This study explores predictors of school choice decisions in American college football recruitment. We combine data about individual athletes’ recruiting activities with social media data to predict which school the athlete will choose among those that have offered him a scholarship. While previous works have approached school choice as a rational decision process, our results indicate that a bounded rationality model incorporating social factors and heuristics may be more appropriate. We explore how the actions taken by athletes during recruitment can be interpreted as early signals of athletes’ preferences and find that models incorporating social media features consistently outperform the baseline model with only off-line recruiting features. In addition to better understanding the school choice decision, this work can help coaches to effectively allocate recruiting resources and inform social media strategies during recruitment.

Published as: Ali E. Abbas, Jay Simon, Chris Smith (2017) Introduction to the Special Issue on Decision Analysis and Social Media. Decision Analysis 14(4):227-228. https://doi.org/10.1287/deca.2017.0364

Female carriers of a BRCA1 or BRCA2 genetic mutation face significantly elevated risks of cancer, with 45%–65% of women developing breast cancer and 15%–39% developing ovarian cancer in their lifetimes. Prophylactic surgery options to reduce cancer risk include a bilateral mastectomy (BM), bilateral salpingo-oophorectomy (BSO), or both surgeries. No comprehensive model providing recommendations at which age to perform the surgeries to optimize quality-adjusted life years (QALYs) exists. Using available clinical data, we develop a Markov decision process model of a mutation carrier’s health states and corresponding transitions, including age-dependent breast and ovarian cancer risk, distribution of each cancer subtype and stage, and mortality. We convert the problem to a linear program to solve for the optimal surgery sequence that maximizes the carrier’s expected lifetime QALYs under varying assumptions about individual patient preferences on postsurgery quality of life, fertility considerations, advances in cancer screening or treatment, and others. Baseline results demonstrate that a QALY-maximizing sequence recommends BM between ages 30 and 60 and BSO after age 40. Surgeries are recommended later for BRCA2 mutation carriers, given their lower risk for both cancers compared to BRCA1 mutation carriers. We derive structural properties from the model and show that when a carrier has already undergone one surgery, there exists an optimal control limit beyond which performing the other surgery is always QALY maximizing.

Stochastic dominance is often used to study preference between different distributions of outcomes. In the stochastic dominance literature, distributions of outcomes are often assumed to be known. However, complete distribution information is rarely available in practice. In this paper, we study weighted almost first-degree stochastic dominance (WAFSD) under limited distribution information. In particular, we show that it is possible to determine WAFSD with linear canonical utility based on expected rewards when outcomes are bounded from below. Furthermore, we illustrate how WAFSD based on more general forms of canonical utility functions can be ensured when additional moment information is available. The key insight is that finite distribution moments can be sufficient for revealing clear preferences in practice, despite the fact that finite distribution moments are generally insufficient for ensuring preferences across all utility functions.

We present a study on the evolution of publications about applications of proper scoring rules. Specifically, we consider articles reporting the use of proper scoring rules when either measuring the accuracy of forecasts or for inducing honest reporting of private information within a certain context. Our analysis of a data set containing 201 articles published between 1950 and 2015 suggests that there has been a tremendous increase in the number of published articles about proper scoring rules over the years. Moreover, the weather/climate, prediction markets, psychology, and energy domains are the four most popular application areas. After providing some insights on how proper scoring rules are applied in different domains, we analyze the publication outlets where the articles in our data set were published. In this regard, we find that an increasing number of articles are now being published in conference proceedings related to artificial intelligence, as opposed to traditional academic journals. We conclude this review by suggesting that the wisdom-of-crowds phenomenon might be a driving force behind the recent popularity of proper scoring rules.

We use results from a multiyear, geopolitical forecasting tournament to highlight the ability of the contribution weighted model [Budescu DV, Chen E (2015) Identifying expertise to extract the wisdom of crowds. Management Sci. 61(2):267–280] to capture and exploit expertise. We show that the model performs better when judges gain expertise from manipulations such as training in probabilistic reasoning and collaborative interaction from serving on teams. We document the model’s robustness using probability judgments from early, middle, and late phases of the forecasting period and by showing its strong performance in the presence of hypothetical malevolent forecasters. The model is highly cost-effective: it operates well, even with random attrition, as the number of judges shrinks and information on their past performance is reduced.

Supporting strong, resilient, and integrated critical infrastructure is vital to upholding the U.S. economy and its national security. Arguably, no sector of the economy could exist without the reliable and predictable networks on which it depends. To date, the research on interdependent infrastructure has concentrated on describing the sources of the dependencies and developing models for predicting performance and cascading disruptions after a hazardous event. However, these models fail to capture the perspective of the operators of these networks and how competing, independent objectives lead to suboptimal investment decisions and hence suboptimal network performance. Rather than focus on how interdependent infrastructures operate and possibly fail, we take the perspective of their operators and ask why they make the decisions that they do. The goal of this paper is to demonstrate how strategic interdependencies may impact performance of coupled systems by shifting investments away from what is collectively best toward decisions that are more myopic and optimal from the perspective of a single infrastructure. Through our model, we make inferences on the level of investments networks make, relate this to performance, and provide policy recommendations on how to promote reliable infrastructure.