Predictive Coding and Hamlet: Madam, how like you this play?
In Hamlet, his mother the queen is quoted as saying, “The lady doth protest too much, me thinks.”
In Shakespeare’s time the word “protest” was meant as the lady vows or declares solemnly, too much.
So do the devotees of predictive coding, protest too much.
A recent article in favor of further brick walls around the mysterious-to-many predictive coding process stated, “In Da Silva Moore, Judge Peck recognized that TAR, also known as computer-assisted review or predictive coding, is an acceptable way to search for relevant electronically stored information. “
Judge Peck did no such thing. What Judge Peck did was to set forth predictive coding as a glamour-child for technology assisted review (TAR), but he did not completely reinvent TAR by setting predictive coding as the only method.
Furthermore, Judge Peck did not detail the specifics of the predictive coding model he had in mind. And, since the parties had agreed to the processes to be followed, he neither sanctioned any particular method nor did he discourage transparency. As the Honorable Judge John M. Facciola describes it, Judge Peck commented on the “high road” of litigating:
“In approving this arrangement, the court observed that such “transparency” with respect to seed sets was proper and important. Given the convenience that disclosure provided, the court recommended that this practice should generally be adopted: Such transparency allows the opposing counsel (and the Court) to be more comfortable with computer-assisted review, reducing fears about the so-called “black box” of the technology. This Court highly recommends that counsel in future cases be willing to at least discuss, if not agree to, such transparency in the computer-assisted review process.”
Should seed sets be subject to work product protections as the author promotes in the article? Certainly there are types of seed sets that must enjoy privilege protection; judgmental seed sets in which the lawyer has chosen certain documents would be protected. But, randomly generated seed sets have no legitimate basis for protections and could be helpful in an opposing party’s becoming secure with the producing party’s methods.
Judge Peck suggests other means are available to test whether predictive coding has been properly done: determining whether there are gaps in the production, quality control review of samples from the set of non-responsive documents, and statistical estimation of recall at the end of the review.
Statistical recall is, unfortunately, not a complete picture of the methods used in creating the seed sets, testing the seed sets or in arriving at the final predictive coding results. Gaps in the discovery is a time tested method used even when we were dealing with old fashioned paper; but if the statistical estimations appear in order, will the Court respond favorably to an objecting party? If the non-responsive document sets show nothing, will that be the end for any objections to the process?
The author concludes her article with: “Consequently, we are left with one more opinion approving the use of TAR; but, the debate will continue as to how much parties should “meet and confer” before employing such technology, and whether courts have the authority to require production of seed sets without party agreement.
Yet another win for producing parties? Call out the cheerleaders!
Will the bevy of articles rejoicing over predictive coding from vendors, defense counsel, and, yes, even judges be such a weight around our collective necks that not only will predictive coding be the absolute default procedure, but we will be successful in burying all the other types of TAR by simply christening predictive coding with the new moniker?
Predictive coding is a tool. It is a viable tool. There are other tools that can be just as viable, given the right circumstances. A media campaign in favor of a particular approach to ESI management is unwise. In addition, it is unhealthy to the discovery process to encourage cooperation, even collaboration except in a single area, controlled only by one party and subject to the potential manipulation of that party; housed behind brick walls.