Legal Analytics for Counsel Selection

When a company enters litigation, it must select litigation counsel. The management of the company would usually call its peers and ask about past experiences, review counsels’ fields of technical expertise, evaluate their success rates, and then invite a handful of law firms to make a pitch. The invited attorneys would propose a case strategy and predict their probability of success. The management would then select a counsel based on the presentations given and their personal impressions.

Ex Parte is a start-up formed about two years ago that is trying to change appellate counsel selection for patent cases.[1] The company collects data for every appellate case since 2004 and combines it with a database of all practicing lawyers. It then uses a proprietary algorithm to identify a lawyer with the highest probability of success for a specific case. When Ex Parte analyzed the recent appeal by 10X Genomics, the company predicted that an average lawyer would have a nine percent probability of winning the case but hiring one particular litigator would increase the probability of success to 25 percent.

The attorney recommended to 10X Genomics by Ex Parte expressed concerns over the ethical implications of this analysis. For one, the lawyer no longer did patent appeals. Also, according to the American Bar Association’s model rules for professional responsibility, lawyers should not create an “unjustified expectation” of a future result based on their part performance.[2] But this is what Ex Parte may be doing – creating expectations of future performance based on past statistics.

In the U.S., the work of Ex Parte is protected by First Amendment rights and companies are not restricted in using its predictions to choose a litigation counsel. However, if a lawyer were to project his own probability of success based on his past record, that would be considered unethical. France has a law completely banning software companies from providing legal and judicial analytics.[3]

Additional concerns revolve over the ability of the software to account for specific facts of the case and the quality of data available from law firm websites. In the case with 10X Genomics, Ex Parte recommended a lawyer who no longer practiced patent appeals because the law firm website had not been updated with his current status. Ex Parte’s information may also interfere with the client’s relationship with counsel, where the software analysis of the probability of success in a case contradicts the counsel’s.

Also, “the law of small numbers” cautions that percentages may appear greatly superior or inferior based on small data sets – the appellate practice is so small that most lawyers only get a few cases in their lifetime. In the example of 10X Genomics, the counsel recommended by Ex Parte had litigated 34 cases, of which 20 were identified as patent appeal cases. The Court of Appeals for the Federal Circuit, which is the sole appellate court for patent appeals, has heard more than 4,000 patent cases since 2004.

With computers infiltrating almost every aspect of our lives, the use of computer-assisted legal analytics is only expected to increase with time. With new technology comes the need to evaluate new ethical considerations. On one hand, making predictions about future success based on past legal achievements may be considered borderline unethical by the legal community. On the other hand, it can be argued that a company such as Ex Parte merely provides additional information and another viewpoint for clients with which to make informed decisions. Litigations are sometimes driven by unfounded beliefs in the strengths of a company’s position, and data from Ex Parte may inject rationality into corporate decision-making and facilitate settlements.

In addition to companies that are entering litigation, Ex Parte’s services are also retained by law firms seeking better marketing positions and financial firms that fund litigation. With litigation being a $100 billion a year industry, any attempt to help quantify outcomes is an understandable goal. The question is whether and where the lines should be drawn for an ethical balancing act.

[1] Roy Strom, “The algorithm will hire your patent lawyer now,” Bloomberg Law, 8/1/2019.

[2] ABA’s Model Rules of Professional Conduct, Rule 7.1.

[3] Sam Skolnik, “France’s judicial analytics ban unlikely to catch on in the U.S.,” Bloomberg Law, 6/5/2019.