Certainly not. Holmes said that "if you want to learn the law and nothing else, you must look at it as a bad man." Neither lawyers nor clients are in that position in the real world, outside the context of learning the law and nothing else. They are not, for the most part, amoralists. My clients care for many things far more than the law, for which, as what they believe to be a non-creative force, they hardly care at all. About the morality of what they do, on the other hand, they care greatly.

In The Path of the Law, Holmes was careful to point out that the “bad” man was not bad, that the law is not a system of positive or negative judgments reaching a single correct sum. Rather, an “if…then…” mode of thought corresponds to Cartesian geometry, that much more accurately describes a rational actor at any given moment. A typical not-so-bad man wants to know if he speeds, what the fine will be, so that he can compare the time saved to the cost. When he has been later caught, he wants to know if the facts can lead to a dismissal. When the car he drives is a multi-billion dollar conglomerate, the inquiry is correspondingly more complex and sound advice as precious.

Perhaps. But in advising the people who do in fact "drive" such organizations, I would say that few if any seem to me to think in this way. By and large they strike me as both far more averse to legal risks than this supposed monetary weighing of costs and benefits suggests, and far more determined to preserve their self-image as people acting morally in their role as businessmen. I don't know why you think Holmes thinks otherwise on the basis of his social experience among the ruling class. Or what frames your own implicit conclusion in supposed agreement.

Are you sure that identifying Holmes and Langdell as co-thinkers of some thought presented here is accurate? Perhaps they disagree about something important here as much as or more than they agree?

Some go to trial, the vast majority of which reaffirm the lines already drawn. Legal scholars study the decisions, and in turn devise rules exemplified by specific decisions. They look broadly at opinions when compiling casebooks, or compile unique sets for law review articles. Occasionally, the judiciary departs from precedent, either on their own whim or prompted by academia, and muddy up the lawyer’s crystal ball. The dust gradually solidifies into new formations, and the process repeats. In Holmes’ time, a decent ball could be contrived by studying a generation’s worth of cases in a given jurisdiction and drawing functions accordingly

I don't think Holmes thinks this. I don't even think Jerome Frank really thinks Holmes thinks this. I certainly don't think Holmes' language leads directly to this conclusion. He says that courts deciding "easy" cases wind up restating the common law rules over the course of time, so that familiar rules are expressed in the long sequence of reports in different ways. He does not therefore think that legislation fails to happen, or that legal change is largely the result of whims, or that judges are much inclined to listen to professors. On the contrary, he expresses quite a different and more purposive sense of the nature of legal change.

.

Of what is he the example? Do you really present a person who compulsively drives too fast as a model "rational actor"? Do we begin by assuming not only that his judgment is otherwise not impaired, but that the forms of unconscious mental activity that cause his self-destructive and recklessly anti-social behavior have no other effects on his behavior? These do not seem to me well-warranted assumptions, so the whole illustration strikes me as not so "simple" as you maintain.

Perhaps that's because baseball is shockingly rudimentary compared with, say, life?

Good lawyers know how to judge and jurisdiction shop, how to introduce certain types of evidence that have often been effective, when to settle. Most of the expertise is gained through experience. Lawyers, and the client for whom they work, make these calls based on very little quantifiable data: verdict amounts, estimated billable hours, etc. But what if it does turn out to matter that a client wore a white hat on the day he signed a contract? What if that fact could be plugged into a matrix along with a whole range of other variables, and the predicted outcomes adjusted accordingly? Though trials are much more difficult to dissect than baseball games, perhaps a new system parallel to sabermetrics and available to the mass market could revolutionize predictive lawyering as it is currently practiced.

Perhaps the point is not to confuse what Big Data analysis of behavior around the legal system can do with what it can't. Jerome Frank's arguments about the difficulty of using calculating machines to predict the outcomes of social processes in which humans find facts are still applicable. Efforts to use Big Data to predict jury verdicts will of course interest insurance companies even more than they will interest contingent-fee plaintiffs' counsel. One could expect, however, the effects in both Tax Court and DC Circuit agency review cases to be somewhat more modest.

This step didn't follow from the preceding ones.

As time progresses, modeling products will emerge to meet every legal service sector they are applicable to, in a manner scaling to the niche served.

Nor this to its predecessor. So in both cases more thought needed to be communicated, and perhaps devised.

The most difficult task involved is gathering and classifying the available data in efficient terms. For the speeder, enough tickets will have to be logged in to make the inferences at all valid. Either those who will benefit will do it in a crowd-sourced manner, or data miners will need to sit down with whatever public records are available and then charge for it. Obviously, the information necessary for multinational corporate giants to wage the most cost-effective battles of attrition against each other will be correspondingly vaster, and the marketable product equally more difficult to use. Even so, it will still be cheaper than current billing models.

This comes to the implicit and familiar but unjustifiable conclusion that the end of improved social cognition is the replacement of, or impoverishment of, lawyers. No basis is presented for this conclusion, which is most certainly not established by anything that has happened over the last thousand years in the English-speaking world. Maybe in addition to being a somewhat sprung conclusion, unrelated to what passed before, it's also wrong? Why not give yourself a chance to imagine a little more broadly how lawyers might integrate the revolution in social science with the skills they already master and deploy on their clients' behalf?

