A Game Theorist Perspective on E-Discovery

Can the Pain of E-Discovery Shift Practitioners from the Prisoner’s Dilemma to a Nash Equilibrium?

The application of Game Theory to the practice of law is not new. Before e-Discovery took center stage the classic adversarial stance taken by many litigators is often likened to the Prisoner’s Dilemma; where the incentives are aligned to encourage parties to defect in their own self interest in lieu of cooperating for a better mutual reward. The 1979 modification to the Federal Rules of Civil Procedure and the influx of Electronically Stored Information (ESI) have only served to increase the likelihood of non-cooperation in the discovery space. As the scale of data has steadily increased and the cost associated with discovery has grown exponentially, parties have adopted an increasingly black and white approach to the entire discovery process.

Using the lens of Game Theory, this piece addresses the factors that entrenched the adversarial approach and created a Litigator’s (prisoner) Dilemma, the factors driving change and what it might take to get lawyers to pursue a mutually self-interested solution and reach a Pareto Optimal Nash Equilibrium.

Game Theory Defined

At its core, Game Theory is simply “the study of mathematical models of conflict and cooperation between intelligent rational decision-makers.”^[1] It was developed in the 1940’s and applied extensively to economics in the 1950’s and later to group dynamics, psychology and an array of other spaces that involve interactive decision making between self-interested or rational beings.

The limited or imperfect information present in the 26 (f) meet and confer conferences as well as the high risk/reward potential in large scale litigation make the current e-Discovery process a prime real-world case study for game theorists. In the context of e-Discovery, the self interested rational beings are Plaintiffs, Defendants and parties in government investigations and the stakes are primarily monetary, reputational and time-related.

The Prisoner’s Dilemma in a Legal Context

The game that is most often used in discussing the conflict present in litigation is the Prisoner’s Dilemma; It is one of the most well known and understandable iterations of game theory.

The Scenario

Two prisoners are held for trial in separate cells with no means of communication. The prosecutor offers each of them a deal. He also disclosed to each that the deal was made to the other. The deal he offered is this:

• If you will confess that the two of you committed the crime and the other guy denies it, we will let you go free and send him up for five years.
• If you both cooperate and deny the crime, we have enough circumstantial evidence to put both of you away for three years.
• If both of you confess to the crime, then you’ll both get two year sentences.^[2]

This game may begin with cooperation or with competition; but the more times the game is played the greater the likelihood of defection. And an initially cooperative player ceases to be cooperative if they are burned once.

The Scenario Applied to E-Discovery

When applied to the e-Discovery space the Litigator’s (Prisoner) Dilemma, two parties with limited (or no) information about the proposed actions of the opposing side must come to the 26(f) meet and confer to determine the scope for discovery in a case and they have the option to cooperate or take the non-cooperative stance.

The decision of both parties is made simultaneously, but influenced by past rounds of the game (cases). There is a similar imperfect dissemination of information in this situation and the risk reward calculus is tied to monetary interests and overall outcome of the case. As in the original prisoner’s dilemma, there are three possible outcomes that mirror the traditional Prisoner’s dilemma matrix:

• If both parties cooperate they will have the mutual benefit of positive judicial inference, curtailed Discovery cost as a result of tailored dates, and terms or use of technology.
• If only one party cooperates the cooperating party may be stuck with exorbitant e-Discovery costs e-Discovery may be used to force settlement in cases that have merit.
• If neither side cooperates, they may avoid exorbitant e-Discovery cost for only one side, but are still stuck with a large e-Discovery burden and broader, additional costs for their clients, as well as a potentially negative Judicial inference.

What is Nash Equilibrium?

In technical terms, a Nash Equilibrium is present when neither participant could have a better outcome given the likely decision of the other party. The movie “A Beautiful Mind” offered a real life demonstration of attaining Nash Equilibrium with a scenario where four male students are at a bar with one extremely attractive woman and four less attractive women with her. In one scenario all the males fawn on the attractive woman, thereby alienating her friends and in the end the attractive woman leaves with none of the men. In a second scenario, each male focused his attention on the woman he had the highest likelihood of attracting given the likely choices of his friends. Each male ended up with a date but it was with one of the four less attractive friends.

The risk reward matrix for e-Discovery going into the 26(f) conference places the Nash Equilibrium squarely in the mutual non-cooperation category. Asking for everything and giving nothing creates a cost burden on the front end of a case that can drive settlement of matters with merit. If one party cooperates and the other opts to be adversarial massive discovery costs or over broad request often resulted; incentivizing parties once “burned” to adopt non-cooperative stances looking forward to avoid the larger loss of cooperating when the other side opts not to.

Not all Nash Equilibriums are created the same, and they do not always provide the socially or financially optimal outcomes. There is a sub category within Game Theory called Pareto Optimization of the Nash Equilibrium that can be attained. An outcome of a game is Pareto Optimized if there is no other outcome that makes every player at least as well off and at least one player strictly better off. That is, a Pareto Optimal outcome cannot be improved upon without hurting at least one player. Often, a Nash Equilibrium is not Pareto Optimal implying that the players’ payoffs can all be increased.^[3]

 

Nash Equilibrium and Pareto Optimization in a Post-ESI World?

Nash Equilibriums do not always provide the socially or financially optimal outcomes. In the example from “A Beautiful Mind” all four men ended up with a date, but no one ended up with the ideal outcome of a date with the most attractive woman ^[3]. E-Discovery is currently in a similar situation where the objectively optimal outcome is not being attained. In the first calculus the risk reward matrix for e-Discovery going into the 26(f) conference places the Nash Equilibrium squarely in the mutual non-cooperation category. Parties once “burned” by non-cooperation opt to adopt non-cooperative stances looking forward to avoid the larger loss of cooperating when the other side opts not to. This is known as obtaining a Non-Pareto Optimized Nash Equilibrium.

