Abstract: In the past decade, game-theoretic applications have been successfully deployed in the real world to address security resource allocation challenges. Inspired by the success, researchers have begun focusing on applying game theory to green security domains such as protection of forests, fish, and wildlife, forming a stream of research on Green Security Games (GSGs). We provide an overview of recent advances in GSGs and list the challenges that remained open for future study.

Abstract: In this note, we discuss several settings in which algorithms can provide good outcomes when given only ordinal information We focus especially on voting mechanisms in settings with spacial preferences, and on the notion of distortion

Abstract: A Walrasian equilibrium outcome has a remarkable property: the induced allocation maximizes social welfare while each buyer receives a bundle that maximizes her individual surplus at the given prices. There are, however, two caveats. First, minimal Walrasian prices necessarily induce indifferences. Thus, without coordination, buyers may choose surplus maximizing bundles that conflict with each other. Accordingly, buyers may need to coordinate with one another to arrive at a socially optimal outcome---the prices alone are not sufficient to coordinate the market. Second, although natural auctions converge to Walrasian equilibrium prices on a fixed population, in practice buyers typically observe prices without participating in a price computation process. These prices are not perfect Walrasian equilibrium prices, but we may hope that they still encode distributional information about the market. To better understand the performance of Walrasian prices in light of these two problems, we give two results. First, we propose a mild genericity condition on valuations so that the minimal Walrasian equilibrium prices induce allocations resulting in low overdemand, no matter how the buyers break ties. In fact, under our condition the overdemand of any good can be bounded by 1, which is the best possible at the minimal prices. Second, we use techniques from learning theory to argue that the overdemand and welfare induced by a price vector converge to their expectations uniformly over the class of all price vectors, with sample complexity linear and quadratic in the number of goods in the market respectively. The latter results make no assumption on the form of the valuation functions.

Abstract: Understanding when equilibria are guaranteed to exist is a central theme in economic theory, seemingly unrelated to computation. In this note we survey our main result from [Roughgarden and Talgam-Cohen 2015], which shows that the existence of pricing equilibria is inextricably connected to the computational complexity of related optimization problems: demand oracles, revenue-maximization and welfare-maximization. We demonstrate how this relationship implies, under suitable complexity assumptions, a host of impossibility results. We also suggest a complexity-theoretic explanation for the lack of useful extensions of the Walrasian equilibrium concept: such extensions seem to require the invention of novel polynomial-time algorithms for welfare-maximization.

Abstract: We show that algorithms that follow the relax-and-round paradigm translate approximation guarantees into Price of Anarchy guarantees, provided that the rounding is oblivious and the relaxation is smooth. We use this meta result to obtain simple, near-optimal mechanisms for a broad range of optimization problems such as combinatorial auctions, the maximum traveling salesman problem, and packing integer programs. In each case the resulting mechanism matches or beats the performance guarantees of known mechanisms.

Abstract: The first ever human vs. computer no-limit Texas hold 'em competition took place from April 24--May 8, 2015 at River's Casino in Pittsburgh, PA. In this article I present my thoughts on the competition design, agent architecture, and lessons learned.

Abstract: Fair allocation of indivisible objects under ordinal preferences is an important problem. Unfortunately, a fairness notion like envy- freeness is both incompatible with Pareto optimality and is also NP-complete to achieve. To tackle this predicament, we consider a different notion of fairness, namely proportionality. We frame allocation of indivisible objects as randomized assignment but with integrality requirements. We then use the stochastic dominance relation to define two natural notions of proportionality. Since an assignment may not exist even for the weaker notion of proportionality, we propose relaxations of the concepts --- optimal weak proportionality and optimal proportionality. For both concepts, we propose algorithms to compute fair assignments under ordinal preferences. Both new fairness concepts appear to be desirable in view of the following: they are compatible with Pareto optimality, admit efficient algorithms to compute them, are based on proportionality, and are guaranteed to exist.

