1. Cracking the Code in the Eastern District of Texas: A Case Study in Scientific Litigation Practice George Speckart, Ph.D.

12. Copyright 2014 Litigation Psychology, LLC. All Rights Reserved.

11. 11 About the Author Dr. Speckart received his Ph.D. in Psychology from UCLA in 1984 with a specialization in personality measurement. He has been active in the jury consulting field since 1983, and has conducted over 600 mock trials and focus groups in pre-trial research for numerous types of litigation. Dr. Speckart has worked with litigators in over 150 jury selections, beginning with Dalkon Shields cases in 1983, the Agent Orange litigation in 1984, and Exxon Valdez litigation in 1994. His area of emphasis has shifted to patent litigation over the past decade as a result of increased demand for assistance in this complex area of jury psychology.

4. 4 have been used without a description of the precise procedures that constitute these approaches, and so the present section illuminates the actual processes involved in bringing about the results observed in the East Texas litigation outcomes. Briefly, these processes are partitioned into three methodological typologies, reflected in the three types of published articles mentioned in the previous section: 1) trial simulation research; 2) juror profiling; and 3) substantive knowledge of how the specific types of cases under study (in this instance, patent cases) are construed by jurors. These three domains interact with each other, for example, as one initially tackles the problem in a given venue, initial exploratory research provides insights into the third domain (substantive knowledge of how jurors problem-solve these particular cases, whatever they may be). Once this knowledge is obtained, it is used to guide research design and implementation (the first domain), making it progressively more predictive. As more research is implemented, methodological shortcomings become eliminated, and greater knowledge of juror profiling (the second domain) accrues simultaneously. In short, it is a process of extended diligence and evolution, based on multiple research exercises, with interactive and synergistic benefits realized over time. Thus, in the case of the East Texas problem, initial research leading to substantive knowledge (domain number three) actually occurred in other patent cases in other states first and was later found to also hold for East Texas. Predictive capabilities of the research were subsequently maximized once it was determined how trials in East Texas were conducted – for example, beginning with an initial tour de force performance on the witness stand by the plaintiff inventor. Precisely matching the venue characteristics of recruited panels for the research against those actually seated in the courtroom trials also required multiple projects to achieve optimal results. Actual simulation of the courtroom environment in such research was yet another factor leading to predictive validity. Finally, recall the use of the phrase “methodological shortcomings become eliminated,” a substantial portion of this process quite simply involves research skill, protracted sweat, vigilance and stamina in weeding out procedural errors, preparation short cuts, imbalance in demonstrative exhibits, and other nagging “bugs” that undermine predictive validity. Getting all of the moving parts “right” to the point that research actually wields predictive power is a labor of devotion, skill, and resources that is not a casual undertaking by any means. Additional factors also came into play that will be discussed momentarily, but first, the three components or domains under consideration are examined in closer detail. Trial Simulation Research. One particularly important aspect of patent litigation is the fact that, since these are invariably federal court cases, and since federal court judges typically do not allow jurors to be interviewed, little – if anything -- can be known about how verdict and damages decisions are made absent valid trial simulation (i.e., “mock trial” research). In short, without valid research, we do not know what jurors are doing or why, insofar as verdict and damages decisions are concerned. If we are relying on the research to tell us what jurors are thinking, how do we know that the research is

9. 9 here is that jurors substitute “accessible” personal interest stories even more readily in patent cases because they do not understand the technology, and so successful theme development often entails “backward engineering” the verdict form so that jurors will answer key interrogatories in a manner dictated by the “accessible stories” that can be woven from the fact scenario in a tactically beneficial manner. Development of these stories, in turn, is a function of learning from jurors what resonates with their perspectives, on the basis of trial simulation research conducted in advance. Conclusions Our biggest lesson has been that invalidity issues tend to center around these “accessible personal interest stories” more so than do infringement issues, and that jurors typically process invalidity as a dispositive case fulcrum point before infringement. If a patent case is really about misappropriation, as jurors tend to figure, then what is the worth or value of what was stolen? In short, what is the “validity” of this patent? Psychologically, our research in cumulative fashion has revealed that patent cases are typically tried “backward” and that invalidity should be the first issue addressed, not the last. This is consistent with the finding that virtually every defense verdict in East Texas has occurred with invalidation of the patents in suit, not by upholding validity and finding non-infringement. All of this progress in the last decade points to a lot of work, and indeed, when asked for strategic recommendations emanating from the research, the most important admonition is that these cases are not won through clever deductions or analytical reasoning. Instead they are won by hard work, investment of time, energy and money, and most of all, out-preparing the other side. There is really nothing mysterious about the action plans described presently – it is only a matter of systematic application of known methods in psychological research designed to predict human behavior. It only appears mysterious because very few trial teams have the time or inclination to do the work (or alternatively, in this day and age, clients do not want to spend the money, even though the economic benefits are many hundreds of times the initial investments). If “known methods of psychological research designed to predict human behavior” sounds mysterious, this should not be too much to ask of a jury research practitioner. Unfortunately, very few decision makers actually do ask for these types of qualifications. What has happened in East Texas is something that should have happened in each and every one of the problematic trial venues in this country.

