Step-by-step, predictive coding is gaining the acceptance in US litigation which was anticipated for it during last year. The latest advance is that the judge in Global Aerospace Inc. et al., v. Landow Aviation, L.P. dba Dulles Jet Center, et al who approved the defendant’s use of predictive coding has now gone on to approve its results. In a jurisdiction which is driven to the point of obsession by precedent in the broadest sense of that word, almost to the exclusion of first principles, this matters very much.
The story is briefly summarised in an article by Legal Technology News called Predictive Coding Completed in Global Aerospace Case. More detail can be found in an article on the OrcaTec website called Global Aerospace: Predictive Coding Results Approved By Judge for 1st Time. It was OrcTec’s software which was used to achieve the approved results.
Other jurisdictions are faintly bemused by the idea that a judge’s approval in a particular case should matter more than the objectively measurable results themselves, but that’s US discovery for you. The bit which matters lies in this paragraph from the OrcaTec article:
Upon post-predictive coding testing, OrcaPredict was found to have achieved 81% Recall plus 80% Precision. Precision means that 80% of the documents OrcaPredict predicted would be responsive were found upon review actually to be responsive.
Given that the target set by the court was 75% Recall where human review generally achieves 50% or less, it is hard to see that judge could have come to any different conclusion. What matters is that time and money has been saved conducting an exercise whose output is measurable. A starting document population of 1.3 billion was reduced to roughly 173,000 documents in an exercise of which Thomas Gricks of Schnader Harrison Segal and Lewis, the defendants’ lawyers, said
Using a strict manual review process, just culling the irrelevant documents would likely have taken as much as 20,000 man-hours to review, at an anticipated cost of $1.5 million. We were able to generate results that measurably exceed typical manual review, at a fraction of the cost, and we were prepared to produce the vast majority of the responsive documents within a tight timeframe.
The effect of this decision, when added to other developments in US jurisprudence, is effectively that the party who objects to the use of technology assisted review, at what will almost inevitably be a higher cost and over a longer timeframe, will have to justify that extra cost and time. The strongest case which a lawyer can make against predictive coding is that manual review was good enough for his or her predecessors. They did not have to deal with the volumes and costs which we face today.