You can't tell till you bet.
Seriously. I know of no way to tell for sure if the price of anything is going down tomorrow, next week, next month, or next year. The only thing I know is if the price is lower now that it was at some prior moment.
When I am wrong I stop the loss.
I believe it helps to test methods using the computer and many years of price data.
Jesse Livermore authored a book that I like a lot and it covers short selling. Livermore's book is titled "How To Trade In Stocks", ISBN # 0-934380-20-1. The book is about 111 pages long. I remember paper trading the method and testing profitable. The problem with Jesse Livermore's method is he does not have a computer and must develop his methods manually. Mr. Livermore does not appear to analyze risk extensively, he vaguely mentions risk once or twice in the book. I suspect Jesse Livermore's lack of focus on risk is the cause of some or all his big losses.
Judging from trading volume I believe the Livermore method is widely used.