Challenge 3: Explain Why Percentile Beats Average Here — Possible Solution ==================================================================== THE SCENARIO ------------- 100 interactions in a session: 95 take 50ms, 5 take 800ms. WHAT THE AVERAGE WOULD SUGGEST ---------------------------------- Average = ((95 × 50) + (5 × 800)) / 100 = (4,750 + 4,000) / 100 = 8,750 / 100 = 87.5ms An average of 87.5ms would classify this page as "Good" under the chapter's (and Chapter 1's) threshold of 200ms or under — the average gets pulled down close to the 50ms figure that represents the vast majority (95%) of interactions, largely burying the 5 much slower ones inside that blended number. Anyone looking only at the average would reasonably conclude this page feels consistently fast. WHAT A HIGH-PERCENTILE INP SCORE WOULD SHOW ------------------------------------------------ INP specifically reports a high-percentile value (around the 98th percentile, per the chapter) — with 100 interactions, the 98th percentile lands within (or very close to) that group of 5 slow interactions at 800ms, rather than being smoothed out by the 95 fast ones. This means the reported INP would be much closer to 800ms — solidly in the "Poor" range (over 500ms, per Chapter 1's threshold table) — rather than the deceptively reassuring 87.5ms average. WHY THE HIGH-PERCENTILE VIEW IS MORE USEFUL ------------------------------------------------- The chapter's own reasoning applies directly here: "a single terrible interaction sticks in a user's memory even if most interactions felt fine — averaging would wash that out." A real user experiencing this page doesn't feel an averaged 87.5ms; they feel 95 fast, unremarkable interactions AND 5 genuinely sluggish, noticeable ones. Those 5 slow interactions are exactly the kind of experience that damages a user's overall impression of the page's responsiveness, regardless of how many other interactions felt fine — and a metric built around averaging would systematically hide that problem from anyone monitoring the page's Core Web Vitals, while a high-percentile metric surfaces it clearly as something actually worth investigating and fixing.