It looks like you're new here. If you want to get involved, click one of these buttons!
If time is no issue, I get 80-100% of the questions correct. On an actual LSAT or a PT, I tend to run out of time with approx. 3-5 questions remaining, causing me to guess them.
When doing Problem Sets ("PS"): The first time I do a given PS, I do poorly. The second and thereafter, I get 80-100% of the questions correct - even if I don't come back to the same problems for a considerable amount of time. The problem is on an actual LSAT, I obviously won't have experienced the questions before.
What is the best method to improve my performance in LG at this point? To grind new PTs (which are a limited resource), or to re-do old Problem Sets?
Comments
Hello @forthewinwin
On every game that either you got a question wrong, went over time, did not get all the inferences JY got or you did not feel comfortable, you need to redo do that game over. I was in the same boat as you when I first started. I spend about month of straight games and sprinkling in LR and RC, I would keep track of all the games that gave me any sort of trouble and I would do those over. I would also work on specific types of games at a time, and once I was good a a specific type of game I would through some other types that I was not familiar with and then start working on the first type and the new second type. I did this until I made all the inferences JY did. Now over the last 5 timed exams I have a 0.8 average wrong for LG. It takes time, but once you do enough of the games all of the games start to look familiar just with different named games pieces and scenarios. Another thing that I did was break the stimuluses down and be able to pick out key words that help me know which game type I was dealing with so I knew the type of board to draw. Once I did this my mastery of games really took off. for example sequencing will have a lot of words like, before, after, next to, in front of, behind, ect., grouping games will have placed into, selected for, teams of. double layer sequencing will have, on group of say teachers and they will have to be matched with students, o papers to grade, so there are two sets of pieces that have to be matched, and sequenced at one time. I always take a few seconds to think of the moving pieces versus the non-moving pieces and their relationship, plus if it is a tricky game I'll peak at the acceptable situation question to get an idea of the game board. defiantly do not break new PTs unless its absolutely necessary. 1-16 and 1-35 is plenty. I can share with you the google spread sheet I made with all the CC games broken down by game type and by time it should be completed, however, there are other things on the doc as well, like LR notes, RC notes definitions of words I didn't know. But the main thing is I color coded the games I needed the most help with.
In the later PTs, the strategy is that for games that are 5 questions or less, you don't spend time to set up all sub gameboards. You let the questions with additional premises lead the way. You do these questions first. Doing so saved me a lot of time. I don't know if you have already done it or not. Just wanna share my observation.
I haven't tried that, as even looking over all the questions quickly detracted valuable time. Or having to later re-visit a question and re-familiarize myself with it.
I've been averaging -5 on the LG part, and I usually end up guessing 3-5 questions in each section due to lacking time. Thus the time constraint seems to be a major cause of my error rate.
But I shall try doing the additional premise questions first.
I did notice that I tend to waste the most time on the "X can be in any of the following...", and the first "Which one of the following is an acceptable arrangement..." questions that usually follows a new scenario.
In LR and RC, this also has been a problem for me.. Sometimes with odd terminology I misunderstand what a given response choice is inferring, or what a question is asking. e.g. Questions about an author's attitude, where the attitude is defined with an odd word I don't understand.
@forthewinwin getting familiar with what the question stems are asking comes with time. However in RC it is important more so than LR to fully read the stem and make sure you exactly what the stem wants. Also, the Core Curium should have flash cards to drill for LR stems. Picking up on small words and phrases for example "ingenious idea", or "cleverly explains" would be an indication of the author's tone. For those question in particular, where they give you like ACs of 1 word or short phrases to describe the tone, I cross off the modifiers of the words and just think of the core word, phrases like 'scholarly neutral', I would just pay attention to neutral. For words that I come across and don't know the definition, I use the context clues to the best of my ability and move on. Don't sink more time into a question you're likely to get wrong(or potentially get right but then leave 2+ questions to guessing), one of the biggest things that helped my score was skipping and coming back to questions.
The Logic game section is designed in a way that more often than not, if you do the additional premises first for the games that have 5-6 or fewer questions, you can use the gameboards set up for such questions to answer other questions. Also, your right and wrong answers of the additional premise question type can give you clues to quickly eliminate wrong answer choices in other questions. I think in recent tests the questions are more interdependent than the old ones.
Questions with additional premises I am referring to are the questions that say If you place a game piece in a slot, what must/could be true?
Setting up all possible sub gameboards gives us peace of mind, and the people who created the test know it, so they try to trick you to set up gameboards before diving into the questions when doing so is not necessary. If you fall for the trick, you will lose valuable time.
To see if setting up all sub gameboards is necessary, quickly scan the question stems after writing down all the rules. Out of the 5/6 questions, if the first question is asking possible complete setup and at least 2 of the questions are additional premise types, you most likely do just fine without setting all sub gameboards. They may have the last question as rule substitution.
That also has been a prime time waster for me. I very rarely get one of those questions incorrect, but they take considerable time since I have to re-draw boards.
I just tried this, but results have been mixed. For example, on PT41, I had to guess 6 questions. Though it may have to do with PT41 having the last LG puzzle where very few inferences could be made.
EDIT: In general, does this strategy work? Just looked at the PT41 and this game seems unsual.
I've had mixed results. Success seems to depend on the nature of the games. Not having sub boards drawn out, even if not very detailed or conclusive, has often caused me to take longer to solve the additional inference questions.
Hi, I am in the same boat (if not a worse boat) then you, as in, I am even slower on the LG. So take my advice with a grain of salt. However, I will tell you what is helping me drastically speed up over the last few problem sets.
That would be, the concept of "recognizing the action." As you are writing out the rules, the "action" would be the place where rules are chained together, where rules trigger outcomes, basically the flex points or swivel points, if you will, around which the pieces are swung around.
Recognizing the place in the game with the most action has helped me speed up a lot because usually the majority (if not all) of the questions will hone in on that one focus zone. If you really mentally understand the "action," a lot of the answers can be visualized, instead of drawn out, saving precious time.
The "Which one of the following is an acceptable arrangement..." should typically be your fastest question, not your slowest. You just run down the rules one at a time, and each rule usually knocks out one answer choice.
Sometimes they throw you curveballs on this one, as in, you need inferences to figure it out, but it's designed for the most part to be very straightforward.
Usually if I notice that if a given condition is linked with a bunch of others - and fills up a substantial part of the board - I quickly draw out a board with that scenario. Very often the additional premise question will refer to that scenario, or a scenario where that scenario is NOT true.
What really hits me hard is if I do NOT draw additional boards at the beginning, and then this causes me to take substantially longer on the additional premise questions. I have this problem the worst with grouping games with repeating items. I did tried PT89 for example, and the last game caused me to make 3 question guesses. By coincidence my score for the LG section was -3.
My LSAT write is next Tuesday, and I suffer immensely at LR and LC (averaging -10 in each), so it's imperative I perfect my performance in LG.
you aren't alone in your issue - under timed conditions I almost always have about 5 questions left ... I am struggling to get any faster.
What I'm going to try next is to use the Analytics feature, and to grind problems of the specific question types I'm getting errors on for LR and RC, or for the question types I spend the most time on per question anyway.