Occam's Razor Mindset

Stop the debate. Occam's Razor says: test the simplest, most checkable explanation first—not because it's true, but because it fails fastest. A 40% drop overnight is a step change, not a slow decay. That points to a discrete event: a code deploy, a config change, a tracking pixel break, or a payment processor outage. Competitor launches rarely cause overnight cliffs; UX regressions usually degrade gradually.

Your first move: check your analytics and deployment logs for the last 48 hours. Did you ship anything? Did your A/B test tool flip a flag? Did your CDN cache suddenly serve a stale bundle? Look at the funnel step-by-step—where exactly did users stop? If signup page load errors spiked, it's a technical glitch. If traffic is fine but form submissions vanished, it's likely a tracking or integration issue.

Only after ruling out the mundane should you even glance at the competitor. The razor isn't about being right—it's about minimizing wasted hours. Start with the hypothesis that requires the fewest assumptions: "We broke something." Investigate that first. You'll know within an hour, not a day.

Start with the drug, not the zebra

You're holding the razor correctly. Two days of lethargy, a new antihypertensive started last week — that's a temporal correlation any intern should chase before entertaining autoimmunity. The most probable explanation is a medication side effect: beta-blockers, central alpha-agonists, or even a diuretic-induced electrolyte shift can cause fatigue within days of initiation.

The practical sequence

1. Review the medication list — confirm the exact agent, dose, and start date.
2. Check vital signs — orthostatic blood pressure and heart rate; a drop suggests the drug is working too well.
3. Run a basic metabolic panel — sodium, potassium, creatinine, and glucose. This catches electrolyte disturbances and renal impairment without a single exotic test.

Why this is the right move

Occam's Razor isn't about certainty; it's about testing the hypothesis with the fewest assumptions first. A medication side effect requires one assumption — the drug. Autoimmune disease requires several: new pathophysiology, latent onset, and a trigger. If the basic workup is clean, then escalate. But don't order the barrage until the simple explanation has failed. Hoofbeats, not zebras — especially when the patient just changed their prescription.

Check the Wifi First — But Only as a Test, Not a Verdict

Your two explanations are not equal. The teaching-gap theory assumes a systemic failure across a full lesson. The wifi-drop theory assumes one technical glitch affecting a subset. Fewer assumptions, faster to verify, and directly observable: pull the quiz logs. Did those five students actually open the quiz? Did their attempts show partial loads or timeouts? That takes ten minutes. Rewriting curriculum takes weeks.

When you hear hoofbeats, think horses, not zebras — but still confirm it’s a horse.

If the logs show the quiz loaded fully for all five, the wifi theory dies cleanly. Then, and only then, examine the teaching gap: compare their answers to the same question from prior quizzes, check if the other 25 students aced it. If the five are scattered across different reasoning errors rather than one shared misconception, that points to individual gaps, not a curriculum hole. Use Occam’s razor as a filter, not a final judge. Test the simplest explanation first, and let the evidence either confirm it or force you to the next. You’re not choosing what’s true — you’re choosing what to check first, and the wifi is the cheapest check that yields the clearest signal.