Let’s talk about gaming the legal system.
There is a product called Ravel that got me thinking about this issue, and I have been sitting on it for a little while.
These guys specifically work at (they claim) making law more accessible to everyone, by streamlining case decisions, and making it clear what comes out of cases amongst other things. In addition, they seem to have a pretty solid grasp on the ideas of mapping out data in a fairly accessible format – for example, their chart for determining at what levels of court a specific issue has come into dispute is really interesting, as you can see here.
So I think that, at a very basic level, Ravel might be working to improve access to justice for the average person. They have a free version of their subscription plan that you can sign up for, and it includes a 7 day trial for their more premium levels, which have greater case analytics, and more powerful tools. I think that as a free tool, Ravel is pretty cool.
Now let’s talk about their monetisation a little bit, and the problems that I have with their model.
Specifically I take issue with one feature in their program, called judge analytics. The idea with this tool, as you might guess, is to analyze the judge that will be passing judgement on your case. It covers everything from the past judgements they have made, their reasons, what they have cited in the past, the specific language they use, the list goes on and on. Frankly, the only thing I think isn’t on there is probably what the judge will eat for breakfast that morning. But don’t worry, I’m sure that’s coming in 2.0.
My issue with this kind of analysis is that the matter of law before the judge is quickly diluted, and becomes substantially more about playing the judge, than arguing the law. Take a game of poker. In a game of poker, you can play the cards (the “law”) and make decisions based on the hand you’re dealt (the client that hires you and the facts you have). Or, you can choose to play the other player. The analogy falls apart a bit here because the other player should really be the other lawyer, but let’s ignore that for a moment and pretend the other ‘player’ is the judge. When you know everything about the ‘other player’ then that means that you don’t really care much what your cards are. If you say the right word at the right time and cite their favorite authority, maybe your facts don’t matter. The same way that playing a hand of poker perfectly can involve never taking in your cards.
I think that when we get to the point where the judge’s every move can be foreseen and predicted by technology like this it threatens the rule of law. The idea is that judge’s are intended to be these arbiters of justice, above the rest of society, yet influenced by it; removed from it to make their decisions in an unbiased manner, but conscious of the biases that they have inherently. When you lay out these biases plain to see, it removes a chunk of what makes judges and the law special. It becomes more about knowledge, and less about what the law really is. It doesn’t matter how good a lawyer you are. It doesn’t matter how innocent your client is of the accusation. If the other lawyer knows how to charm the judge just right, the guilt or innocence of your client doesn’t matter.
That being said, clearly there are avenues of recourse for losing a case that should justly have been won. You can appeal, you can go for judicial review, you can rail and scream at the top of your lungs and cause a ruckus in the press. But at the end of the day, this kind of response merely bogs down the process of justice. Your client goes away unhappy. Your case goes unresolved.
Justice falls by the wayside.
As an aside, this is obviously an excellent tool for litigators, and has the potential to really transform the way that litigators act in the courtroom. In fact, Ravel even states something very similar on their marketing page.
Overall, I think that Ravel is likely a good thing. This kind of increased and simplified access to case law is a huge boon to society. I think that the drawbacks come from their monetization model, which I think will make them a lot of money, but in a way that I think has the potential to take away from real justice.
“Ars, Lex Iuvenesque Inventores: hinc Futurum”
Greg, thanks so much for a really great first student post of 2016. I hope it gets some discussion going!
As I was reading your righteous objections to gaming the system with judge analytics, I found myself thinking “you do realize litigators do this already, right?” Based on my personal anecdata from my firm days, a huge amount of time and energy are spent on figuring out what judge you might get and how to pitch things to appeal to that judge. There is a whole industry of litigation consultants who pitch strategies on things like what colour background to print visual aids on to elicit the response you want from a judge or jury. Not kidding.
And then I also remembered how stunned I was by all this when I encountered it. It was rather a long way away from what I thought law was all about. It was another one of those “huh, I wonder why no one tells you about this in law school?” moments.
Data analytics on judges do something in a more sophisticated and complete way that litigators have been doing based on their informed intuition since … probably Cicero’s day, or Socrates’ … the trial of Socrates might even be a cautionary example of what can happen if you go for pure argument over “gaming the system” or manipulating your audience.
Another important part of using analytics is more effective prediction of what the outcome will be – and of course who the judge is is a variable in that. The benefit there is that clients can make better informed decisions about whether it’s worth proceeding.
