Anti-supported. The major problem is that, no matter how many cases the system stores, they’re not able to reason out how one case is relevantly similar to another. They just check for cases that match up with fixed criteria that are preset by the designer. Adding more cases wouldn’t solve the underlying issue that these systems are really just a matching program, not a reasoning program.

Unsupported. Case-based systems might be more successful (at least in the eyes of their proponents), but nothing suggests that they’re based on a “simpler view” of reasoning. It’s just a different method. Rule systems reason by rule-application using explicit legal rules, while case-based systems reason by analogy using precedent. We have no reason to think that reasoning by analogy is simpler than rule-application.

Strongly supported. These systems already have specific criteria for similarity among cases. The problem is that these criteria, regardless of what they are, are always preset by the designer, which essentially makes these an inflexible matching program. They can’t actually reason out how a certain precedent may or may not apply in unique or unexpected ways.

Anti-supported. The author states that no legal reasoning system can independently provide expert advice.

Unsupported. The way these systems try to reach decisions (reasoning by analogy instead of by rule-application) might be considered more ambitious (though that’s unclear). But their goal is the same: to “help to resolve legal disputes by reasoning from and applying the law.” The author never suggests that the purpose of case-based systems is any different from that of rule systems.