7. 7 (theft), plaintiff jurors – those who are more cynical – were consequently more apt to view the defendant as having stolen the plaintiff’s intellectual property. Thus, plaintiff jurors were more likely to hold the belief that technology had been “stolen” because they were more cynical, and ultimately more likely to award high damages precisely because of such beliefs. In one of the most important cases in which we were involved, four defendants were being sued and the cynicism items were included in our proposed Supplemental Juror Questionnaire (“SJQ”) submitted to the court. Not knowing their significance, the plaintiff “passed” on the SJQ and it went to the prospective jurors. Striking strictly on the basis of answers to the cynicism questions, a disproportionate number of minorities were removed from the panel by defendants. Plaintiffs immediately issued a Batson challenge, whereupon it was explained to the court that only the race- neutral SJQ items measuring cynicism were used as a basis for the strike decisions. The Batson challenge was summarily denied, and the jury – a jury that the defense wanted – awarded $3.5 million across four defendants, or $875,000 each (plaintiff’s request was approximately $70 million). Continued use of the items is apparently starting to arouse the suspicions of some local plaintiff counsel, who see us in court in East Texas on multiple occasions. In the most recent one, plaintiff counsel stopped us after jury selection in which similar results were obtained and sneered, “You are so predictable.” At least the results were, as the jury in that case awarded only $1.5 million against our client, after plaintiff had requested over $60 million from the jury. It is not intended to infer that these results were a sole function of the psychological measurement technology implemented in jury selection. As stated previously, damages suppression is the result of a convergence of factors, ranging from attorney preparation; witness training; state-of-the-art visual exhibits; theme development; and other tactical measures. Such measures are obviously instrumental in accompanying the research endeavors described presently to create the overall benefits of reducing 8-,9- and 10-figure verdicts into damages figures that are 1-3% percent of those amounts. Incidentally, cynicism is not the only operative construct that requires close attention in jury selection in these cases. Various indicia of information processing styles – how jurors assimilate data, store (or fail to store) complex evidence, and similar attributes are also strongly predictive. Thus, for example, individuals whose work entails detailed information or measurements – engineers, programmers, technicians, etc. – are less likely to find infringement, since for them, minor differences tend to be significant. Psychologists use a construct entitled “information need” to describe the amount of data, information or evidence that a person characteristically requires in order to make a decision. This construct has been labeled “cognitive sophistication” by other psychologists, but from the standpoint of strategic implementation, it can most reliably be assessed and evaluated through the systematic construction of the SJQ. 4 Substantive Knowledge of the Cases. As described particularly in article “Navigating the

