New logic games = plug and chug?

dba415
Posts: 87
Joined: Wed Feb 23, 2011 9:50 pm

New logic games = plug and chug?

Postby dba415 » Mon Sep 24, 2012 4:56 pm

With the combination of the nature of the new games but also the new layout with extra space, does it seem like the LSAT is suggesting that plug-and-chug is the best way to solve some of these games?

I think you might be wasting precious minutes trying to think of any key inferences, when you can be creating scenarios, which you would eventually be doing anyway after you can't come up with any unlocking inferences.

User avatar
beezneez
Posts: 22
Joined: Mon Apr 02, 2012 11:53 pm

Re: New logic games = plug and chug?

Postby beezneez » Mon Sep 24, 2012 7:09 pm

new games? omg wtf?????? (please let me know more, thanks!)

User avatar
CalAlumni
Posts: 214
Joined: Fri May 11, 2012 11:58 am

Re: New logic games = plug and chug?

Postby CalAlumni » Mon Sep 24, 2012 8:31 pm

dba415 wrote:With the combination of the nature of the new games but also the new layout with extra space, does it seem like the LSAT is suggesting that plug-and-chug is the best way to solve some of these games?

I think you might be wasting precious minutes trying to think of any key inferences, when you can be creating scenarios, which you would eventually be doing anyway after you can't come up with any unlocking inferences.


My strength in the old LG's was creating a great diagram based upon important inferences I discovered and then quickly getting through the questions with little hassle or too may hypos. Newer games seem to have less major inferences and therefore time consuming hypos have to be made on virtually each question which is time consuming--but it is especially time consuming if you spent a nice chunk of time creating a great diagram that subsequently doesn't help you.

User avatar
cahwc12
Posts: 941
Joined: Mon Jul 02, 2012 4:49 pm

Re: New logic games = plug and chug?

Postby cahwc12 » Tue Sep 25, 2012 8:11 am

CalAlumni wrote:
dba415 wrote:With the combination of the nature of the new games but also the new layout with extra space, does it seem like the LSAT is suggesting that plug-and-chug is the best way to solve some of these games?

I think you might be wasting precious minutes trying to think of any key inferences, when you can be creating scenarios, which you would eventually be doing anyway after you can't come up with any unlocking inferences.


My strength in the old LG's was creating a great diagram based upon important inferences I discovered and then quickly getting through the questions with little hassle or too may hypos. Newer games seem to have less major inferences and therefore time consuming hypos have to be made on virtually each question which is time consuming--but it is especially time consuming if you spent a nice chunk of time creating a great diagram that subsequently doesn't help you.


I think you'll be surprised to find that in all but a very few (maybe 2 or 3?) games if you redid them without some magnificent initial diagram highly dependent on inferences and instead just wrote the rules and went to the questions, you would almost always have finished the games in equal or less time.

One strength of the new approach as well is that because the games don't hinge on crucial inferences anymore, the game times are not nearly as volatile (4 minutes if you catch the inference(s), 8-9 minutes if you don't).

Swimp
Posts: 493
Joined: Sat May 26, 2012 9:32 am

Re: New logic games = plug and chug?

Postby Swimp » Tue Sep 25, 2012 8:18 am

I think the new layout confirms that the shift toward hypo-heavy games isn't just people's imagination.

I don't really care one way or another (LG are my weak point either way), but I have to admit I don't understand why the testsmakers made this particular change. Seems like a dumbing down of the games section.

User avatar
cahwc12
Posts: 941
Joined: Mon Jul 02, 2012 4:49 pm

Re: New logic games = plug and chug?

Postby cahwc12 » Tue Sep 25, 2012 10:04 am

Swimp wrote:I think the new layout confirms that the shift toward hypo-heavy games isn't just people's imagination.

I don't really care one way or another (LG are my weak point either way), but I have to admit I don't understand why the testsmakers made this particular change. Seems like a dumbing down of the games section.


Because if you get lucky and happen to notice some key deduction in the game, you can absolutely crush it. By eliminating those insightful deductions, rather than base it on luck, it's based on your synthesis of a set of rules.

As for the space change, the other sections are eight pages, and some people write larger than others. I've never had issue with space in a game, but I also write very small, so that gives me a natural advantage over someone who writes larger. I think the space change is entirely minor and not due to any lean toward increased hypothetical scenarios. It's just one less thing for people to complain about (lack of space).

In fact, before the June tests were handed out at my test center, two people next to me were complaining about how the LSAT didn't allow you any scratch paper for games since they never seemed to have enough space. I'm sure they were pleasantly surprised.

bp shinners
Posts: 3091
Joined: Wed Mar 16, 2011 7:05 pm

Re: New logic games = plug and chug?

