Litigators with a Data Fetish: Moneyballing Testimony

I am obsessed with depositions and, in particular, their importance to the outcome of cases. That obsession extends to deposition data as well as to the psychology of the players involved: the attorneys taking and defending depositions; the witnesses being deposed; as well as the court reporters. I’m fascinated by these people and their often-predictable behaviors. I’m fascinated because each of them influences—sometimes subtly, often radically—what occurs during a deposition and therefore, inevitably, the outcome of any case.

If you’re involved in litigation, and if you care about outcomes, then the importance of testimony is difficult to overstate. Facts matter in a case—of course they do—but it is the testimony that interprets those facts and that puts them into context that wins and loses cases. Documents and dry facts put juries to sleep – literally (Google it). But the drama of sworn testimony, surprise admissions and denials, well-crafted questions and impactful answers – good, bad, or ugly – move juries and judges and attorneys and their clients and everyone else in ways that few people really understand.

My team studies depositions relentlessly, measuring everything that occurs and doesn’t occur during these most impactful hours in any case. Among many other things, we can extract and quantify testimony from a single deposition, creating a “Cliff’s Notes” version of its contents, or from a thousand using a variety of tools (e.g., NLP). Once you extract that data and organize it into a borderline crystalline density, you can then mine that testimony for useable data and start to make predictions. And we do.

We monitor and extract the topics addressed and quantify the amount of time spent discussing them (fascinating). Using tools such as AI and Machine Learning, among others, we can extract question and answer pairs, analyzing the questions asked, the answers provided, the objections made and not made, documents used and re-used and not used at all, and the types of witnesses those documents are put in front of.

We look at the lengths of questions and answers, who speaks fast and who speaks slowly and when and during what topics. We look at the propensity of an attorney to talk over another attorney or over a witness, and if their behavior differs if the other attorney/witness is a man or a woman or younger or older or how long they’ve been practicing.

We see whether attorneys regularly and strategically talk over a witness mid-sentence to garble testimony they don’t like, trying to deprive the opposing party of a clean quote to use in motion practice or at trial.

We examine the frequency of breaks and their duration and who calls them and why. We scrutinize aggregate data revealing the habits of the people over time and across disparate matters. Yes, attorneys, I’m afraid you’re just as predictable as everyone else, often more so.

We’re even looking at data related to the emotional state of parties – witnesses and attorneys – over the arc of a deposition, including relative to when topics are discussed.

And we’re looking at correlating the presence of people – attorneys and witnesses – to specific outcomes: trial results (occasionally) but also settlement outcomes, because the presence and absence of specific people on your cases has measurable impacts, even if you don’t know precisely why. In short: we’re measuring everything having to do with depositions and testimony. You should see our gorgeous graphs!

Why on earth would we do this? Seems like a lot of trouble. But the answer is simple: knowledge is power. The possession of actionable data in advance of key events can often provide an overwhelming advantage, especially in highly competitive environments. If possessing asymmetric or advance data wasn’t an advantage, insider trading would be perfectly legal. So would industrial espionage. NDAs would be worthless.

We all recognize the power of advance knowledge in non-legal scenarios. It explains why retailers don’t send their competitors drafts of their Black Friday advertising campaigns. It’s why Apple doesn’t provide details about new technology before launch events. And it’s why catchers in Major League Baseball silently signal the pitches they want thrown, instead of telling the batter what's coming next: 94 mph fast ball, upper right third of the strike zone. After all, when batters know exactly what’s coming, they have all the data they need to knock it out of the park. And here’s my point: the same thing is possible in the arena of testimony. Among many other things, deposition data allows you to anticipate an opponent's style and pace, and type of objections. It helps you understand your own and your own witnesses’. It can tell you an opponent’s likely reaction to objections or interruptions, all of which have real impacts on a witnesses' preparation (and performance), and which can inform a lawyers' strategic preparation for their own role(s) in a deposition. With data, you can start to predict the contours of a deposition before it happens.

How about a concrete example? Consider: if a deposition or trial witness knows what’s coming in advance, they can deliver a perfect answer they practiced to a question they knew to expect. They can cement into the record an answer they can use as a fulcrum to move the case: in motion practice (e.g., summary judgment); during mediation (lower or raise bargaining position); or at trial (to flat out win).

It works fiendishly well on offense, too. Aggregated data on an opponent’s past testimony provides outrageous advantages over that opponent’s contemporary witnesses. That’s especially the case where that opponent’s own attorneys don’t have possession of—let alone insight into—that same past testimony. It allows you to lay verbal landmines, inviting a witness to contradict themselves or sworn testimony from other aligned witnesses from the current case or from the past.

When I explain what we’re doing (“Moneyballing” testimony), many people don’t get it. At least not at first. Why should a lawyer care what an opposing lawyer said or did in a different case not involving their current client? At first blush, it makes no sense. It seems counterintuitive and has never been done before. However, when we explain what we’re doing and why to seasoned litigators, it’s really cool to watch as they mentally unpack the value proposition and its many implications. Often their eyes get wide. One in-house lawyer I recently spoke with whistled low, then accused us of practicing witchcraft. I took it as a compliment. But no, we’re not practicing witchcraft. We’re just litigation professionals saddled with a data fetish. Testimony quants, if you like. Silently obsessing about the power of data and testimony and how it can help you win tight contests you might otherwise lose, or prevail in cases you should win, only earlier and on better terms.

If you’re someone that gets this, and you want to learn more, drop me a line at We're a startup, but we’re ready to scale and we're looking for new partners and friends with an interest in winning cases and helping us pioneer a completely new field of LegalTech: Testimony Intelligence.

203 views0 comments