Happily, we have seen little evidence of fraud during RM. However, due diligence requires we check at least monthly, and in Round 8 we disqualified 11 accounts for improper collaboration.
We cannot offer details, but we can say several would have won survey prizes, and that warning flags included clear ties to a single established forecaster, coincidence of timing and location, and all using a signature survey behavior otherwise unique to that forecaster. When emailed separately, none had a convincing alternate story, and all failed to confirm connections we already knew. In the end we could not completely rule out sock puppet accounts, and we decided to suspend all involved.
RM is very grateful for the diligence of our Forecasters and their continued efforts throughout this research project.