An outcome of a game is considered Pareto Optimized if there is no other outcome that makes every player at least as well off and at least one player strictly better off. That is, a Pareto Optimal outcome cannot be improved upon without hurting at least one player. Often, a Nash Equilibrium is not Pareto Optimal implying that the players’ payoffs can all be increased^[4]. While we are currently in a Non-Optimized Nash Equilibrium, a Pareto Optimization can be attained through modifying the e-Discovery game itself.

Factors Driving Change

Recent and subtle changes in e-Discovery practice and case law are shifting the tides in favor of a more cooperative approach and Pareto Optimization. Formerly, the discovery process was asymmetrically weighted in favor of the plaintiff’s side with the bulk of ESI tied up in the larger defendants. Often the small plaintiff would send a sweeping discovery request asking for years of data, archives and extremely broad keywords in the hope that the Goliath Defendant would opt to settle in lieu of paying massive e-Discovery bills. Today, even the smallest plaintiff has some sort of digital foot print which is potentially relevant and discoverable and this shift is turning the tides in favor of a more cooperative approach.  From email to shared external hard drives, archived data on company servers, a single plaintiff still has a large potential e-Discovery burden.

Additionally, the bench is increasingly weighing cooperation of the parties in deciding whether to apply cost shifting from the producing party to the requesting party of other monetary sanctions.  Federal Rules of Civil Procedure provide an avenue for shifting such costs – Fed. R. Civ. P. 26(b)(2)(B) – when the ESI sought by the requesting party is not reasonably accessible or according to  Rules 26(b)(2)(C) and 26(c) when courts find that the burden or expense of proposed discovery outweighs its likely benefit, or that they need to protect a party from undue burden or expense.

This cost shifting combined with an increase of sanctions for contentious discovery requests has helped balance the risk/reward calculus in determining whether or not to cooperate in the discovery process. Options such as rule 502 (b)’s claw-back, production of data with minimal review with the caveat to claw back and date determined to be privileged or confidential, has also served to make the e-Discovery process a more even playing field. Although each of these solutions has limitations and potentially negative repercussions, they are all still serving to level the playing field.

Cooperation Proclamation

The high stakes of e-Discovery drove the bench and Sedona Conference to espouse a more cooperative stance in an effort to reintroduce proportionality to the increasingly burdensome discovery process.  The conference argued that “[o]ver contentions discovery is a cost that has outstripped any advantage in the face of ESI and the data deluge”^[5] The Cooperation Proclamation was published in 2008 and is a short document that argues that if lawyers work together during the discovery phase, the merits of the underlying dispute are more likely to get a fair hearing. Specifically, it calls on lawyers to “work more collaboratively during the discovery phase so that greater time and attention (and money) can be spent on litigating the merits of the underlying dispute.”^[6]

Initially, many hailed the proclamation as altruistic and lacking teeth since the first iteration of the Cooperation Proclamation, but in the end the perceived altruistic behavior can serve attorneys self interest and save clients’ money on both sides of the table.  The threat of sanctions, increasing likelihood of cost-shifting loser pays all, reduction of asymmetry in ESI burden between plaintiff and defendants and technological advancement that have all contributed to this fundamental shift in perspective. These factors have reigned in discovery cost and contributed to reshaping the matrix and risk reward model for e-Discovery.

Concluding Thoughts

When looking at the e-Discovery Litigators Dilemma as it was first conceived the only way to ensure the more socially optimized solution was to count on people to act in an altruistic, trusting manner. But, scholars, lawyers, judges and e-Discovery professionals looking to reduce the adversarial nature of e-Discovery ought to stop trying to change the human nature and its tendency towards self interested decisions and risk aversion; and look instead to changing the game itself. We are already beginning to see the case law to support the risk reward calculus shifting to a Pareto Optimized Model.  If the Judiciary continue to reward efficient and proportional discovery practices and punish people leveraging e-Discovery as a weapon to force settlement as well as look to alternatives to reign in cost (predictive analytics, or advanced technology) we will continue to see a positive trend toward cooperative e-Discovery.

Citations
[1] Roger B. Myerson (1991). Game Theory: Analysis of Conflict, Harvard University Press, p. 1. Chapter-preview links, pp. vii-x.
[1] Game Theory formal definition:
Let (S, f) be a game with n players, where Si is the strategy set for player i, S=S1 X S2 … X Sn is the set of strategy profiles and f=(f1(x), …, fn(x)) is the payoff function for x   S. Let xibe a strategy profile of player i and x-i be a strategy profile of all players except for player i. When each player i   {1, …, n} chooses strategy xi resulting in strategy profile x = (x1, …, xn)then player i obtains payoff fi(x). Note that the payoff depends on the strategy profile chosen, i.e., on the strategy chosen by player i as well as the strategies chosen by all the other players. A strategy profile x* S is a Nash equilibrium (NE) if no unilateral deviation in strategy by any single player is profitable for that player, that is:

[2] http://perspicuity.net/sd/pd-brf.html
[3] http://www.youtube.com/watch?v=CemLiSI5ox8
[4] Shor, Mikhael, “Pareto Optimal” Dictionary of Game Theory Terms, Game Theory .net, <http://www.gametheory.net/dictionary/ParetoOptimal.html> Web accessed: 2.12.12
[5 & 6] http://www.thesedonaconference.org/content/tsc_cooperation_proclamation/
bullseyeviewofcooperation.pdf

Related Posts

No related posts.

Add new comment

Name (required)

Email (will be hidden) (required)

Web

Notify me of follow-up comments by email.

Notify me of new posts by email.