19 Responses to NEWS FLASH: Surprise Ruling by Delaware Judge Orders Both Sides To Use Predictive Coding

  1. [...] Ralph Losey and his e-Discovery Team blawg comes this surprising opinion out of Delaware Chancery Court: EOHB, Inc., et al v. HOL [...]

  2. craigball says:

    It’s not clear to me why the Judge is so confident that use of a *common* review tool is calculated to save money or time. Certainly, no one has studied whether training a predictive coding tool against one collection will necessarily produce better results against a different collection when the same issues are implicated. Absent such cross-training advantages, it’s hard to fathom why use of a common predictive coding algorithm will foster economy or efficiency here (compared to use of separate vendors using agreed-upon protocols).

    C’mon. Fess up! Did somebody spike the Kool-Aid?

    Ralph Losey, is that a flask in your pocket? Go to the Vice-Principal’s office NOW!

    • Dominic Jaar says:

      Craig,
      That probably reminds you of one of these discussions we’ve had (interestingly, one with no Kool-Aid involved whatsoever): It was last year, prior to my conference on cutting-edge eDiscovery technologies (including predictive coding) at the ABA Techshow. I was sharing my view of a future world where the courts would offer a big black box in which parties would dump their documents and which would spit out relevant material based on the motions filed and a couple of key exhibits (in other words, a prehistoric electronic judge). I remember how you told me that it would perhaps happen when you’ll be dead but certainly not in the USA, perhaps in Canada because we’re “so warm ‘cause it’s so cold”…
      While I’ll admit this order does not suggest that your government will make any investment in the short term that would improve (I would have written “maintain” but thought it’s a bit late…) access to justice, it certainly gets you closer to my vision than I would have expected, at least, while you’re alive! 
      That being said, I’ve been suggesting for years now (before the advent of predictive coding) that, while we wait for parties and lawyers to abide by (read: read and understand) the TSC Cooperation Proclamation or more simply to master technology as you and very few other lawyers do, they should be ordered to use the same vendors. At first, I suggested that as an in-house counsel because I realized that:
      1) Vendors unitary costs were inversely proportional to the amount of data they were ingesting;
      2) By having one case repository (with different access privileges), duplicates and afterward near-duplicates could be dealt with proactively;
      3) Documents could be produced my merely providing access to the documents that should be produced, which would:
      a. serve to educate lawyers and judges to become technology agnostic review platform wise;
      b. reduce the exchange (unsecured in many case) and duplication of documents on too many servers and computers (something I truly hated for the company I worked for) over which corporate clients have no control;
      4) The vendor could act independently, as opposed to supporting whatever claim made by their client, thereby providing the court with an objective view about what?, why?, how?, etc. perhaps even a master/neutral to break a tie.
      Nowadays, with PC, I’m even fonder of the idea. Most of these technologies improve their reliability proportionally to the volume of data they ingest. Accordingly, dataset 1 could be comingled with dataset 2. The tool would cluster the new dataset and show each party only the documents that belonged to them in each cluster. In order to take it a step further, while agreeing with you that “no one has studied [publicly proven] whether training a review tool with predictive coding technology against one collection will necessarily produce better results against a different collection when the same issues are implicated”, I would suggest that agreeing with the other party about concepts/issue codes and seeds/training/example documents for each of them, and running the audited result against the comingled dataset would at least set aside any debate about culling methodology… Furthermore, I would bet that the precision and recall rates would be higher, and the resulting responsive documents more relevant than if each party had used predictive coding (or any other methodology) on their own.
      While I’m not in a position to defend the right of the court to render such orders in your jurisdiction, I certainly applaud the fact that the abovementioned unpublished study will hit the public record. I truly hope the right vendor and technology are used!

  3. matt says:

    Thanks for sharing Ralph. This takes judicial activism to a frightening new level.

  4. [...] I urge you to take a look at the interesting dialogue in the commentary section of my last blog: NEWS FLASH: Surprise Ruling by Delaware Judge Orders Both Sides To Use Predictive Coding. Craig Ball, the Texan from New York who is no stranger to controversy,  begins with a short [...]

  5. [...] NEWS FLASH: Surprise Ruling by Delaware Judge Orders Both Sides To Use Predictive Coding [...]

  6. Interesting – in a negative sort of way – that so many previous discussions stress the need for cooperation between counsel and then in a situation where cooperation does not seem to be a problem, the bench appears to strong-arm anyway.

  7. [...] NEWS FLASH: Surprise Ruling by Delaware Judge Orders Both Sides To Use Predictive Coding [...]

  8. [...] Losey, an attorney for Jackson Lewis, reported last week that a Delaware judge took matters into his own hands by proactively requiring both parties to show [...]

  9. [...] Losey, an attorney for Jackson Lewis, reported last week that a Delaware judge took matters into his own hands by proactively requiring both parties to show [...]

  10. [...] it is a defense trap, or a plaintiff’s trap, or the latest after EORHB, Inc., that it is a judge’s [...]

  11. [...] Laster in Delaware Chancery Court has made e-discovery history, again, with a surprise bench order requiring both sides to use predictive coding and to use the same vendor,” writes Ralph Losey for the E-Discovery Team blog. “This [...]

  12. [...] Chancery Court, another person much talked about at Georgetown, although he did not attend. See NEWS FLASH: Surprise Ruling by Delaware Judge Orders Both Sides To Use Predictive Coding. There are many Judge Learned Hand types waiting in the wings, I am sure. They need to remind us of [...]

  13. [...] judicial endorsement of TAR.  Da Silva Moore, Kleen Products, Global Aerospace, Actos and EORHB are all interesting, but they have just begun the judicial discussion on the use of TAR.  Far more [...]

  14. [...] NEWS FLASH: Surprise Ruling by Delaware Judge Orders Both Sides To Use Predictive Coding. [...]

  15. […] Delaware Judge and Predictive Coding:  Comments from Ralph Losey, Craig Ball and Dominic Jaar –http://bit.ly/SkifxE (Ralph […]

  16. […] judicial endorsement of TAR.  Da Silva Moore, Kleen Products, Global Aerospace, Actos and EORHB are all interesting, but they have just begun the judicial discussion on the use of TAR.  Far more […]

  17. […] NEWS FLASH: Surprise Ruling by Delaware Judge Orders Both Sides To Use Predictive Coding. […]

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

Follow

Get every new post delivered to your Inbox.

Join 3,106 other followers