«isn't significant proof of decline, although since it's 2 of 5, a 40% decline seems notable to me» Any percentage value is irrelevant because a change of ±2 a

«isn't significant proof of decline, although since it's 2 of 5, a 40% decline seems notable to me» Any percentage value is irrelevant because a change of ±2 among 500 retailers is statistically insignificant. If it was ±20 it would be significant. What if we went from 1 to 2 accepting BTC, would you say "acceptance went up 50% among top 500 retailers?" It would be highly misleading and improper to report it as such.