Maybe being able to do this with more powerful and accurate tools is a qualitative change, though. It’s one thing if lawyers can make an informed guess about litigation strategy based on the particular judge’s foibles. It’s quite another if an algorithm can figure it all out without any need for argument and eloquence, the quintessence of what lawyers do.
If that troubles you, just wait till you read and hear what Ben Alarie has to say about the “legal singularity…”
(Also, it may comfort you to know that, apparently, this stuff doesn’t work that well in Canada, at least not yet, because our data sets aren’t big enough: http://www.slaw.ca/2016/09/11/the-big-data-problem-for-ai-in-law/)
I’m not sure if you have read much John Grisham (which I know is definitely not wholly factually based haha!) but he goes straight to the topic of gaming the judge to an obscene amount, and ignoring your actual case. That’s one of the reasons I started thinking about this issue in the first place!
And yeah, it is a current issue in law, that people already game the system as much as they can. But I do think the qualitative change is problematic. When the information is at your fingertips in seconds, rather than through hours of experience and research, it fundamentally changes the way that we view the legal system.
I am EXTREMELY glad to hear that the data is just not there for Canada! In fact, I think we might have a more robust LAW system than the states, and that could be one of the reasons for it.
Thanks for your thoughts! I appreciate the comment!
Great topic Greg!
I am wondering if judges could use this data to possibly limit their potential biases by reviewing it regularly to see what their own tendencies are? Or possibly the administrators of the court system could review such information on a regular basis to see if there is anything that they should be doing about it?
I know it is not necessarily bullet proof information, but if these tendencies can be used by lawyers maybe the judiciary should be seriously looking into it. We discussed some similar issues and studies in Professor Jones’ Advanced Advocacy class last year. There were a number of studies that we looked at relating to the biases that impact decisions outside of the pure law of cases. It was very interesting and worrying, but I think having access to this information in up to date analytics could certainly help to remedy the problem. Hopefully the US judiciary has already jumped on this and purchased their subscription to Ravel to aid in limiting the use of this information in “gaming” the system.
I do think that it would also be great if, as Professor Sykes stated, these numbers could be used effectively to determine whether or not people move forward with their claims. Something like this could certainly make the legal system more accessible if litigation that is doomed to fail or at a minimum is not worth the time and effort, never even makes it to a lawyer’s office, never mind a court room.
Nathan
Ooooh Nathan that’s a cool idea to turn it around! I like that thought process.
Question I would be curious to hear your answer to though – do you think that, even if we are told our ingrained biases about certain viewpoints and such, are we able to change them? Better yet, do you think a judge could change them?
I seem to recall Prof. Jones mentioning there is some research on that – even when people were told that an experiment was measuring how they were affected by “heuristics,” the effects were still there. I think there might be discussion of this in Daniel Kahneman’s Thinking Fast and Slow. Anyone have more accurate / less vague information about this?
Great post Greg!
Just to let you know though, a number of litigation micro-finance companies already do include similar prediction stats in their algorithms. In other words they take it seriously enough to invest on it.
Notice, that even on WestlawNext there is a whole section devoted to litigation quantums, which is a whole another area of predictions on its own (*a similar concept though).
Tbh I also don’t think judicial predictability is a vice. If anything, it’s probably one of the a hallmarks of an advanced judicial system. Remember Fuller’s hypothetical King Rex scenario from LP? His citizens hated him for making things up adhoc! His rules lacked predictability and any sort of coherence.
I do realize that the example is rudimentary and the issue you have raised is far more complex, but I think we may both agree that predictably on its own, isn’t such a bad thing. I mean if someone is offering you a software that will help you make more accurate predictions, then why not?
I think that the key issue I have with it Anoosh is the potential of the judge predictability to take away from the raw law. I know (hopefully I have learned this over the last two years!) that the law is a series of grey areas, with judges simply trying to shrink the amount of grey over time, or at least make the shades clearer. My problem is the potential of future implications of the technology; the idea that theoretically a lawyer could walk in knowing what exactly they need to do to convince a judge based on the judges biases, not on the law itself.
That being said, I recognize your point about King Rex. But I think that judicial bias has more of a potential to create a King Rex scenario than pure law; when decisions are based exclusively on law, I think they become predictable, up until the point of novel issues, in which case we probably need some level of judicial bias and recognition of the fact that they are not robots.
Long post. Hope I touched on some things you find interesting!
Links (Litigation Finance):
-https://trialfunder.com/how-it-works (US)
-http://bridgepointfinancial.ca/# (CAN)
Great article on the topic:
-http://www.newyorker.com/business/currency/what-litigation-finance-is-really-about