Hello, everyone. My name is David Burnett. This is part two of my two hour on data analysis and litigation. I'm pleased to be joining Quimbee for this presentation. Um, the. I encourage you to. Watch part one first. So this is. But this covers non-overlapping information. So I encourage you to watch part two as well. So just a very brief recap. This topic is about why. Um. Why? Data is important. In litigation, why it's becoming increasingly important. Data analysis, both at the pleading stage and complaints and related motions to dismiss and also in discovery, both fact and expert discovery. I talked about strategic and economic benefits of using data because data seems empirical, scientific. It seems objective, even though, um, you know, one can quibble about the data and how it's analyzed. Nonetheless, um, it is sort of numerical and has a certain solidity based on that. Um, economically talked about how using big data to prove your cases and to plead them allows plaintiffs to scale up the scope of cases dramatically from individual in scope to market wide or society wide. And the two examples that I talked about were mortgage backed securities litigation, which implicated the entire market for mortgage backed securities. Um. And in terms of society wide, I talked in part one about opioids litigation, where thousands of cases have been filed in every state in the country against companies involved in the market for manufacturing, distributing and selling. Prescription opioids. So the point is that using data is a way to bridge the gap from individual to collective proof. The challenge, as I touched on in part one, is that you have to use that data persuasively. You have to be mindful of the applicable standards for admissibility, whether Di or Daubert or Frye. And, um. Yeah. And then you have to do. Of data and analyze it and explain it in a way that's persuasive to the judge and jury. So again, very briefly recap about my background. I'm currently a senior counsel at Motley Rice, which is one of the country's largest plaintiffs firms. My work involves representing plaintiffs in a broad range of cases opioids, litigation, securities, class actions, mass torts involving pharmaceuticals and also environmental litigation. Where I'm representing the state of New Jersey and a Superfund case. Prior to that, I worked at the litigation funder Burford Capital, where I evaluated whether I thought Burford should invest in cases based on. Uh, legal and economic analysis. And as will explain, the litigation funding industry also used this data. And then prior to that, I worked for ten years at Quinn Emanuel in New York, representing plaintiffs in litigation arising from the financial crisis. Um, the. The reason why I was inspired to present this topic is that data and data analysis has been a recurring theme in my career. Um. One. One thing I didn't mention in part one is that I. I enjoyed math when I was a kid. I feel comfortable with numbers, so not every lawyer is going to enjoy data analysis and feel as comfortable doing it as another. And that's part of the beauty of working on these big cases is that there's kind of a role for everybody. So I've naturally gravitated to this type of work. Um, and so nonetheless, even if you aren't planning to become a data guy and, you know, working with economists and other experts on on numbers intensive analysis, nonetheless think any litigator, whether on the plaintiff or the defense side, should have some um, working understanding of, of of how data can be used and the pros and cons because even if you are the business generator or the rainmaker, um, you know, even if you're the person on top, you still need to have some familiarity with with how the people working in the trenches with the data about how they're doing that. All right. So after having. Um. Oh, and just recap here. As I as I talked about the the challenge with data is you need to present it in clear and compelling form. The way to do that often is to use formulas, spreadsheets, graphs, other visual representations that aggregate a lot of information into a summary form that that makes it easy to understand. Um. The perhaps goes without saying, a lot of this work on big cases where you're using a lot of numbers, there's a lot of information to process. Oftentimes that requires consultants and testifying experts. And that they could be experts in economics, epidemiology, marketing, medicine, statistics, etcetera. And along the way, it's not only experts that are using the data, but in my experience, also employees, whether they're employees of the state, employees of the company, plaintiff or defendant, oftentimes individuals to explain. Uh, and describe complex data sets. Okay. So in the last hour, in part one, I talked about structured finance litigation and litigation, and I talked about opioids litigation. And this hour I will give three more examples from my experience of three other areas of law in which data is frequently used, that is securities, litigation, antitrust litigation and litigation finance in securities. Um, data can be used lots of different ways, but but one way is in proving proving wrongdoing in antitrust, where you're talking about market wide wrongdoing, data can be used to describe patterns of manipulation of markets and in litigation, finance funders will sometimes use data to analyze cases, value and how the funders think the cases are going to turn out. So without further ado, let's talk about those three. Um. Three examples and then I will end with some recommendations, sort of an overview of how to use data in all different phases of litigation. Okay, so, um, this first example in securities class actions is from a case