I’ve waited a bit to write this post because I wanted to see what my colleagues were saying about the latest opinion from Judge Peck. In ED circles, a new ESI opinion from Judge Peck is more highly anticipated than the next Bruce Springsteen CD, except maybe in the Facciola household where The Boss is revered just below … well, actually I’m not sure his status is below that of anything in the Facciola home except, of course, Mrs. Facciola.
Earlier this week Judge Peck opined in the case of Hyles v. New York City ( No. 10 Civ. 3119 (S.D.N.Y. Aug. 1, 2016) that proportionality trumped TAR. And he didn’t beat around the bush about it, stating in the very first paragraph of the order:
“The key issue is whether, at plaintiff Hyles’ request, the defendant City (i.e., the responding party) can be forced to use TAR (technology assisted review, aka predictive coding) when the City prefers to use keyword searching. The short answer is a decisive “NO.” “
His reasoning was, of course, that the absent an agreement of the parties as to a specific search protocol, the applicable standard is the Sedona Principle 6, which holds that
Responding parties are best situated to evaluate the procedures, methodologies, and technologies appropriate for preserving and producing their own electronically stored information. (The Sedona Principles: Second Edition, Best Practices Recommendations & Principles for Addressing Electronic Document Production, Principle 6 , www.TheSedonaConference.org).
Well the Twitterverse exploded with comments about how Judge Peck declined to order the parties to use TAR this and Judge Peck pulls back on TAR enthusiasm that. In the spirit of the impending football season all I can say is “Come on man”
First off, we all know that Judge Peck has never ordered anyone to use TAR. He’s entered orders in several cases where the parties agreed to use TAR, a factor that had not happened in the Hyles case.
And this lack of a fundamental understanding of the fact that in the legal profession (that’s profession, not industry) the word “order” can be either a noun or a verb. I lay this lack of understanding squarely at the feet of the ever increasing assimilation of eDiscovery software and services companies by people who have no legal background. I’ve said it before and I won’t go off on that particular rant again here.
Rather I’d like to just point out one part of the proportionality debate that seems to be missing. Judge Peck in Hyles mentions cooperation and speed of process, and refers to the Tax Court decision in Dynamo Holdings Ltd. P’ship v. Comm’r of Internal Revenue 143 T.C. 9, 2014 WL 4636526 at *3 (2014, which spoke to the same considerations.
But in his decision, J Peck notes on page 3 that “… in general, TAR is cheaper, more efficient and superior to keyword searching.”. I think that if I say “not so fast” one more time in a column that I’m going to hear from Lee Corsos attorneys but I have to say that I don’t believe the issue of “cheaper” has been clearly established. Even in Hyles, J Peck says at Fn 2 that “The Court acknowledges that some vendor pricing models charge more for TAR than for keywords. Usually any such extra cost is more than offset by cost savings in review time.”
I respectfully argue that there has been no empirical validation of that statement that I have seen. Now it may very well be that vendors have filed briefs in matters that address that point or even presented substantiation for such a positon during the submission of attorney fee claims in cases that I have not seen. So what I’d really like to see is a case study that shows the efficiency based on price savings not time savings of TAR on a particular set of documents.
Not that time savings is irrelevant or should not even be the deciding factor. But it should be just that: a factor. One factor of several to be weighed in the process of which tool to use.
I note with interest that David Horrigan, E-Discovery Counsel and Legal Content Director at kCura, in a blog post on another case this week, the 10th US Circuit Court of Appeals’ decision last week in Xiong v. Knight Trans (see http://blog.kcura.com/relativity/blog/facebook-follies-10th-circuit-sees-a-keyword-fail-in-social-media-e-discovery ) mentioned that “We’ve always been skeptical of attempts to use the 1985 Blair and Maron study to argue that keyword searches are only 20 percent accurate”. I’ve also disagreed with the general proposition that TAR is always better than keyword searching ( see my post Reports of the Death of Keyword Search Are Greatly Exaggerated at http://www.advanceddiscovery.com/blog/2016/04/reports-death-keyword-searching-ediscovery-exaggerated/ ) and I think the point here is the same one that Judge Peck makes on page 5 of the Order in Hyles.
“ It is not up to the Court, or the requesting party (Hyles), to force the City as the responding party to use TAR when it prefers to use keyword searching. … While Hyles may well be correct that production using keywords may not be as complete as it would be if TAR were used (7/18/16 Ltr. at 4-5), the standard is not perfection, or using the “best” tool (see 7/18/16 Ltr. at 4), but whether the search results are reasonable and proportional. Cf. Fed. R. Civ. P. 26(g)(1)(B). “
And as one consultant in our field (who wished to stay above the fray so I will not use his name) said to me recently:
“Choice of predictive coding, managed review (with or without validated search), or just validated search does NOT pre-determine success. … So a bad protocol might lead to poor results in all three, and a good protocol might turn south in all three if calibration and QC is missing, or if it is improperly applied. Ultimately it is only “better” if a reasonable production is made without substantial critical documents left on the cutting room floor.”
As I said several years ago in another column about another issue (https://docnativeblog.wordpress.com/2009/02/18/its-the-archer-not-the-arrow/ ) :
“It’s the archer not the arrow”.