Before Blue Eyes

-- By JacksonAlberts - 06 Apr 2013

The Bad Man I, or Every Client

According to Oliver Wendell Holmes, all clients, indeed all capitalists, are bad men. For they simply want an answer to the question “if…then…” that maximizes their own return or minimizes the penalty exacted by society, whether it is called a tax or penalty. In The Path of the Law, Holmes was careful to point out that the “bad” man was not bad, that the law is not a system of positive or negative judgments reaching a single correct sum. Rather, an “if…then…” mode of thought corresponds to Cartesian geometry, that much more accurately describes a rational actor at any given moment. A typical not-so-bad man wants to know if he speeds, what the fine will be, so that he can compare the time saved to the cost. When he has been later caught, he wants to know if the facts can lead to a dismissal. When the car he drives is a multi-billion dollar conglomerate, the inquiry is correspondingly more complex and sound advice as precious.

The Old Crystal Ball

That bad man has access to the best prediction money can buy. His attorney is well versed in the “crystals” that are the law and the “mud” within which a litigator may wiggle. Following Holmes and Langdell, the study of law boils cases down into rules, which predicate the terms of the settlements by which the vast majority of conflicts are resolved. Some go to trial, the vast majority of which reaffirm the lines already drawn. Legal scholars study the decisions, and in turn devise rules exemplified by specific decisions. They look broadly at opinions when compiling casebooks, or compile unique sets for law review articles. Occasionally, the judiciary departs from precedent, either on their own whim or prompted by academia, and muddy up the lawyer’s crystal ball. The dust gradually solidifies into new formations, and the process repeats. In Holmes’ time, a decent ball could be contrived by studying a generation’s worth of cases in a given jurisdiction and drawing functions accordingly.

Coming Change

Since the turn of the last century, all sort of lenses have radically improved in ability, cost, and ubiquity. And though Shepherdizing has given way to Lexis searches, another revolution is due. The tools available at present are cumbersome enough for research; they have little to no predictive ability. For that, attorneys must rely on their experience and knowledge, and clients must pay accordingly. Change is in the air, however. Consultants now offer legal data analysis to in-house counsel. Law schools are beginning to offer courses in “Empirical Legal Studies.” But this is only the beginning. As the volume of data available increases exponentially year by year and the tools for analyzing it follows Moore’s law, the impact on the legal field will eventually catch up.

The Bad Man II, or the Example

Take that simplest bad man, the rational speeder. An experienced commuter, he knows by now which stretches of road he is likely to encounter police on. He also knows the cost of a ticket, the time value saved and its corresponding worth to him. All this information is held by his in-house legal knowledge. When he does get a ticket, he might turn to outside help to fight it with advanced legal tools, such as questioning the citing officer. Or perhaps he goes in to fight it himself, with no strategy other than the hope the officer will not come to court. Or he calculates that the odds of success and the time cost are not worth it, and simply pays the ticket.

The New Crystal Ball

Currently, even for very sophisticated clients, the legal analysis done is shockingly rudimentary, compared even with say professional baseball. Good lawyers know how to judge and jurisdiction shop, how to introduce certain types of evidence that have often been effective, when to settle. Most of the expertise is gained through experience. Lawyers, and the client for whom they work, make these calls based on very little quantifiable data: verdict amounts, estimated billable hours, etc. But what if it does turn out to matter that a client wore a white hat on the day he signed a contract? What if that fact could be plugged into a matrix along with a whole range of other variables, and the predicted outcomes adjusted accordingly? Though trials are much more difficult to dissect than baseball games, perhaps a new system parallel to sabermetrics and available to the mass market could revolutionize predictive lawyering as it is currently practiced.

The Bad Man III, or How to Most Efficiently Serve Tomorrow's Clients

For the ticketed speeder, the most relevant piece of information is not the size of the fine, but how likely it is that he will have to pay it. While the information may already exist in formal or informal form on how often officers in general show up for court, or even officers of a given department, much better can be done. In this simplest scenario, the client needs to know if a given officer will show up to a support a ticket on a given date. And if enough tickets were plugged in, trends would certainly begin to emerge. Eventually, a formula could be adduced that accurately predicted the cost of that given infraction, then packaged into an app or website. As time progresses, modeling products will emerge to meet every legal service sector they are applicable to, in a manner scaling to the niche served. The most difficult task involved is gathering and classifying the available data in efficient terms. For the speeder, enough tickets will have to be logged in to make the inferences at all valid. Either those who will benefit will do it in a crowd-sourced manner, or data miners will need to sit down with whatever public records are available and then charge for it. Obviously, the information necessary for multinational corporate giants to wage the most cost-effective battles of attrition against each other will be correspondingly vaster, and the marketable product equally more difficult to use. Even so, it will still be cheaper than current billing models.

• #Set ALLOWTOPICVIEW = TWikiAdminGroup, JacksonAlberts
• #Set DENYTOPICVIEW = TWikiGuest

Note: TWiki has strict formatting rules for preference declarations. Make sure you preserve the three spaces, asterisk, and extra space at the beginning of these lines. If you wish to give access to any other users simply add them to the comma separated ALLOWTOPICVIEW list.