that, um. Where Motley Rice was co-lead counsel for the class, along with Robbins Geller. This is not a case that I worked on, but it's one that I have some familiarity with. My colleagues worked on it. The case was filed in 2016 in California. Plaintiffs alleged that Twitter overstated the amount of users that were using Twitter's platform. Plaintiffs alleged that Twitter fraudulently changed the way that it tracked and reported user engagement in 2015 to mislead shareholders and inflate its stock value. Plaintiffs allege that the company switched from describing monthly active users to daily active users, um, focusing internally on the daily number without disclosing that to investors. And we see that Twitter stock dropped dramatically in value in price and value when the executives admitted on an analyst call that user growth was actually slowing dramatically. So. Um, as you can imagine, just from the way I'm describing it, this case was very heavily premised on the numbers, you know, the daily versus monthly user engagement and how those numbers were changing over time. The suit took issue with statements made in press releases and public filings about monthly active users. Claiming the company switched its performance metrics to focus on daily active users without disclosing it on a motion to dismiss. In 2017, the court agreed that the omission of daily user metrics was misleading. And then on a motion for summary judgment in 2020, the court found that certain statements were misleading and that a jury could conclude that Twitter was at least reckless in failing to disclose declining user engagement. So. Um, the case settled in. Um. In September 2022. Last September, Twitter agreed to pay $809.5 million in settlement. The case settled the morning. The jury selection was set to begin. It was the second largest securities class action ever obtained in the Ninth Circuit. So this case again didnt work on this personally, but I know that it settled in no large part because of the strength of testimony from plaintiffs expert who testified that Twitter should have disclosed the daily user metrics, not just the monthly user metrics. Um, so we, we the plaintiffs had engaged a professor who gave persuasive testimony about the core issue of the daily versus monthly metrics, and that professor was able to understand the topic and explain it. Working with my colleagues was able to explain that topic in a way that was persuasive, um, to the court. And evidently Twitter felt that if they had gone to trial, if they had gone forward with the trial, that they must have faced a risk of loss there in part because of that persuasive testimony about the data. So that was a that was a good experience. Um, uh, city of Providence versus Bass. This was a cautionary tale that I'm introducing here. Um, work on this is a securities class action, um, where Motley Rice was co-lead counsel with two other firms. This is a class action filed against the stock exchanges, the stock of the New York Stock Exchange, Nasdaq and bats on behalf of all investors. Uh, in the in the stock markets, except for high frequency traders. So the allegation was that the stock exchanges offered products and services to high frequency trading firms known as Hfts, which allowed the Hfts to gain an advantage over ordinary stock market investors. And we allege that the stock exchange failed. The three stock exchanges failed to disclose those products to the market and they failed to disclose the advantage they were giving to HFT. Hft is our type of investment firm whose investment model was premised on beating other investors to trades, using sophisticated technology to get microscopic gains on other investors. And if you add up those microscopic gains over the long term, those hfts could make a huge amount of money through technological advantages over others. And the allegation is that. Um, services offered by the stock exchanges like proprietary data products and co-location of trading centers, as well as complex order types allowed Hfts to get a benefit over other investors. The lawsuit was inspired in part by Michael Lewis's book Flash Boys, which people may have heard of. This case was incredibly data intensive because we are alleging manipulation of millions, if not billions, of trades across the entire stock market over many years. This is a very complicated case, very incredibly broad reaching. Mean, you're talking about the entire stock market over many years. Um, so there were a number of factual challenges to this case. We had to identify who the hfts were, how they were different from ordinary investors, how they were using these different trading strategies to manipulate the market, and then how the stock exchanges were involved in that. Um, we also had to show, you know, how, how the HFT is trading worked on a granular basis. The role of stock exchanges in in that and we had to show how the ordinary investors, the non hfts were harmed. We also had to show that the specific individual lead plaintiffs had standing by showing that their specific investments had been adversely affected. Um, so as you can imagine from the way I'm describing this, this ended up being a very complicated and difficult case to prove. So in discovery, um, we obtained from the defendants, um, enormous, enormous trove of trading data, um, showing, uh, records of stock market purchases and sales over, um, a market over the entire market over a long period of time. So as you can imagine, um, this data set was ginormous and even ingesting the data and, you know, processing it and storing the data was incredibly time consuming and expensive. Um, it was also something that we had to extensively negotiate with the defendants who