Help for someone who apparently sucks at LG? Forum

[Fiera]

Hi guys, I'm semi-freaking out here. I took the Powerscore course and my LR and RC are very solid, but I'm just too slow with the LG. It takes me too long to figure out all the inferences to start the game and I make too frequent, silly errors. Does anyone have any book recommendations or tips for the games--I really need a 170 and this section is killing me!

Right now I'm scoring around -(0-4) in total LR and -(2-3) in RC, and still improving. LG is all over the place and terrible, like -10.

How many tests are you doing, too, for those of you aiming or Oct?

Thanks guys!

[lilhugsy24]

Honestly if you've already done powerscore then I think the best thing for you to do is to just drill constantly. There really isn't any book that's just gonna automatically help you get to -0. LG is just more about constant practice so that so after a while you know exactly how to attack a game because you've already done so many like it. I know how it feels to be struggling in LG as that used to be my worst section (-11) and honestly I think the best thing to do is to just drill games.

Plus the more recent games nowadays seem more about making hypos for individual questions than finding that crucial inference like in the older games. That's just my take on things though.

Good luck with your prep and HTH.

Saying how much you miss in each section might help people give better advice as well so you may want to put that down in the OP.

[Fiera]

Thank you lilhugsy, I think drilling would help as well but I feel like I don't have enough material (as in real LSAT games) to just do them all day long and still have enough lsats left over to do 2 tests a week til Oct. Did you use any books or materials that weren't necessarily real LSAT questions that still helped you?

I've been scoring in the -(0-4) range in LR and - (2-3) in RC, and feel like I can improve even more. The LG is a disaster and all over the place, like (-10)ish.

[northwood]

do each game three times. divide about 12 tests LG section by game type, and go through them each 3 times. first just try to figure out the nuances of the game. second pass ( like the first, it should be untimed) try to answer the questions, and check for accuracy. Third time ( timed) answer each question and am for accuracy ( 100%). when you go back, explain the setup, the right answer choice, and the wrong answer choices. YOu should do each attempt at most once a day.( mayble a bit longer so you dont konw the right answer choices/ set up from memory). helped me tremendously.

[RCinDNA]

I do about 7 - 10 games, starting from easy ones (ranked 1) to the most difficult (ranked 4) in the Cambridge book. Then I listen to Manhattan LSAT videos for tips on the set up or go through the games again to figure them out myself. When I first started drilling the Games, for example, I would go through the entire section through rank 1 to rank 2 a few times to figure out how the inferences worked and the set up before attempting the harder ones.

Are you making sure to drill different types? Sometimes, I'll just do the game types I'm used to, so without meaning it, I will stop drilling the ones I am least skilled at doing (probably confidence).

[cloudhidden]

I'm going through every previously released game, and doing most of them three or more times. I never thought the day would come, but I went from once having a -16 (and that was after reading the LGB) to having my first perfect section. Logic games are messy and all about finding the quickest way from point A to B, not about how well you adhere to particular methods. Through trial and error, I learned how I could pick up time on any variation of any game type. But that's not to say to use trial and error on the questions, that approach just eats up time.

[Psib337]

Go over everything again. I took Testmasters before the June test and was getting -8 to -10 on LG almost every time and I was either not finishing sections or just barely finishing them. After the June test I was most concerned about LG and after I went through the LGB I realized there was just a lot of stuff I missed the first time around and since then games have been my favorite section. I also think maybe some things in the LGB were phrased in a way that just made more sense to me, Testmasters was great but sometimes you just need to hear the same thing explained in a slightly different way. Also, keep practicing, you'll start seeing the inferences. My diagrams used to just be a bunch of not rules and letters put in slots when a rule said, "R goes third." Once you do enough games some things will become obvious.

[bp shinners]

When people take too long on games, it's usually because they're not sticking to the script, and they're trying to A Beautiful Mind deductions/answers. Those of us who rock the games tend to have a set method that we ALWAYS apply, and it doesn't involve rolling our eyes back into our heads and doing crazy mental calculus.

When people make stupid mistakes on the games, it's usually because they're trying to remember rules to save themselves time, so they don't check their deductions against the rules. This is a mistake (like adding milk to your omelet to add density).

The LSAT, for the most part, isn't hard. It's simply a matter of doing the same, simple steps over and over again without making a terrible mistake. The difficulty (outside of a few hard questions) comes from not making those mistakes over the course of 100 questions, each with 5+ steps, when every step lends itself to error/tries to make you slip up.

