Something normally distributed that’s gone on seemingly too long is bound to end shortly; but the longer something in a power-law distribution has gone on, the longer you can expect it to keep going.
Prior distribution assumption might give you a hint about your prediction. Normal or scalefree(power law) ?