were concerned about data security. So. Um, the challenge and what we discovered. Was that? It was very hard to match up the different sources of data about individual trades in a way that would allow us to trace the path of a specific investment through all the different intermediate steps and all the different third parties that were involved in one investor's decision to buy buy a stock because they're they're broker dealers and other intermediaries, investment managers who are all involved in the process of executing a trade. And what we determined was that it was very difficult, if not impossible to match up these different sources because there is no one comprehensive source of all the data about trades. The information that the stock exchanges provided us was only part of the picture. So that was an example of this is an example of of the difficulty. Proving proving a case, which is incredibly complicated. On on the issue of data where you're trying to get data from different sources. And in this case, we did not, unlike the cases where we had a lot of the information ahead of time and we knew what to expect from the information that we would get in discovery. In this case, we did not have the trading data before we filed the case and we weren't sure what it would all look like. And the. Data and processing it and matching up the different data sources were more complicated than than I at least had expected. So unfortunately, in this case, it needed it. It led. Whereas the Twitter case led to a great settlement. This case, um, led to dismissal on summary judgment and also granting of the motions in limine to dismiss, to to reject the testimony of our expert. Um, the court found that there was a disconnect between the facts of the case and our theory of defendants liability when compared with our data and our opinions. Um. In short, this case just became a little too unmanageable. There were, you know, data coming from too many different sources and too many layers. So the it's a cautionary tale about the challenges of using data. In some cases, we're able to thread the needle. And like in the opioid litigation, you're able to layer data from different sources using different experts to reliably explain that data and present it and vouch for it. And it all meshes together well. Other times it just turns out that data is is incomplete or it's too hard to work with or it doesn't show what you hope it would show. And so not every case works out well when it comes to the data analysis and proving your case. All right. So the next example of a antitrust class action that we'll talk about that was very data intensive is in re credit default swaps. So this is one of a series of huge antitrust class actions based on manipulation of different investment markets. This was a case that was brought by Quinn Emanuel. Quinn Emanuel got involved in these cases while I worked there, but was not personally involved in this case since I was focusing on those cases that I described earlier. Quinn Emanuel represented the class of plaintiffs in this credit default swaps case. The subject matter is complex and very difficult to explain, but in short. Credit default swaps are a type of financial product whose value depends on an underlying asset. In this case, the underlying asset was a debt. Sophisticated investors used credit default swaps as a way to balance out risks in their portfolios. Um, as you can tell, it's a very complicated topic. It involves very complicated data analysis, which we will talk about. I will spoil the lead by letting you know that, um, the good news for Quinn Emanuel and the plaintiffs was that these, this case settled on an aggregate basis across 12 defendants or so for 1.8, $6 billion. So that gives you a sense of the the weight of the allegations and the the evidence that plaintiffs were able to collect, including persuasive data analysis that would have allowed them presumably to do well at trial, which is presumably why the defendants agreed to settle for that. So the claims here were claims under the federal Sherman Antitrust Act and also under state law for breach of contract and unjust enrichment. The plaintiffs were an institutional investors. They alleged that the banks conspired to keep new participants from entering the credit default swaps market and also keeping the price for trading and credit default swaps artificially high, which plaintiffs alleged cost them tens of billions of dollars. The class was composed of all investors who bought and sold credit default swaps over a multi-year period. The total annual market for credit default swaps is valued in the tens of trillions of dollars each year, but it fluctuates widely with economic conditions. So this, just like those cases, involved the entire market for mortgage backed securities and just like opioids, involve the entire market for prescription opioids across the entire country. This credit default swaps case also involved an entire enormous market. In this case, for these credit default swaps, um, on the level of tens of trillions of dollars. Um, as with the cases, um, my former colleagues at Quinn Emanuel got passed motions to dismiss by incorporating extensive quantitative analysis prior to filing the complaint to help them plead wrongdoing. So the lawyers worked with consultants, statisticians who had who looked at public data about the credit default swaps market and were able to, um, from that plausibly allege patterns of manipulation of market by looking for aberrations in the data. So the complaint was able to include detailed description of how the credit default swaps market was allegedly being manipulated. So the Judge Cote, from the judge, Denise Cote, largely denied