Abstract: This is the second annual collection of profiles of the junior faculty job market candidates of the SIGecom community. The twenty four candidates for 2017 are listed alphabetically and indexed by research areas that define the interests of the community. The candidates can be contacted individually or via the moderated mailing list ecom-candidates2017@acm.org.

Abstract: Optimal mechanisms are often prohibitively complicated, leading to serious obstacles both in theory and in bridging theory and practice. Consider the problem of a monopolist seller facing a single additive buyer with independent valuations over n heterogeneous items. Even in this simple setting, it is known that optimal (revenue-maximizing) mechanisms may require randomization [Hart and Reny 2012], use menus of infinite size [Daskalakis et al. 2015], and may be computationally intractable [Daskalakis et al. 2014].In a letter here last year, Babiaoff et al. [Babaioff et al. 2014a] described their attempt to alleviate the problem by showing that a constant fraction of the optimal revenue can be obtained by a simple mechanism. In this letter we argue in favor of a related research direction: finding the optimal simple mechanism. We survey our recent results in this setting [Rubinstein 2016] and draw attention to the question of what is a "simple" mechanism?

Abstract: In this letter we briefly survey our recent work [Cai et al. 2016]. In it, we provide a new duality theory for Bayesian mechanism design which is quite general, and applies for any objective the designer wishes to optimize, and for arbitrary agent valuations. We then apply our theory to auction design settings with many independent buyers who have independent values for many items, and are able to provide a unified proof of several recent exciting works on this front [Hart and Nisan 2012; Li and Yao 2013; Babaioff et al. 2014; Yao 2015; Chawla et al. 2007; Chawla et al. 2010; Chawla et al. 2015]. These works all show that simple mechanisms are approximately optimal in various settings. In some cases, our principled approach yields greatly improved approximation ratios as well.

Abstract: We study dynamic pricing policies for ridesharing platforms such as Lyft and Uber. On one hand these platforms are two-sided: this requires economic models that capture the incentives of both drivers and passengers. On the other hand, these platforms support high temporal-resolution for data collection and pricing: this requires stochastic models that capture the dynamics of drivers and passengers in the system.We summarize our main results from [Banerjee et al. 2015], in which we study the role of dynamic pricing in ridesharing platforms using a queueing-theoretic economic model. We build a model of two-sided ridesharing platforms that captures both the stochastic dynamics of the marketplace and the strategic decisions of drivers, passengers and the platform. We show how our model can help explain the success of dynamic pricing in practice: in particular, we argue that the benefit of dynamic pricing over static pricing is not in the optimal performance, but rather, in the robustness of its performance to uncertainty in system parameters.

Abstract: We study the causal effects of financial incentives on the quality of crowdwork. We focus on performance-based payments (PBPs), bonus payments awarded to workers for producing high quality work. We design and run randomized behavioral experiments on the popular crowdsourcing platform Amazon Mechanical Turk with the goal of understanding when, where, and why PBPs help, identifying properties of the payment, payment structure, and the task itself that make them most effective. We provide examples of tasks for which PBPs do improve quality. For such tasks, the effectiveness of PBPs is not too sensitive to the threshold for quality required to receive the bonus, while the magnitude of the bonus must be large enough to make the reward salient. We also present examples of tasks for which PBPs do not improve quality. Our results suggest that for PBPs to improve quality, the task must be effort-responsive: the task must allow workers to produce higher quality work by exerting more effort. We also give a simple method to determine if a task is effort-responsive a priori. Furthermore, our experiments suggest that all payments on Mechanical Turk are, to some degree, implicitly performance-based in that workers believe their work may be rejected if their performance is sufficiently poor. In the full version of this paper, we propose a new model of worker behavior that extends the standard principal-agent model from economics to include a worker's subjective beliefs about his likelihood of being paid, and show that the predictions of this model are in line with our experimental findings. This model may be useful as a foundation for theoretical studies of incentives in crowdsourcing markets.