The conclusions drawn by the TFA are far from proven, albeit based on some kind of "research" (it is claimed). Arguably, the phenomenon, and the issues it raises are at least an order of maginitude more complex, as comments here have amply demonstrated. "Trolls" may serve a useful social purpose after all, because they may provoke reflection on many levels including: the rightness of one's assumptions, awareness of the trolls' pov/talking points (always useful), the trolls' (possibly hidden) agenda, state of mind/mental health, one's own capacity to be easily provoked to anger by something that ought to be ignored or regarded as merely a source of transitory amusement, a challenge that calls for a witty riposte. The possibilities are almost limitless. I believe the 'success' of trolling as a strategy of discourse is highly overrated by TFA, and ephemeral at best.