defendant's motion to dismiss in September 2014, holding in part that plaintiffs alleged injuries are not speculative because to support the claim the plaintiffs had paid inflated prices when trading in credit default swaps. Quote, the complaint references several sources. Modeling by CME and Citadel Research performed by some dealer defendants, statements of SEC employees and an economic analysis commissioned by plaintiffs so that that last. That last category, the economic analysis commissioned by plaintiffs. That's a reference to the complaint statistical analysis that I just referred to, where the the plaintiffs worked with consultants to help plead their allegations about manipulation of the market. Um, and that use of data was found sufficient on on a motion to dismiss to, to help pass the motion. And it's significant also that we see reference in this in this complaint or sorry, in the in the quote that I gave you. Judge Cote referred to modeling by CME and Citadel research by dealer defendants. In other words, it wasn't just data from plaintiffs. It was also data that they were able to pull from other sources that helped to plead their allegations. All right. So that was a case that turned out well. Um, another case that I'll describe here, this was another manual antitrust. Market wide harm this case. This case also involves Cohen, Milstein and Labaton Sucharow. The case accused a group of banks of rigging auctions for Treasury Department bonds and other securities, including setting, sharing confidential information and setting prices, as well as competition in a secondary market. Unfortunately, the judge in this case dismissed the complaint twice, finding the plaintiffs failed to provide smoking guns of the alleged conspiracy, even though the complaint had included a statistical analysis of Treasury auction results. So like in the credit default swaps case, the plaintiffs here had retained consultants who came up with a statistical analysis of public data about Treasury auctions. The analysis purported to show that patterns of market manipulation changed when there were disclosures about, um, about a DOJ investigation of the market. So the analysis was intended to show parallel activity by the banks that reflected a conspiracy. But the judge held that the analysis was not sufficiently detailed to single out individual banks. Plaintiffs had argued that they couldn't get any more detailed data at the pleading stage. Nonetheless, like the footbridge case that I referred to in part one involving the court here found that. At the pleading stage, the plaintiffs needed more detail about the specific defendant. So, um, you know, just an example of how, um, even with the best data you can find, you're not always going to win. In the cases. We were fortunate that once we got the detailed forensic analysis that we plaintiffs were able to use at the, uh, at the pleading stage, in those cases, we were able to get a nearly 100% success rate on motions to dismiss. Um, nonetheless, this antitrust case and the, um, yeah, this, this, this Treasury auctions case is, is an example of how even with good data you're not always going to. Um. All right. Next example from antitrust. Is this commodity exchange? Gold futures options trading. Uh, last example that I'll give here. Um, this case involved an alleged conspiracy to manipulate the price of gold over an eight year period. Um, the judge, Judge Caproni, denied defendant's motion to dismiss in 2016in part because plaintiffs included allegations about 300,000 quotes for gold prices. The data The court found that this data was sufficient at the pleading stage. Um, however, here, this, this decision that I'm going to talk about involves sort of a side issue about one cautionary note that in the process of plaintiffs using, um, confidential proprietary statistical analysis at the pleading stage where they're using undisclosed experts. Um, defendants sometimes will, um, press to get that analysis in discovery. I saw this in the cases where defendants would sometimes press us to give the full, um, full result of our forensic analysis that we use to plead misrepresentation. Um, same thing in, in, in this antitrust case where the defendants, um, whether purely for strategic reasons or because they actually needed the, the information the defendants pressed to get plaintiffs. Um, modeling that they had used in the complaint. So defendants filed a motion to compel because the plaintiff said, no, that's our work product. Um. The plaintiffs. Also in this case, just like in the cases number one, we said the modeling is work product. Number two, we would say it's not relevant because we're not going to use that initial analysis at the pleading stage, but rather, I'm sorry, we're not going to use that after the pleading stage. So once you get into discovery, whether an case or these antitrust cases, you will get more data and perhaps better data that will allow you to prove your case, which is different than the data that you were able to come up with at the pre complaint stage. So we plaintiffs would argue that the data that they used at the at the pre complaint level was not. Not relevant and defendants are not entitled to it. However, in this case, Judge Caproni granted plaintiffs granted defendant's motion to compel required plaintiffs to give defendants all materials created or considered by the plaintiff's consultants prior to filing the amended complaint. Um, Sarah had already produced materials from the consultant's analysis to the extent it had been disclosed in the complaint. Um. But did not want to give analysis related to material that was not in the complaint. The court inferred that the plaintiffs had withheld