They exist: nothing goes well on them; even though the game may be changed the ill-luck remains. Two tries should be enough to tell if one is in luck to-day or not. Everything is in process of change, even the mind, and no one is always wise... But a shrewd person will not decide on the day's luck by a single piece of good or bad fortune.
Baltasar Gracián, 1647 (tr. Joseph Jacobs, 1892)
Drop the astrology and keep the observation, which is sound: performance is not constant. Sleep, health, mood, and a hundred invisible factors mean that some days your judgment is sharp and some days it is not, and the dangerous part is that the bad day does not announce itself, it just feels like the world being difficult. The two-try test is a practical instrument. When two independent attempts go badly, treat it as information about you rather than about the world, and switch modes: stop deciding, stop negotiating, stop sending anything consequential, and do the maintenance work that a poor day cannot damage.
Practice
- Keep a maintenance list for bad days: filing, cleaning, routine calls, exercise. Work that cannot be spoiled by poor judgment.
- Apply the two-try rule before blaming circumstances; two misfires in a morning usually means the instrument, not the terrain.
- Protect your best hours for the decisions that matter, and never schedule the irreversible ones at the end of a hard week.
Use this when
- Use this before signing, sending, or confronting; a single postponement costs nothing and prevents the costliest class of error.
- Use it to read others too: the boss having a bad day is weather, and today is not the day for your proposal (maxim 235 times the ask).
Ignore this when
- Don't build a superstition; this is about your own state, not about omens, and the fix is rest, not ritual.
- Don't use bad days as a standing excuse; the log will tell you honestly whether you are having off-days or avoiding work.
A picture of it
The bid was due Friday and by eleven on Thursday morning two things had gone wrong: a calculation error caught by an assistant, and a call that ended with the client sounding puzzled. The old version of him would have pushed through and submitted that evening. Instead he shifted the afternoon to filing and a long walk, slept properly, and rebuilt the bid on Friday morning in three hours. The submitted version was materially different in two numbers, both of which mattered. He had not become smarter overnight. He had stopped operating on a day when the instrument was reading wrong.
Two misfires in a morning means the problem is you. Switch to maintenance and decide tomorrow.