5. 5 sufficiently valid to be truly reliable? One way to know if the research is valid is if it predicts. Under the general paradigms that are accepted as reflecting true scientific progress, one first obtains the ability to predict, and from the ability to predict comes the ability to control. Thus, if we are able to predict, then the ability to control is right around the corner. Consequently, documentation of the predictive validity of the research was the first step to establishing the reliability of its conclusions. 1,2 As discussed in these articles, many litigators are skeptical about the ability of jury research to accurately predict, but at the same time, they are procuring services in a field with absolutely no barriers to entry and no qualifications or credentialing standards. The American Society of Trial Consultants recently held a vote on whether its members should be required to hold any type of credentials, and the proposition was voted down. While accurate prediction of trial outcomes is still not generally achievable in all cases, the observation of cases in which accurate prediction did occur in the present efforts (a substantial proportion, exceeding 80%) was a signpost suggesting that the associated findings could indeed tell us what the key operative factors were in precipitating damage awards. Another important signpost was the ability to take findings suggested by one project and determine that the inclusion of additional evidence or arguments suggested by such findings could indeed create the desired results in subsequent research for the same case. Thus, the litigators in Forgent v. Echostar mock tried their case four times (after “losing” the first three mock trials) before finding the “holy grail” that would ensure a defense verdict. Of all of the cases across which the $1.4 million average damages (cited previously) were computed, this case was the only one that was mock tried four times – and one of the few for which an outright defense verdict was obtained. Coincidence? Hardly. Four of the five defendants settled out of this case for $28 million and only Echostar held out, relying on the research, to obtain the defense verdict at trial. To give another example, in one of the cases, three mock juries in the highest damage case involved awarded $2 million, $2 million and $12 million – an average of $5.3 million, and the actual jury verdict was $5.4 million. Compare this to a situation where a project produces three mock juries that each render a defense verdict but the real trial produces substantial damages (a common problem when using “budget” jury research providers). Which project can be counted on to reveal subtle nuances of juror problem-solving patterns that will illuminate a trial strategy that actually works in suppressing damages? Which research findings are sufficiently reliable to use as a basis for re-formulating a defense that has the desired impact on the courtroom floor? From this vantage point, the relationship between prediction and control becomes self-evident. Quite simply, as the accuracy of the research continued to improve, the ability to use the associated results to control trial outcomes continued to become more and more effective. Since Forgent v. Echostar in 2007, defense verdicts in East Texas have been coming faster and faster, with five overall and two this year (2013). In East

6. 6 Texas patent matters, the process is most often related to identifying systematic patterns of miscomprehension. Since very little of the entire case fact scenario is accurately assimilated by jurors, the challenge is to identify those areas of miscomprehension that harm the defense and find ways to remediate those, and conversely to also identify those areas of miscomprehension that help the defense and enhance those. This approach is particularly relevant for “patent troll” litigation in which junk patents are being asserted, and for which the case only holds “merit” from the standpoint of a lay audience who does not fully comprehend the entire case. Juror Profile Advancements Getting the themes and arguments right, of course, is only part of the story, and a cogent argument can be made that trial outcomes are as much a factor of who hears the case as they are of what is presented. In Forgent v. Echostar , the attorneys were concerned about prospective jurors’ knowledge and experiences connected with DVRs (Digital Video Recorders), satellite dishes, and similar technology, being convinced that hands-on use of such devices would regulate jurors’ infringement and invalidity opinions, since the case at hand involved related technology. Subjecting the litigators’ hypotheses to empirical test, correlational analyses were performed that examined the extent to which such knowledge and experiences actually influenced subsequent verdict outcomes. The results revealed no statistical relationship – in other words, knowledge and experiences connected with this technology did not affect verdict outcomes in the four mock trials (the analyses were conducted on a combined sample size of 96 respondents who participated in the four projects). Instead, the items that were predictive – the measured variables that did regulate verdict outcomes – were variations of the same ones that had been identified seven years earlier in a paper describing the “Universal Plaintiff Juror”. 3 A more refined identification of the key measurements that were strongly related to verdict preference was eventually accomplished through correlational analyses conducted in replication across successive research projects. Once such replication was achieved, the identified measurement items were considered “must have” in any Supplemental Juror Questionnaire submitted for a defendant (since the items reliably identified high-damage plaintiff jurors to serve them up for strike by defendant. Working for a plaintiff, these items would be the ones that the trial team would most emphatically not want, since they would identify and expose the best jurors for their own side in that case, thus making them vulnerable to strike by opposing counsel). These predictive items measured a stable temperament characteristic that psychologists refer to as cynicism – the tendency to view the world as sinister, malevolent, or predatory in nature, particularly with regard to those in a position of power within corporate management (this is distinguished from an oft-confused concept known as skepticism, which is a tendency to doubt a proposition without sufficient proof. Skeptical jurors are good jurors for defendants!). As research had revealed that plaintiff jurors typically process patent cases in terms of the question of misappropriation

