I like the puzzle prediction effect.
I have some notes.
Often, the performer presents a framed prediction covered in a cloth on a stand. Then, they take something like a bag full of puzzle pieces and get the spectator to choose a puzzle piece at random. The cloth gets removed to display a completed matching puzzle with one missing piece. The missing piece of the puzzle is the one the spectator chose randomly.
My first note is more of a question. I don't have the answer, but does the selection process need to be random? Why is our instinct with a trick like this to make the selection process as random as possible? Doing so turns the trick into a coincidence effect. Should an effect like this be a coincidence trick?
Often, a coincidence-matching effect is a coincidence at both ends. That's what makes it feel like a coincidence. Two spectators randomly select face-down cards, and they both match. Voila. A coincidence.
Sometimes, a coincidence-matching effect is a coincidence at the performer’s end only. That's what makes it feel like a coincidence. The magician hits a pool ball around a table covered in face-down cards. The spectator freely names a card, and they match. Miracle. Coincidence.
What I find odd about the puzzle trick is that it's only random on the spectator's end. The magician was incredibly purposeful with their prediction. They spent several hours doing a puzzle, glueing it into a nice frame and leaving out one puzzle piece. Their end of this trick is far from coincidental.
This unravels the effect for me.
To counter it, I feel like the selection of the puzzle piece needs to involve zero choices from the spectator or one very intentional choice. Either that or you really sell the fact that you genuinely lost one piece of a puzzle at random.