the consultant's analysis that did not support their allegations and had only disclosed the favorable analysis. And the court held that that would be unfair to the defendants. Um, the court found that that would be, quote, selective and misleading. It would benefit plaintiffs and prejudice defendants. Because plaintiffs statistical presentation was central to the court's decision. The plaintiffs have plausibly alleged defendant's participation in price manipulation. So I include this decision here just as a warning about one of the side effects that can arise from a heavy analysis on statistics in a complaint is that defendants. You know as a countermeasure. My. Force you or try to force you to disclose. Disclose that data and the courts may grant that. So that's something to be mindful of when you're working with tenants or when you're working with consultants. So just like the cases and the opioids cases and just like the Twitter case, um, the securities class action that I described, the antitrust cases tended to work out well. And again, just like those other cases, a lot of that should be credited to the, uh, the ability to successfully harness, um, large data sets that were compelling at the pleading stage and then also in discovery to be able to prove these very broad, wide ranging, complex cases. So the credit default swaps case settled for 1.9 billion in aggregate across 12 different banks. Is to fix another class action. Antitrust settled for 500 million. Gold settled for 150 million, and Libor settled for 182 million. All right. So I'm going to briefly talk about litigation finance. Um, every litigation funder does things a little differently, but, um, the basic idea is that litigation finance is a specialized area within the private equity world where private funders offer money to plaintiffs or plaintiffs lawyers to bankroll their cases in exchange for getting a return on the funders investment if the case turns out well. So if the case settles, or if you win at court and you get damages, then the funder, um, depending on the the deal, the funder would get some percentage of the plaintiff's lawyers recovery. So I worked at Burford Capital, which is one of the oldest and best known litigation funders as an underwriter. My job was to evaluate individual investments, to decide whether I thought Burford should invest in those cases. And like any underwriter in litigation finance, my job was to look at the legal and economic merits of the potential investment in in a case. Um, and in doing so was trying to predict how the case would turn out to decide whether I thought the case was going to result in a good settlement or, or large damages such that it would be a good risk for us. Um, and the funders assessment of the legal and economic merits then drove how much money, whether we were willing to invest, how much money and on what terms. Um, and that analysis is common to any funder. Um, the way I would do that as an underwriter was to review case law review orders in the case, speak with the lawyers. Um, you know, sometimes you would find that you would conclude that a case was too risky or too small or too difficult. Um. In the process, data ends up being a large part of how litigation funders analyze cases. You know, funders are sort of stripping cases down to their economic, um. Foundation and the funders are looking at a filed case or an unfilled future case. As a financial asset, which has some value based on the degree of harm and the amount of damages and the settlement value and, you know, the leverage that the plaintiffs and the plaintiff's lawyers can bring to bear. Um, so, you know, funders are looking at lawsuits the way that commodities brokers look at a bale of hay as something that has value. And it's just a question of assessing how much value that asset has and therefore whether that asset is a good risk for the funder. So in the process, um, you know, a lot of the analysis that funders do is based on the specific facts of the case. Um, and you know, the, the facts of the case, the relevant case law. Um, but then you're also, many funders will incorporate data analysis because if you're thinking about a lawsuit as an asset, you know, that lawsuit is one of, of thousands of lawsuits out there. Um, and, you know, there are ways to standardize the analysis. So you can, you can look at things like, you know, how long does it typically take a judge in this jurisdiction to rule on this type of case? The goal there is to try and make the process of analyzing lawsuits scientific and not just sort of anecdotal and not just sticking your finger in the air, but trying to make it as precise as possible. So that that was a very interesting way. You know, as a litigator, you're used to thinking about cases just sort of within the four corners of that case typically, whereas with funders, you're you're looking at cases as financial assets and you're thinking about, um, you're thinking about it in terms of, you know, different criteria of financial analysis of assets, in large part by comparing that case to other cases and using data to try and standardize the process and and make the underwriting decision more accurate. Okay. So that was a brief description of litigation finance and how it uses data. Um, brief comment here about vendors. So in my experience, a lot of work involves not only consultants and experts, but also vendors. Um, vendors can be hugely helpful, um, in helping lawyers to process large data sets, um, at all phases of the case. There is a constellation of lawyers out there which which tend to circle lawyers like sharks wanting to their credit to wanting to help lawyers make cases more efficient and effective. Um, the vendors