8. 8 Road to an Intellectual Property Verdict – Planning, Execution, and Things that Go Bump in the Night” 5 , jurors’ interpretation of patent cases involves perceptual anomalies and various cognitive foibles that appear to be unique to this litigation, as well as thought processes on key decisions (e.g., infringement, invalidity) that do not seem to “map” well into other types of litigation (e.g., product liability, contract cases). After working in patent cases across the country, it appears that these perceptual idiosyncrasies are more characteristic of the type of case (intellectual property and particularly patent cases) then they are of the venue (in this case East Texas), and as mentioned earlier, most of the observations documented in these articles “Navigating the Road to an Intellectual Property Verdict – Planning, Execution, and Things that Go Bump in the Night” 5 and “ “The Jury Trumps Everything,” Inside the Minds: Litigation Strategies for Intellectual Property Cases” 6 were first observed in patent litigation in California and elsewhere, and only later found to be applicable to East Texas as well. In short, the venue – East Texas – was not found to be as “unique” as we thought in terms of the perceptual distortions that jurors make. What is unique about East Texas is that jurors have a stronger presumption of validity, and that this presumption – combined with the supposition that, if we are in court, there must be a good reason (i.e., the defendant is probably “guilty”) – appears to be linked to the proclivity to award high damages. Additionally, their resistance to abstract reasoning causes jurors to deflect the notion of a hypothetical negotiation, so that they tend to side with running royalties that create larger numbers. At least, such was the case in the days when “the train went all the way to the end of the line.” Recent advancements include more powerful attacks on the presumption of validity (starting with aggressive voir dire queries, e.g. “Would you be surprised to learn that the Patent Office would have to be set up to operate differently if not for your role in checking these patents?”); more imaginative means to induce jurors to consider what a “hypothetical negotiation” actually entails; and more effective strategies for taking on infringement and invalidity arguments generally (e.g., increased emphasis on “Eureka!” or “independent development” stories, painstaking witness training with different, earlier inventors associated with defendants; etc.). One aspect of patent cases that is found to be systemic in jury psychology issues generally is the notion that jurors do not deliberate based on what happens in the courtroom – instead, they deliberate based on what they store, retain, and retrieve later from their memories. Thus, the transfiguration of evidence that occurs during the process of storing and later retrieving information from memory plays a vital role in shaping trial outcomes. The key technical features that infringement and invalidity experts place as central to their testimonies often do not emerge as factors in deliberations at all. Instead, the arcane nature of most of the key issues in patent litigation dictates that jurors will substitute other more “accessible” issues – personal relationships, licensing agreements, market performance, individuals’ own stories – for the more technical arguments that are fundamental to the actual case issues. But the question arises, “Is that so much different from what happens in complex business litigation, antitrust, or even toxic cases?” The answer

2. 2 I N 2001, DURING A MEETING WITH EAST Texas trial counsel, one of the more esteemed local litigators remarked that, with respect to juries awarding damages in the District’s patent cases, “Once the train leaves the station, it goes all the way to the end of the line.” The appearance of 8-, 9- and even 10-figure verdicts in patent cases in this notorious venue has generated a tremendous amount of interest, controversy, research, and even legislation in the last decade or so. Various approaches by defendants in dealing with the problem have been formulated to circumvent the obvious dangers in this venue, ranging from attempts to limit claims through federal legislation, to “venue- shifting” as a means to transfer out of the Eastern District altogether. The present discussion explores, instead, meeting the threat head on, by finding means to obtain defense verdicts or limit damages through the use of “psychological technology” – that is, the application of psychological research methodology to solve a real world problem. Most importantly, this discourse is not about a venue – it is about a method . In addition, it is not oriented toward a problem, but rather toward a solution . Application of the method to be described has resulted in average damages for the last fourteen jury verdicts of $1.4 million – about 2% of the damages amounts requested – working for defendants in East Texas patent cases throughout the 2006-2013 time frame (including five defense verdicts). So quite obviously, the train does not go to the end of the line in East Texas any more. It should be noted that, from the standpoint of some litigators who work in East Texas, damages have been reduced as a result of other factors as well, such as, for example, recent court rulings limiting the types of arguments that can be made by the damages experts. From this standpoint, the magnitudes of the damages awarded are a function of the arguments made by such experts. From the vantage point offered presently, damages are a function of the perceived merits of the case . While this is not an either/or question – both perspectives have merit and indeed overlap substantively – we will be approaching the issue based on a psychological propensity to award damages that is a function of jurors’ motivations and reactions to the entire case. Even from the perspective of these damages experts, from our understanding, their testimonies are contingent upon various substantive aspects of the entire case fact scenario (e.g., the actual content of invalidity and infringement arguments). Consequently these experts have been telling their clients that they need to be brought into these cases earlier so that their testimonies can incorporate these substantive aspects of the case that comprise the entire “story.” In essence these experts are taking a position consistent with that being offered presently, namely, that damages are a function of the entire case that is presented (by either side), and it is this presentation of the “entire case” that is primarily affected by the methods that are the subject of the present discussion. Moreover, this treatise is not really about “conquering East Texas.” It is about how the psychological technology works, and it works in a manner that is non-venue specific . In other words, the technology is not “tailored” to this venue, but rather does work, and has worked, generally in