So here's what I do for every single game I've ever done:

1: Read the introduction and make your setup

2: Read the rules and symbolize them using your preferred method

3: Recheck your symbolization of the rules

4: Identify any player that doesn't show up in the rules - this person can go anywhere, and is thus important to keep in your back pocket as a random

5: Make deductions - 'deduction' is a fancy word for 'combination of 2 or more rules'. There's nothing special about it - you're literally looking for the same letter/player/group to show up twice, and that's about it.
a) Check your first rule. See if it interacts with any other rule (either because they share a player, both talk about the same slot/group, both limit distributions)
b) Repeat step 5a for each other rule

6: Make scenarios
a) Check to see if any single player (or group of players) is so constrained that they're only able to fit in 2-3 different places. Make scenarios based on these placements, especially if they interact with any other rules. Bam, you now have 2-3 skeletons that define all the possibilities. This usually comes in the form of blocks for ordering games, an ordering chain with a 'focal point' of one player in ordering games, a Must Be Together rule in grouping games, or a player that shows up in 3 or more rules in any game (count each part of a multi-part rule for this purpose).

b) Check to see if any single slot is so constrained that there are only 1-2 different players that will go there. Make scenarios based on those players in that slot. Bam, you now have 1-2 skeletons that define all the possibilities. This usually comes from an option in an ordering game (if at least one of the option players shows up in other rules), or a group that is almost-but-not-completely filled out in a grouping game.

c) Check if there's anything else that limits the possibilities to 2-3. Make scenarios based on this. This is a catch-all based on those weird rules that sometimes show up (like mauve dinos). If you have a rule that you know is weird (because it's not one of the normal rules for that type of game), think about scenarios based around it.

7: Go to the questions
a) Elimination (which of the following could be a complete and accurate) - Don't look at your work. Read a rule, eliminate an answer or two. This is the fastest way, by far, to approach these problems.

b) Conditional ("If _______", or anything that gives you a new piece of information) - Draw a new diagram with the piece of information and any deductions you made in your setup. See if the new piece of info interacts with your first rule. Then, your second, and so on until you've gotten through all of the rules. If one of the rules lets you make a deduction, start over again with the first rule (but this time you can skip the rule you already used to make a deduction). If you have scenarios, see if the new piece of info limits you to 1-2 of them, and use them.

c) Absolute ("Which of the following must/could be true/false?") - If you made the deductions, you should be able to answer these without doing any work. Use your setups/scenarios to answer them. If you don't have the answer in your setup, you missed a deduction. If this is a "Which of the following Must Be True?" question, the answer will be the deduction, and you can add it to your setup. THIS IS THE ONLY TIME YOU SHOULD EVER ADD ANYTHING TO YOUR SETUP AFTER YOU FINISH IT AT THE BEGINNING OF THE GAME!

8: Do the Happy Dance.

Hope this helps!

[cloudhidden]

I agree about demystifying the section. Even dramatic improvements can result from simple execution. For instance, just knowing the random variables for could be true questions can save a lot of time that might otherwise get wasted on drawing up unnecessary hypotheticals for answer choices that are not likely correct. Despite my substantial improvements, I don't think I had any truly eureka moments, I just kept chipping away at it and saw how I could do some things more efficiently.

[Swimp]

Drilling is the way to go with LG. Make them routine. This will happen if you do them enough times and stick to the script. I never had any epiphanies while drilling LGs, but one day I looked up and realized that I'd been consistently getting almost all the questions right. I still feel like I'm grinding more in LG than in LR or RC, but I choose the right answers.

The only question type that still throws me off is the 'Which of the following rules has the same effect as rule X?" ones. Whether or not I get those generally just depends on whether I have enough time left over to diagram each answer choice.

[cloudhidden]

Yes, but I found that those can sometimes be suprisingly easy like when the rules say that two variables must be next to each other and that one of them must be before another one, well, if the question asks what rule substitution would have the same effect as that one variable being before the other variable we know that saying the variable that the first variable must be next to occurs before that third variable gives us the same result. Other times, though, these questions can be very difficult, but I don't just automatically skip over them anymore.

[Swimp]

Yes, but I found that those can sometimes be suprisingly easy like when the rules say that two variables must be next to each other and that one of them must be before another one, well, if the question asks what rule substitution would have the same effect as that one variable being before the other variable we know that saying the variable that the first variable must be next to occurs before that third variable gives us the same result. Other times, though, these questions can be very difficult, but I don't just automatically skip over them anymore.

Yeah, I agree. They just feel a little more abstract and less mechanical than the other question types.

[bp shinners]

The only question type that still throws me off is the 'Which of the following rules has the same effect as rule X?" ones. Whether or not I get those generally just depends on whether I have enough time left over to diagram each answer choice.

Here's the mechanics for it:
For these questions, the rule you're replacing generally has two parts to it ("H is before L, and H can't be first."). If it only has one part, skip analysis B below.

Step 1) See if there is anything that lends itself to showing up in a right answer (this is the closest to divining an answer I'll get, but if you skip this step, you'll still get it right - it'll just take you a bit slower). By this, I mean that if you're trying to replace 'H is before L' and a rule states L must be immediately before M, then look for an answer with 'H is before M'; analyze that one first.

