Using Millnet’s Technical Expertise to Advance Your Legal Drafting in Predictive Coding and Technology Assisted Review Cases

Author: Kathleen Kristiansen, Senior eDiscovery Consultant, Millnet

Background to Pyrrho

The Millnet Consulting Team provided insights, workflows and recommendations regarding Predictive Coding for its client Taylor Wessing.  These were incorporated into Associate Edward Spencer’s witness statement on behalf of the Fourth Defendant in Pyrrho Investments v MWB Property Ltd.

Master Matthews states in Pyrrho that his landmark judgment “owes a considerable debt” to Spencer’s witness statement on Predictive Coding.  This document aims to show you how you can also incorporate the same technical expertise into any legal brief on Predictive Coding.

In this document we are using the term Predictive Coding to describe the process whereby software is used to analyse the conceptual content of a document to determine whether a document is “Relevant” or “Not Relevant.”

While Pyrrho was the first time the English High Courts approved the use of Predictive Coding, this template can be used for positioning your case worldwide for Predictive Coding and Technology Assisted Review.

The Pyrrho Judgment – Master Matthews’ 10 Factors to Include in Your Legal Brief

Following Pyrrho and the more recent BCA Trading case in the UK, the question on the mind of every litigator with more than 100,000 documents has become: “is my case suitable for predictive coding and technology assisted review?”  Shortly this question is followed by: “how do I describe this technology to my client, to the other side and, ultimately, to the court?”

Master Matthews included in the Pyrrho judgment a helpful list of 10 factors to consider in your witness statement or request.  Millnet can help you consider these points with the requisite technical jargon and costs expertise.  After you have decided that your case is suitable for Predictive Coding, with consultation from a technical expert, it is time to update your legal drafting with the appropriate technical language.

Master Matthews’ well publicised ‘10 factors’ can be helpfully split into 5 factors for the legal team and 5 factors for your technical experts, such as Millnet:

Pyrrho factors for the legal team to consider and include:

  1. Other Jurisdictions’ use or approval of Predictive Coding or Technology Assisted Review and relevant case law
  2. The Civil Procedure Rules and Practice Directions do not prohibit Predictive Coding
  3. Time before trial
  4. The parties agreement of the use of Predictive Coding
  5. The value of the claims

Pyrrho factors that your technical consultants at Millnet can assist you with:

  1. The total number of electronic documents
  2. The cost of a linear manual review
  3. The cost of using Predictive Coding
  4. How Predictive Coding leads to accurate outcomes
  5. How Predictive Coding can be more accurate than a linear review

Factors that Millnet Consultants can address in drafting your statement on Predictive Coding:  

Total Document Counts
Millnet can help you estimate the number of documents based on the size of your data set or give concrete post-deduplication numbers following data processing.  Typically, your case is ripe for a predictive coding argument if it includes over 100,000 documents for review.  In the case of Pyrrho, there were 3.1 million documents.

Costs and Proportionality of Manual Review
While proportionality arguments do not rely solely on costs, this is an important factor and essential to understand as soon as possible in any litigation.  From the total document count, we can extrapolate the cost of a traditional linear document review using outsourced or law firm resources.  In Pyrrho, the estimated cost of manually reviewing the 3.1 million documents was several million pounds.  Master Matthews found this to be disproportionate to the claim, which concerned an amount in the tens of millions.   

Costs of Technology Assisted Review or Predictive Coding
Using document counts and Millnet’s expertise in estimating how many rounds and documents a subject matter expert would need to spend training the system, we can extrapolate the cost of your Predictive Coding exercise.  In Pyrrho, the estimated cost of the Predictive Coding exercise ranged from £182,000 to £469,000.  The wide spread of the estimate was the result of comparing both insourced and outsourced solutions for ultimate review of the documents found “relevant” by the system.  Master Matthews found even this wide range of values to be more proportionate to the value of the claim.      

Accuracy of Predictive Coding
Another important point made in the Pyrrho case by Master Matthews is his finding that there is no evidence that using software leads to a less accurate outcome than a fully-manual process.  Taking it one step further, he stated that Predictive Coding actually may be more accurate than using “dozens of lower-grade fee-earners” as is typical in a standard, large-scale document review.  Millnet can help you articulate why your case is likely to obtain more accurate results as a Predictive Coding matter, as well as the statistics and algorithms behind the technology.

Predictive Coding as a Diagram:


Additional Documentation

Every day Millnet is positioning its clients to take the lead in cases on Predictive Coding and Technology Assisted Review.  We provide technical language for your disclosure/discovery protocols, keyword analysis, round reports, and Electronic Documents Questionnaire (“EDQ”), as well as estimated costs for settlement discussions.  Whilst most of these matters never make headlines, our clients are advised to flag the potential use of review technology early to couple their legal expertise with our leading technical expertise for their case.

Image Credit

Senior eDiscovery Consultant

Kathleen is a qualified solicitor with over 10 years’ experience in the European legal and eDiscovery market with a particular focus on optimising review workflows and positioning clients for judicial approval of technology assisted review.  After training and qualifying as a solicitor at Slaughter and May, she practiced as a litigation associate at Gibson Dunn in London.

Following a secondment as in-house counsel at UBS in Switzerland, Kathleen gained hands on technical experience as a Senior Project Manager at Epiq Systems.  Most recently, she led an eDiscovery and Document Review department where her cost budgeting and advice on Technology Assisted Review was ultimately incorporated into the Pyrrho Investments High Court decision.

As a former Magic Circle and Big Law solicitor with hands on Relativity experience, Kathleen brings a unique perspective to the team helping to quickly answer questions at the intersection of law and technology.

This entry was posted in News. Bookmark the permalink.
  • Contact

    Contact Us

  • Approved Infrastructure