can be super helpful. They can also be expensive. I'm just going to touch on a few vendors that that can be helpful with data analysis. Um. So one example is with document review. As we know, in large cases, um, you can have tens of thousands if not hundreds of thousands of documents involving sometimes millions of pages worth of documents that need to be reviewed. Um, law firms will sometimes use predictive coding to help streamline the process of both offensive and defensive document review. Um. So rather than having a contract attorney look at every single document, every single document, just like the use of statistical sampling in cases that I described, you can review a sample of documents and then have, um, have a computer program apply that analysis to, to a larger sample of documents that need to be reviewed to streamline the process so that not every single individual document needs to be looked at by a human being to get it coded. Um, artificial intelligence is obviously a hot topic in litigation. It's beyond the scope of this presentation. But I will just very briefly note that AI uses enormous computing power to cull through huge datasets, huge batches of information, to identify patterns, and to create new content that could be anything from poems to artwork. Um, there is concern about supplanting the use the role of professionals, um, even including briefs. Um, it's an example of a powerful use of data and litigation. Again, it's beyond the scope of this presentation. It is not something that I've personally used or have much experience with. So I'm not including it here. But I will just, um, note that it's out there. This presentation is focused on my personal experience, which is with human driven data analysis like spreadsheets and simple formulas. And I will note that, um. Across all sorts of different applications in society. People are excited about the potential value of AI to help with advancing knowledge. In my experience, I'm not using AI has been perfectly sufficient to plead and prove complex cases involving datasets, so don't think that AI is would have been necessary to any of these cases. It would be interesting to think about retroactively how AI perhaps could have helped those cases or would have made the outcomes different. But in any case, that's a whole that's a whole separate topic. Um. Other examples of how vendors are used to organize data. One example that I list here is in securities cases. You're looking for the plaintiff with the largest losses, and to do that you need to track, um, institutional investors that have invested in a given company's securities. Um, you know, whether Twitter or any other case. Um, so there are law firms out there that will track, um, portfolios of institutional investors such as pension funds, to keep track of what their losses are related to alleged wrongdoing. So that's one example of how vendors track data in a way that is foundational to securities cases. Um. Another example here is jury selection. There there's a company out there that will offer to help plaintiffs with, um, picking jurors based on a statistical analysis of how, um, past jurors have ruled, um, based on, uh, based on demographic data. So the common theme here is using vendors to try and organize cases, make them more efficient, um, make them possible in some cases, like with tracking client portfolios, to look for losses and to make cases more effective. So, you know, if you can make cases more data driven, more fact driven, more scientific, like when it comes to jury selection, um, you can make decisions as a lawyer that are based on hard evidence rather than just intuition. All right. So in the remaining ten minutes of this presentation, I want to give some overarching comments about some strategic thoughts on how to use data in different phases of litigation and different. Parts of the case. So the first slide is using data and working up a case. Then I'll talk about using data in drafting complaints. In fact, an expert discovery at summary judgment in trial and in mediation and settlement. So the first point here, working up a case, i.e., you know, when a first when a case first comes to you as a plaintiff's lawyer and you're thinking about what do I want to prove, you know, who are the plaintiffs, who are the defendants? What are the causes of action? What is this case all about? The key point here is at this initial strategic stage, you need to already be thinking about do you need data to prove this case? If so, where is that going to come from? Is it reliable? How much will it cost to get it? Who you're going to get it from? What will that data likely show? You can't wait until Discovery to think about data. You can't procrastinate that topic because as we saw with the high frequency trading case. Um, sometimes you get into discovery and you can't get. And end up thanking the whole case. So. You know, part of part of the work of working up a case as a plaintiff's lawyer is to think about evidence and expert testimony, including data. So in opioids cases, for example, where we were looking at where a key topic in the case was the number of fatal and non-fatal overdoses in a given state or county or city. We needed to make sure that we were going to be able to get reliable data about the number of overdoses because that's such a core topic. It would be a problem if you weren't able to get that in discovery. Um, another key point here is that when you're thinking about how to get data, if possible, it it is ideal to not have to rely on defendants or third parties for that information. Best to be able to find that data yourself from some other third party where it's publicly available or where you can buy that