B: , with one of them being the sound you described. BUT... then it becomes important for you (or someone else) to make sure that the assignment of the associations is random. If you don't do that, then you will almost automatically associate the sound you predicted to the team you want to win. The associations in this case >>must<< be random. THEN, you can look at the associations (before the event) and bet on the one that your predicted sound is associated with - whether it's the team you wanted to win or not.