Lots of previous work has been done on scoring rules, prediction markets, peer to peer evaluation models but none of these are decision making. Scoring rules is small (qualified) agent based, prediction markets is multi agent but not focused on decision making and peer to peer evaluation is about eliciting information from some ground truth.
Decision making work that has been done has not dealt with conflicting incentives.
We have demonstrated that no decision scoring mechanism is completely robust to intrinsic competing recommender incentives, and that the best performance can be achieved through the use of the Quadratic Scoring Rule. However, when a rational briber is the cause of the competing incentives, we can get a no-manipulation result, as long as the mechanism has sufficient budget relative to the maximum influence that recommenders can have over the decision. For multiple recommenders, we can allow the total recommender influence to grow as √1/n while preserving the same incentives. We also show that dependent recommender beliefs can cause an additional violation of strict-truthfulness, but that this can be resolved with a general, decoupling construction.