Initial growth rates of malware epidemics fail to predict their reach

Abstract Empirical studies show that epidemiological models based on an epidemic’s initial spread rate often fail to predict the true scale of epidemic. Most epidemics with a rapid early rise die out before affecting significant fraction population, whereas pace some pandemics is rather modest. Recent suggest this could be due heterogeneity target population’s susceptibility. We study computer malware ecosystem exhibiting mechanisms resembling those biological systems while offering details unavailable for human epidemics. Rather than comparing models, we directly estimate reach from new and vastly more complete data parallel domain, offers superior insight as concerns outbreaks. find highly heterogeneous distribution susceptibilities, nearly all outbreaks initially over-affecting tail distribution, then collapsing quickly once depleted. This mechanism restricts correlation between growth its total reach, thus preventing majority epidemics, including fast-growing outbreaks, reaching macroscopic population. The few pervasive malwares distinguish themselves via following key trait: they avoid infecting tail, preferentially targeting computers unaffected by typical malware.