Step 2) Analyze each answer choice over the following three metrics - if it fails on any of them, cross it out.
A) Does it replace the first part of the rule (in my example, "H is before L")?
B) Does it replace the second part of the rule (in my example, "H can't be first")? (remember to skip this step if the rule being replaced only has one piece to it)
C) Does it tell you any new piece of information? ->This is the hardest metric, see strategy tip below.
If your answers are "yes, yes, no", then you have your answer.

Strategy tip: Analyzing along metric C is, by far, the hardest part of the analysis, as well as the most time consuming. As such, I recommend (if you have an extra 30 seconds or so) to skip it during your first pass through the answers. Eliminate the ones that fail along A or B. If you only have 1 left, you don't have to bother with C - you've got your answer. If you have 2 left (or, very rarely if ever, three), then do the analysis of C for the remaining answers. It's a bit slower, but it could save you time overall if you only have 1 answer left.

[Fiera]

Wow, thanks for all the awesome advice guys! I was starting to think there's some magic key to LG that I just haven't been able to find, but now I'm definitely motivated to not give up. I've started re-doing games (which I hadn't thought about doing before, thanks!) and it's been rough, but I'm just going to keep drilling and drilling. I'm hoping it will click before Oct 5! I'm trying out your guys' strategies and tips too--hopefully I can come back and report a difference!

[LSAT Hacks (Graeme)]

I'm an LSAT tutor, and I often go through the same games 5-15 times with students. I've noticed that I keep seeing new things up until around the 5th time through a game. I really recommend repeating games until you know them COLD. If you know your way through 20-40 games, you'll be better able to spot the hidden paths that exist in all games.

7Sage has a video guide to logic games repetition. They also have free logic games videos if you need help with the deductions on the games.

[albanach]

LG is probably the only section that it's helpful to repeat tests in, so don't worry about using up your material.

I liked to keep a couple of games in my pockets and whenever I had ten minutes free I would do a game.

Follow the PS techniques and practice a lot. You'll get there.

[sam25]

I get thrown by some of the phrasing. For example, the Feb '99 tree game's last 2 rules. This is my best stab at diagramming them. Does anyone have a definitive diagram for these rules?

If yews are not in the park, then either laurels or oaks, but not both, are in the park.

Not Y --> (L --> Not O) and (O-->Not L)

.....or is it

Not Y --> L<-|->O

If it is not the case that the park contains both laurels and oaks, then it contains firs and spruces.

L<-|->O --> F AND S


Is "if it is not the case that both L and O" just another way of saying "L or O, but not both?"

[bp shinners]

I get thrown by some of the phrasing. For example, the Feb '99 tree game's last 2 rules. This is my best stab at diagramming them. Does anyone have a definitive diagram for these rules?

If yews are not in the park, then either laurels or oaks, but not both, are in the park.

Not Y --> (L --> Not O) and (O-->Not L)

This isn't quite right. You're saying here that Yews not being in the park creates two conditional relationships, which it does. If you have Y, you need to have L or O, but you can't have both. With this rule, L, O, and Y could ALL be out of the park.

.....or is it

Not Y --> L<-|->O

Not quite right again, for the same reason. In fact, this is the shorthand for the first one you wrote above, so they're logically equivalent.

I'd diagram it as (~ stands in for 'not'):
~Y-> (L or O) and (~L or ~O)

If you don't have Y, then you have at least one of L and O (L or O), but you don't have both L and O (~L or ~O).

If it is not the case that the park contains both laurels and oaks, then it contains firs and spruces.

L<-|->O --> F AND S
Is "if it is not the case that both L and O" just another way of saying "L or O, but not both?"

This might be the meanest rule ever written. Your diagram means, "If L and O can't both be in the park, the F and S are in the park." Again, you have a sufficient condition that is itself a logical rule, which isn't the case here.

Let's take it part by part (which is what you should do for tricky rules):

"If it is not the case that the park contains both laurels and oaks..."
It's saying here that at least one of these two is NOT in the park, so my sufficient condition is "~L or ~O" - that just means one is definitely out (and both could be out).

"...then it contains firs and spruces."
Easy enough - "F and S"

So we get:
~L or ~O -> F and S
and
~F or ~S-> L and O

so either
1) F and S are in
2) L and O are in
3) All 4 are in
But as soon as one of those four is Out of the park, we know that two others are definitely in. 3 seems like an outlier to me that's likely to show up on one question to make sure you didn't miss it.

[sam25]

Thanks bp shinners! With all the misinformation out there on this game, this helps a lot. Looking forward to trying it again with the correct rules.

[Fiera]

I have to say, I was skeptical about how helpful repeating games would be, but after doing just that and taking a practice exam, I scored a 173 and only got 3 LG wrong/incomplete!!! this is huge for me--I don't want to be overly optimistic but this is issue awesome motivation, eeeek. Thanks guys!