Ideas for Now: CALR Linkrater

Somebody mentioned brief banks on Twitter a few days ago, and that got me to thinking about the types of information that I use to steal from other partners’ and associates’ briefs when I practiced. It was mainly the procedural stuff. I would always look at the way they framed burdens and standards, and the cases they used. For any new attorney, you know how valuable a good brief bank can be. And for researching, good appellate briefs are excellent sources of information.

So during this thought, it dawned on me that a brief bank is something of a ratings system. When I was practicing, our firm used a DOS-based program called PC Docs to tag documents with all kinds of information, including the type of document (e.g., “motion for summary judgment,” “motion in limine”). This allowed you to search for documents based on the tags and build your own motions, briefs, etc. If I used someone’s previously briefed argument, I would double check all the cases and replace or supplement as needed. I remember specifically that one of our attorneys had constructed a particularly good summary judgment brief, and it became the standard for just about all MSJs filed after.

During this process of creating, using, and re-using briefs, each of us validates the decisional authority contained within them. It’s like saying “hey, I think these are the best cases to use for this topic.” The more briefs using those cases, the greater the validation, or popularity, among your colleagues.

Now, what if we could tap into that ratings system? If we could, I would call it something like “Linkrater.”

Under the Linkrater system, when you are researcing cases using your favorite CALR program, each case—including the one you are reading and every case cited within that case—would be ranked (say by a number or stars) in terms of popularity. Let me be clear about something though. Popularity obviously isn’t a reason to use a case, but it might influence me when looking at a long list of cases following a standard keyword search. I might choose to sort by popularity first, just to see if the most “popular” cases are on point.

But for popularity rankings to mean anything, the underlying process for ranking has to be able to do some magical things. First (and this is where the magic comes in), the system must know what cases are being cited and for what propositions of law. In other words, a very flat document like a brief must be tagged so all of its citational goodness can be utilized. I think something similar to Thomson West’s Key Number taxonomy would be helpful here (they use it for everything else, why not briefs?). Second, the system should give you the option of rankings by geographical location (maybe city, county, state). So if you office in one county, but are briefing an issue in another county, the rankings would reflect lawyers’ preferences in that other county. It might also be interesting to see what other hyperlocal distinctions could be made (e.g., by lawyer, firm, or judge). Finally, the system should let you sort data by popularity ranking.

Thomson West claims to have access to nearly 2 million briefs. Who knows how much information is on PACER (but hopefully Carl Malamud will free them all one day). And other companies, like JD Supra, are building their own collections. So there’s plenty of information out there. For me, the idea of Linkrater is simply this: beyond analysis, what else does a brief say about the law?

[Image Augie Schwer.]