information yourself because defendants are going to fight you every way on giving you data. And um, you have less control over the process. Same thing with third parties, even if they aren't. Um. Even, you know, as a non defendant. Third parties are often loyal to the defendants and reluctant to give you information because they don't want to be involved or they don't want to hurt their relationships. So. That. Can be difficult, so it's best to try and get get the data yourself. In addition to the data, you want to think early on about how you're going to process that. You know, what vendors will you need? What consultants will you need, What testifying experts will You need to understand and present the data. So, for example, with the cases, we started lining up statistical experts very early in the case who would be able to help us with sampling. Um, and one note is that within a type of case, you know, within opioid litigation. The good news is that a lot of the analysis overlaps and is in fact very similar across cases. The methodologies are similar and often the vendors and consultants and experts are similar. So that does streamline the analysis a bit. If you come up with one methodology that you're then able to replicate across cases. One other general point here is that you should spend money on data collection in proportion to a case's value. You don't want to spend millions of dollars on consultants and purchasing data and hiring experts if the case isn't worth. Cases that I've talked about today, as you've seen from settlements, are all. They all sell for hundreds of millions, if not billions of dollars, in which case it's okay to pay top dollar for experts, but not in small cases. Um. All right. So in drafting complaints, the same principle of think early on about, you know. The what data you're using, how it will not only get passed a motion to dismiss, but also how it will help you prove a case. So you already need to be thinking about your pleading burden with expert analysis when it comes to the data. The general goal here is to incorporate case specific data and data analysis to get passed a motion to dismiss and avoid the dismissal by by judges on the basis that your claims aren't specific enough. The more data you have from the more sources, the better. And again, none of these cases are. All also are incorporating facts drawn from confidential witnesses, from newspaper articles, from governmental analyses, from other public information that the defendants have put out, um, you know, from academic studies, etcetera. Um, nonetheless, data is, is often critical. So you want to make sure the data is persuasive, that it's authoritative, that it supports your claims, that it's case specific. And in the process of doing so, often lawyers, as I've noted, such as in the and the antitrust cases, lawyers will often use consultants and vendors to help them with pleading and alleging information about data sets in in complaints. Okay. In discovery, you've got fact discovery, followed by expert discovery in some of these cases, all of that, both fact and and expert discovery, um, can be very data heavy. Um. These bullets describe things to think about, in fact, an accurate discovery in these cases. The general point here is, I would say be comprehensive and focus on quality and precision and excellence. Mean to be able to plead. A good case that talks about market wide wrongdoing or society wide wrongdoing to be able to get past a motion to dismiss, to be able to assemble the evidence that you need through, you know, discovery of emails and documents and testimony and expert reports and expert testimony to be able to prove that at trial or to be able to get a big settlement in lieu of trial. Everything has to be excellent. I mean, all of these cases involve the top firms in the country. Super complex topics. You know, the the experts that we used in the opioids cases, for example, they're all Ivy League professors. Everybody's smart. You know, the analysis is all super complicated. So you really have to be on on your on your toes. Um, you know, the more ambitious the cases get, the more complex and this area of, of data analysis is particularly complex. Um, you have to have some familiarity and familiarity with numbers. Um, so that's just a cautionary note that I interject here, which is that, um, you know, these cases can result in huge settlements, but there are so many pitfalls along the way and so many challenges. And you really that's why it's so essential to think ahead and, you know, be strategic early on in cases and, um, you know, go into it with, with eyes wide open about the challenges and, and really think a bit ahead about where the data is going to come from and how are you going to prove it. Um. You know, along the way. Also, even though lawyers are able to rely on consultants and experts, still the lawyers themselves have to have some understanding of the data and how it's analyzed. So even if you're not an economist or an epidemiologist yourself as a lawyer, you're still going to have to be comfortable reviewing a draft expert report on a, you know, for example, a on about epidemiology. And you have to be able to to be comfortable reviewing a peer reviewed study about addiction rates, for example, or you have to be able to be comfortable talking with vendors about statistical analysis that you're putting into a complaint. So that's another sort of global point there. Word of caution that, you know, as lawyers you need to be comfortable working with these complex topics. So, for example, in the opioids cases, I had frequent calls with the experts as they were working on their reports and, you know, helping them to come up with their