Game Theory Applied to online Dating

Game Theory Applied to Online Dating

INTRODUCTION

This paper focusses on some key game theory strategies that come into play during the online dating process and between different apps involved as each attempt to better their positioning. The focus will be on some of the key areas that are of significance. This is important to understand because players face tradeoffs by using certain strategies over others and certain apps are more successful than others in helping players achieve their ultimate goal^[1]. We will discuss this in detail later in the paper.

To be successful an app must be first selected by a user, we will discuss different strategies used by different apps and which apps would be better positioned for people looking for a serious relationship versus a casual one. For players we define success when they are able to meet their objectives i.e a player looking for a serious relationship is successful when they get a serious relationship. However, in order to understand that we will first talk about the various factors one must consider (both as a player or founder of the app) while coming up with a dominant strategy.

Current Scenario:

Applications:

Applications try to create a “thick “market to increase probability of matching their players for e.g. eHarmony has a higher probability of matching players than say veggiedate.com simply because of the sheer numbers. Hence applications that want to satisfy their consumers should not limit themselves by creating an exclusive market. Another factor to consider is that online dating even in today’s day and age has a negative connotation attached to it with users applying the “lemon law” (George Akerlof) while selecting their mate online. The lemon law or adverse selection states that players believe that there is inherently something lacking in the people who are present on the dating app (this phenomenon is contradictory human behavior since the player themselves is present on the dating app). Applications that thus verify information such as The league where members have to be a part of an Ivy league institutions (the information is verified by the company since you need to enter your school login) can reduce the effect of adverse selection. Finally, apps also face the uncertainty attached to imperfect information, it is hard for players to assess a potential candidate given the limited amount of available information. This lack of information can also be detrimental to the application itself since false representation reflects poorly on both the candidate and the app; it could lead to players leaving the app ( it should be noted here that since most of the apps work on network effects of such negative publicity could lead to higher losses for the app than the player).While this phenomenon cannot be mitigated even in “real” world dating, it can be reduced by applications asking players to link their profiles to their social media ( e.g Facebook or twitter). The assumption here is that players not forethought the game and fabricated their evidence on every social media platform with the intention of manipulating dating websites in helping them find their perfect match ^[2].

Players:

Players who use online dating applications can tweak their image by signaling and cheap talk. The difference between signaling and cheap talk is that cheap talk does not negatively alter the payoff for the players. For example, if a player says that he earns more than a $100000 annually (signal of wealth) and that information turns out to be false the matched player is likely to have penalize the other player for stating a false fact (penalty can be in any form). However, if a player whose height is 5’3 ft. states that he is 5’5 ft. tall he is less likely to get penalized for two reasons i.e. the transaction cost of verifying such information is high (the other player will have to obtain documents that have the exact stated height or carry a measuring tape) and two because the other player might not view exaggeration of height as a deal breaker. Commonly used strategies are men who have pictures with an animal to display that they are affectionate and friendly, women tend to display travel photos to indicate that they are cultured and curious. Thus players when formulating their online personas should assess the criteria that he/she would like to highlight and the amount of exaggeration his/ her potential mate would find tolerable. Another consideration for the player is whether they should signal their true intention behind joining the app ( e.g whether they are looking for a serious or casual relationship). It is always beneficial to indicate something which is close to the value for not at the value, this is because human behavior is irrational and perceives signals in the dating context in the form that they want to hear. To elaborate there are a multitude of stories where men have indicated that they are only looking for casual relationships but women assume that this can be changed over time. This contoured view of moves is what makes the application of game theory to online dating applications extremely interesting. Having set the stage, we will now discuss different applications and the strategies (both dominant and mixed) which are used in each of the applications.

Match.com /e-harmony/ok-Cupid^[3]

Match.com and e-harmony are online dating services with web sites designed “specifically to match single men and women with each other for long-term relationship.”^[4]

Prospective members complete a proprietary questionnaire about their characteristics, beliefs, values, emotional health and skills.  These traits are entered by the players themselves, hence playing in to the signaling phenomenon ( which of your traits would your potential partner find attractive). Since information is entered by the player there is a chance that it can be mis- represented, however since the assumed goal of the site is to get a long term relationship, mis- representation will only reduce payoff's for the player. Matching algorithms, which the company believes matches people's core traits and values to replicate the traits of happy couples, use these answers to match members with users the company believes will be compatible. The software technology also evaluates users' behavioral data such as average time spent on the site. There are two games a player need to play here one where he/ she needs use backward induction to decide what traits to highlight to get the ideal partner. The second is the behavioral component which tracks data, here a player can access the economist while on the site to display that he/ she is looking for intellectuals or google search flights if the player is looking for a match that travels or "worldly". There is a fact that needs to be highlighted here that is the site increases your chances of meeting your " ideal " mate but not about how successful a relationship might. This is because human beings have a sense of self distortion and their wants align with what society deems to be compatible with their image of themselves. Combining a distorted self-image and exaggerated data can lead the algorithm to match the player with someone he desires but might not have the highest chance of having a successful relationship with ( which we assume for these sites is the ultimate goal for the players). Hence the dominant strategy in this case would be to display your real traits however not put all the cards on the table I.e  a player should not fabricate / exaggerate information about him/herself , however the should keep their idiosyncrasies at bay at least at the start.