Taking departure in the ScienceAtHome platform, we present two studies on individual and collective problem solving via citizen science games: In the first study we examined players’ social learning strategies when faced with a real quantum physics problem (through a gamified interface). Preliminary results suggest that players engage in adaptive search both in their asocial and social learning strategies, meaning that negative feedback induces larger search distances, while positive feedback translates into smaller, more conservative moves.

In the second study we aim to experimentally test a setup similar to Mason and Watts’ (2012) Wildcat Wells game. We have developed an agent-based (NK) model to establish hypothesis on how collectives solve problems. Preliminary results show that across all parameters, a “not invented here” penalty (a premium added to a solution, should it be derived from one’s own efforts) increases the system performance as measured by the number of the maximum is reached. Results seem to be due the fact that, without the penalty, potential successful local search moves are replaced with social learning moves which drive informational diversity out.  These hypothesis are to be tested in a game available at ScienceAtHome.