That’s a good strategy, but personally, I believe in the "wisdom of the crowd" theory and tend to apply it whenever I participate in prediction contests (whether it’s a bitcoin price prediction or anything else).
What I do is to wait for the last moment before the deadline and calculate the average of all the submitted predictions (sum of all predictions devided by the number of submissions). The result will be my prediction.

Theoretically, this would work if other participants do their own research and don’t just submit random numbers. Also, the more participants there are, the closer the average will get to the actual price.