How to improve from 170's?

brydenwright

I've been scoring in the mid 170s (174-177) for a while now and I don't really know how to make more progress. I've gotten -3 to -5, but it seems pretty random which section I do best at. I've gotten perfect scores on every section across different PTs, and at this point it seems kind of random which one I do best at and which one I am worst in. Any advice on how to make my performance as consistent as possible?
Averages:
LR: -.7
LG: -2.4
RC: -1.7

KevinLuminateLSAT

The best way to improve at your level in LR and RC is to teach the exam. Perhaps you could start offering tutoring or find study buddies whose weaknesses are your strengths and try to answer every point of confusion they ask you about. Through teaching the exam, you'll be forced to articulate your own system for approaching the questions and you'll start to notice areas where your understanding is shaky or seemingly inconsistent with your reasoning in other contexts. I think being forced to resolve those inconsistencies and smooth out the shaky areas should help address at least some of your mistakes in LR and RC.

Another thing that should help is regularly revisiting problems that you've gotten wrong or had trouble with and walking through the exact thought process that would get you to the correct answer efficiently. Think of it like "foolproofing" those problems. It doesn't matter whether you already reviewed the problem, think you understand it perfectly now, and remember the answer. The point of revisiting the problems is to try to make the kind of thinking required on that problem more natural and automatic. Oftentimes the reason you got something wrong is that you were interpreting something in a slightly different way than the LSAT wanted you to or were just not as precise as you needed to be in breaking the stimulus or answer choices down. Although you might understand the problem well upon review, that doesn't mean that the correct reading/thought process would come easily to you if you saw it again. Regularly revisiting problems you've gotten wrong/had trouble with can help reinforce the right reading and thought processes that once escaped you.

For games, assuming that you're already in the habit of re-doing any games that gave you any amount of trouble, I'd suggest a couple of things. First, make sure you re-do games in different ways. Can you try the game again with a handicap? If you hadn't made this inference, would your approach to the game still allow you to solve the game well without too much lost time? If you didn't split the game up front, how could you still confidently work through the game? And if you didn't split originally, could you solve the game with a split? The point of this is not to show that every way of solving the game can work, but it's to help you see what kinds of inferences or approaches are absolutely critical to your way of doing games and to explore what processes you need to add or emphasize in order to notice those inferences or approaches.

In addition, make sure to keep doing games by type. Sometimes people start studying LG just by doing LG sections. But it absolutely helps to have a rock solid system in place for all of the standard game types. Have you done all of the "pure sequencing" games in a row? All games involving "or, but not both" ordering rules or conditional ordering rules? What about all In/Out games involving subgroups? If you don't think you can reel off a complete lesson on each kind of game, general approaches, common inferences tested, then you can definitely benefit from focused practice. In fact, ideally you'd get to a point where you could write your own game that involves the rules, inferences, and tricks that are typically presented in that type.

Cant Get Right

Well you’ve got to get that LG average down. How to go about that depends on a lot of factors, but that’s the obvious standout.

Beyond that, you’ve got to accept a certain margin of error. You don’t have time to outright solve every question on the exam, so you’ll have to be comfortable with a certain amount of risk. If you’re 90% confident in your answer on a given question and it’ll take you a minute to confirm, it’s an unambiguous error to confirm. Yet, that leaves an expectation that you’ll miss similar questions 10% of the time. In review, you need to treat it as a total calamity when you miss it, of course, but under time, you’ve got to accept the risk.

There may be some higher variance strategies which could produce more 180’s but with high inconsistency and dramatic downswings. I always prioritize consistency, and I think that is the correct priority. But if you’re willing to risk the relatively terrible downswings, you might could produce more higher scores than you otherwise would when prioritizing consistency. Your average would almost certainly drop, but with a greater number of higher scores anchoring the top end. Doesn’t seem worth it to me, and not sure it would work at all, but improving on that range is borderline theoretical and that’s my theory.