3. 3 litigation environments throughout the country. The choice of East Texas is utilized presently because that is the forum for which clients were most recently willing to invest in the technology and support the chance to prove its ultimate usefulness, involving multiple replications, in the most compelling manner possible. Obviously, litigation research only occurs where it is funded, and the impetus to fund it in this venue has been considerable, although, as we shall consider momentarily, this is not the only such venue in which defendants have had an interest in damages suppression. We know that the implications of the present discussion extend well beyond East Texas because the usefulness of the technology at issue has been proven before, in other venues. As early as the 1980’s, in serial litigation involving automotive products, the defendant manufacturer’s litigation costs were cut by 30% as a result of implementation of early phases of the technology. In 2003, a major insurance carrier implemented an improved version of the same approach and came in $83 million under budget against its loss reserves in just one department. Finally, from 1995-2008, the largest heavy equipment manufacturer in the world, utilizing the same methodology, kept its damage awards to under $4.5 million over this 14-year period – with no punitive damages awarded in this time span. Once the recession arrived, the company decided to cut its litigation support costs, whereupon early in 2009 (after terminating the use of the presently discussed methodology) the company was promptly hit with a verdict for $56 million, including punitive damages. So we know the technology works, even disregarding the East Texas experience. But East Texas patent litigation has been a topic that has spawned a particularly keen interest within the field of litigation. Tackling the challenges there engaged the interest of numerous Fortune 100 companies as a high-visibility undertaking with significant economic ramifications. Ensuring success required a multi-disciplined effort drawing on various tactics, but again it is emphasized that the approach described presently involved an amalgamation of approaches that had developed and been proven elsewhere. For the interested reader, portions or “building blocks” of the technology have been discussed in other forums, including an explication of the predictive capabilities of the underlying trial simulation research in forecasting trial outcomes (i.e., prediction of jury awards) 1,2 . A parallel body of research 3,4 was developed concurrently identifying juror profile typologies – again, an issue of prediction, but at the individual level (in this case, prediction of individual juror behavior). Thirdly, an accumulating body of knowledge on how jurors problem-solve patent cases was also documented, 5,6 following a series of presentations at the UTCLE Advanced Patent Law Institute in 2007-2008. Then finally, a precursor to the present article was published in Law.com entitled “Taming Texas” 7 that discussed the use of the technology in connection with the defense verdict in Forgent v. Echostar in East Texas in 2007. How It Works Up till this point, terms such as “psychological technology,” “research methodology” and the like

10. 10 References 1. Speckart, G., “Trial by Science,” Risk & Insurance , 2008, 19, no. 3. 2. Speckart, G., “Do Mock Trials Predict Actual Trial Outcomes?,” In House , 2010, Summer, 5 , no. 13. 3. Speckart, G., “Identifying the Plaintiff Juror,” For the Defense , 2000, 42 , no. 9. 4. Speckart, G. and McLennan, L., “How to Tap the Potential of the Juror Questionnaire,” ALI-ABA’s Manual on Trial Advocacy , 2001, American Law Institute -- American Bar Association on Continuing Legal Education, IBN # 0-8318- 1405-5. 5. Speckart, G., “Navigating the Road to an Intellectual Property Verdict – Planning, Execution, and Things that Go Bump in the Night,” For the Defense , 2008, 50 , no. 4. 6. Gilbert, C. and Speckart, G., “The Jury Trumps Everything,” Inside the Minds: Litigation Strategies for Intellectual Property Cases , 2010, Aspatore Publishing, Boston MA. 7. Raymond, N., “Taming Texas,” Law.com , March 2008, http://www.law.com/jsp/ PubArticle.jsp?id=900005504899&slretu rn=20130922173948