analysis and work under working through it on on my own end and trying to understand it myself at the same time. Um, the other global point here is that, you know, even when you're in the weeds of collecting data or analyzing it, try and keep one eye to the end result that you're looking for of how the data fits into the other evidence that you have and how it will help you prove it your case, and making sure that it will be admissible. Um, and finally, as I noted earlier, um, remember that a court may require you to disclose your data analysis in the process of discovery. So just be careful. Um. About work, product and privilege concerns. And, you know, think about the fact that. Your work with experts and vendors and consultants might have to be disclosed. Um. All right. We're getting to the end here at summary judgment in trial. The key point here is that when it comes to expert reports and summary judgment, and particularly at trial, particularly if you have a jury, it's critical to make the data understandable. Oftentimes you're working with huge data sets involving millions of data points. And really you need to find ways to aggregate that and simplify that so that even though you're describing wide data sets, you're doing it in a way that's understandable, that can, you know, they're easy shortcut ways of doing that, like using spreadsheets, using pie charts, line graphs that distills information in the process. She need to decide what data is important, what is not what, What can I leave out of the analysis? You know, generally in trials you're very constrained in terms of the amount of time that you have relative to the amount of information that you'd like to disclose. So that involves some hard judgments about what is the most relevant information that we really need to get in. Um, and then, you know, data can be dry, it can be complicated. People can be scared off by it. So, you know, it's important to make data understandable. It's also important to make it interesting. So you need to find ways to incorporate, uh, data into the case and to sort of strike the right balance between the qualitative and the quantitative evidence. And it probably goes without saying. But when you're talking about data, it's always important to emphasize the objectivity, the authoritativeness, the reliability of that data. You know, when you're when you're talking about, for example, in opioids cases, when you're talking about the number of overdoses, you know, there was a heavy emphasis on talking about the fact that this came from official government sources, that it was comprehensive. There was discussion about how the data is collected and why that's reliable, etcetera. All right. So last suggestion here in mediation and settlement. Um, the point here is that, you know, if you are going to reach agreement on numbers that defendants, that money that defendants are willing to pay, there needs to be some basis for that. That number isn't just pulled out of a hat. Oftentimes, if the case has gone far enough, the parties will have done expert analysis of damages and they can use that as the basis for what they what plaintiffs are seeking and what defendants feel like, um, plaintiffs are entitled to, if anything, damages. Um, the point is that to the extent both parties have done. But damages, analysis and data analysis in the course of the lawsuit that's being settled, whether that information has been disclosed to the other side or not. That information is often helpful in settlement to create some sort of. That might that. Might make the parties more likely to settle because they feel like there's a, quote, reference point for the parties settlement positions and, you know, where they can meet in the middle based on that versus pulling a number out of thin air in the process. Um. You know, within the context of settlement party should consider sharing analysis even if they wouldn't be sharing it in or. Of data and expert testimony, at least with with a mediator, if you have one, if not party to to help get that case settled. All right. So just a recap of this part two as well as part one of my presentation on data analysis and litigation. Um, key takeaways. Data is very powerful. It can help you plead and prove cases. It allows plaintiff's lawyers to scale up the scope of cases from individual to market wide or society wide. It is therefore effective in creating case value, where you see cases that settle for hundreds of millions, if not billions of dollars in value because they're encompassing such broad subject matter. Data is attractive because and effective because it can help make cases more objective and scientific. Even though data is is hotly debated and both sides will criticize the data and the data analysis. Nonetheless, it it it's based on hard evidence. Um, cautionary note that data has to be handled and presented with care. You have to collect it effectively. You have to use it effectively. Present it effectively. Um, particularly in these complex cases where there are so many pitfalls and, and so many issues that can come up. So the general guidance there is just to use reliable data, use good experts, use the best you can in these in these big expensive cases. Um, layer in fact witnesses and fact testimony and fact evidence to complement the data plan ahead of time and. You know, just really be. Be strategic and then ultimately explain it well when it comes to trial. All right. So thank you for joining me today. I appreciate your attention. This was part two of my class on data analysis and litigation. My name is David Burnett. I'm a senior counsel at Motley Rice. My contact information is here in case you need to reach me. Thank you very much.