Psychologists at William and Mary conducted a similar experiment to test whether participants could accurately predict who the winner of the conclave will be based only on photos. They were not asked to predict the winner but to assign a number on trustworthiness and other variables based on the face. The results were then collated.
I have attached the full paper below that describes their methodology and data. However, here are some excerpts:
Much prior research has established that face appearance often influences assessments of others and decisions about them. For instance, the Todorov, Mandisodza, Goren, and Hall (2005) presented participants with images of faces from competing electoral candidates and asked them to rate them on a variety of features, such as trustworthiness and competence. Participants had no knowledge of the candidates except for the face images, but their ratings enabled better than chance predictions of election outcomes. The differences in ratings were even linearly related to the winning candidates margin of victory. The information involved in these types of assessments seem to be gathered from even very brief exposures to the stimuli (Ballew & Todorov, 2007; Willis & Todorov, 2006).
This case is a bit different in that all of the cardinals know each other (somewhat)and are more deeply engaged in the process than the average voter. Still:
The goal of the current study was to assess whether this strategy could be used to predict the outcome of the cardinal conclave tasked with selecting the next Pope. It provides a good test case, because the selection of a Pope is a decision based on detailed information by a group of experts. Congressional elections are decided by many thousands of votes from a popular electorate exposed to media-driven campaigns that can provide conflicting, misleading, and often negative information. In such an environment, there may be no clearly preferential candidate, such that the appearance of the face may be all that is needed to tip the scales. If the same principles apply to the decisions of a conclave of highly educated, experienced, senior leaders, it would suggest that the power of a face to drive decisions is truly powerful.
They conclude that, so far, it appears, based on their research that O'Malley will be selected. This is consistent with buzz from reporters in Rome. Nonetheless it is possible that another top scoring candidate will be selected.
A less specific prediction is that one of the top scoring cardinals will be chosen as the next Pope. Thus, if not O’Malley, then Erdo, or Sandri will be chosen. If none of these three is chosen then perhaps face appearance does not play as big a role in this selection process as in others.
If their results stand and one of those three is elected, then that certainly adds further reliability to studies around how people arrive at judgements.