Legal AI case prediction start-up, Gavelytics, has taken a different approach to rival US predictive platforms and focused only on California, a move that the company believes offers users a far more detailed and granular insight into the potential actions of local judges.
Gavelytics currently covers California courts in Los Angeles (pictured above) and Riverside Counties. Additional California counties will be added soon, the company says.
While the core principles of case analysis and prediction are in many ways the same as other systems, i.e. you use natural language processing (NLP) to fillet key data from litigation documents and then leverage this information into actionable and searchable insights for users, Gavelytics believes, probably quite rightly, that California is such a massive legal market that litigators would benefit from a dedicated, local approach.
While those in the US are well aware of how large California is, it can sometimes be overlooked that the state has a GDP larger than that of France or India, that it has around 40 million people and in 2015 saw 6.8 million court cases filed across its superior courts system that cover civil, criminal, family and probate.
These almost 7 million cases in just one US state in turn led in 2015 to 9,540 jury trials and 480,000 court trials. Or to put it another way, that’s more than 26 jury trials everyday, just in California. The state also has 190,000 active attorneys, which is more than the total solicitors in England & Wales, plus around 2,100 judges.
In which case, one can see the value in a super granular predictive system for just one jurisdiction like this. The company adds that unlike some other US case prediction systems that cover the whole country they really get into every bit of court action across the state, including providing judge behaviour analytics compared right down to a county average.
OK, all well and good. But, for those not familiar with case prediction systems, what does Gavelytics, founded by litigator Rick Merrill and techie Juan Carlos Moreno actually do? Put simply it studies what judges do in the court room so that litigators have a better chance of predicting their behaviour, which in turn should – it is hoped – lead to more successful outcomes for their clients.
As the old adage goes: ‘Knowledge is power.’ And this certainly may be the case when a litigator walks into a court room and stands before a judge, who like all other people has their own habits, tendencies and behavioural biases. Know how your judge will act and you may have an edge against the opposing counsel who may not be as well informed or able to predict how their calls for certain motions may be treated.
The company puts it this way: ‘Imagine if litigators could effectively anticipate how a judge will rule on a pivotal motion. Or make a data-supported decision to disqualify a judge at the outset of litigation.’
‘In the past, litigators would have had to rely on vague recollection or anecdotes from colleagues to gather even a shred of information on trial court judges,’ they add.
Now, for the parts of California it currently covers, you can:
- Learn how a judge tends to rule on over 100 different types of motions, as compared to the county average.
- Review each judge’s tendency to rule for plaintiffs or defendants in bench trials and in motion practice.
- Improve pitches to prospective clients: demonstrate your detailed knowledge of the judge at the outset of litigation
- Custom-tailor your litigation strategy to the judge: adjust your tactics based on anticipated judicial behavior
- Get rid of a ‘problem’ judge: i.e. a peremptory challenge against a judge whose past rulings indicate she will be unsympathetic to your client’s case.
All of this not only hopefully gives you an edge in court, but as the company says allows you to ‘manage your client’s expectations’, and that has to be helpful when a lawyer is besieged by the dreaded question all clients will ask: ‘How likely am I to win this thing?’
So, there it is. But, what does this and other systems, such as RAVEL, which is now part of LexisNexis, and Premonition, and others mean for the legal world?
Firstly, legal AI predictive systems are providing a level of analysis that most law firms simply would not have, as they would not have an objective grasp on the data to do so. Manual review of 1,000s of old cases would be too slow and costly. No client would pay for it.
Secondly, they can – in the right hands – give a lawyer an edge that could mean they have a better chance of winning their case, or if they know they are likely to lose to be able to at least prepare the client’s expectations based on factual evidence.
But, are these productivity and efficiency tools? In some ways, yes, as they cut down the time needed for research. However, the real value is perhaps that they deliver additional value, i.e. they create value that was not there before because without the NLP and data analysis these kinds of insights just would not be available. I.e. this is legal AI creating something new, expanding what lawyers can do and know. And that is a great use case for explaining why legal AI is such a valuable new technology.
If you’d like to see a short video about the platform please see below: