The Tower card showing up right before layoffs feels like proof, but that's exactly how confirmation bias works - you're filing away the hits and forgetting the misses. Every tarot deck has cards about conflict, change, loss, upheaval. In any given week, *something* stressful usually happens to someone, so when you pull a card that could vaguely apply, your brain locks onto it as predictive.
Apps make this worse because they're designed to feel meaningful. They use artwork, dramatic language, and broad interpretations that apply to almost any situation. "The Tower" sounds way spookier than "a major change," so when your company announces layoffs you remember the reading. But if the app had given you the Ace of Pentacles instead and nothing happened, you'd have scrolled past it and forgotten it existed.
Track it properly if you want real data: write down every reading and prediction, the date, and what actually happens. Give yourself a specific timeframe (say, one week) and see how many genuinely come true versus how many are vague enough to match anything. Most people who do this find the hit rate drops to pure chance levels once you stop cherry-picking the matches.