Postby bp shinners » Tue Sep 25, 2012 11:33 am

cahwc12 wrote:I think you'll be surprised to find that in all but a very few (maybe 2 or 3?) games if you redid them without some magnificent initial diagram highly dependent on inferences and instead just wrote the rules and went to the questions, you would almost always have finished the games in equal or less time.


I don't think I could disagree more strongly. These inferences aren't magic things that appear to some when the moonlight is right and the tides are out - they're straight-forward combinations of rules. If you know what to look for, they should take you seconds to find; and what you're looking for is a player/slot/distribution that shows up more than once. If you get them, they make the questions trivial. If I plug-and-chug a game, it takes me much longer than if I make the inferences. Also, if I plug-and-chug a game, there are more chances for me to make a mistake.

So I'll stick with my inferences that increase speed and accuracy.

bp shinners
Posts: 3091
Joined: Wed Mar 16, 2011 7:05 pm

Re: New logic games = plug and chug?

Postby bp shinners » Tue Sep 25, 2012 11:35 am

cahwc12 wrote:Because if you get lucky and happen to notice some key deduction in the game, you can absolutely crush it. By eliminating those insightful deductions, rather than base it on luck, it's based on your synthesis of a set of rules.


Again, there's no luck to it. There are specific things that tell you there are deductions/inferences to be made. If they exist, make them. If they don't, move on.

To the OP - I keep hearing this on the forums, and I honestly think it's group-think infecting how people are approaching these games. I teach each Games section that comes out at least 3 times to tutoring students, and I haven't noticed a decrease in the number of Games that lend themselves to scenarios/deductions, or an increase in Games that require a straight-up plug-and-chug approach. They're there - they might just be hidden a little bit more than usual (which lines up with the overall increase in the difficulty of the exam).

User avatar
cahwc12
Posts: 941
Joined: Mon Jul 02, 2012 4:49 pm

Re: New logic games = plug and chug?

Postby cahwc12 » Tue Sep 25, 2012 1:00 pm

bp shinners wrote:
cahwc12 wrote:Because if you get lucky and happen to notice some key deduction in the game, you can absolutely crush it. By eliminating those insightful deductions, rather than base it on luck, it's based on your synthesis of a set of rules.


Again, there's no luck to it. There are specific things that tell you there are deductions/inferences to be made. If they exist, make them. If they don't, move on.

To the OP - I keep hearing this on the forums, and I honestly think it's group-think infecting how people are approaching these games. I teach each Games section that comes out at least 3 times to tutoring students, and I haven't noticed a decrease in the number of Games that lend themselves to scenarios/deductions, or an increase in Games that require a straight-up plug-and-chug approach. They're there - they might just be hidden a little bit more than usual (which lines up with the overall increase in the difficulty of the exam).


I respect you a lot for the continued awesome advice you give on this forum so it's difficult for me to disagree with you, but I genuinely believe advocating making inferences before going to the questions to be poor advice. I will say that I think this approach can allow you to finish games sections in time, but it is simply not as efficient as not making these deductions beforehand.

By simply writing rules and going to the questions, I consistently not only -0 games sections, but finish with 7-10 minutes leftover to double check most of my answers. This time is saved by not bothering to make non-obvious deductions that the game will elucidate as needed. I can look at the rules and pretty quickly see how they fit together, but I don't take the time to write out scenarios and make guesswork. Instead, I let the questions tell me exactly where those deductions are (and they always will), which is a guaranteed hit versus a more luck-based ("I hope I find it if I do this!") chance hit.

The problems with trying to squeeze deductions out of games are numerous:

- before going to the questions you may or may not get all of these deductions
- the ones you get may or may not be helpful to you in the game
- the work you do is only transferred, not saved, if it is tested; there is no time-savings when you do the work required for Q17 by manipulating the rules before ever reading Q17
- you run the risk of not finding deductions if they aren't there and wasting time merely playing with the rules, rather than applying them directly to the questions

There are only three games out of the 276 (I haven't seen PT39 or 56 games) I've taken where playing with the rules beforehand actually netted me a time savings, and in two of those three it wasn't very much. That means that, in my opinion, 98.9% of the games that have existed on the LSAT, it was not more efficient to spend additional time on games deductions rather than moving on more quickly to questions. And in 99.6% of cases, it wasn't a significant time savings.

Said another way, in my experience, 98.9% of games can be solved more efficiently by not spending time attempting to make these deductions beforehand. It's akin to being told there's treasure on an island and running off digging in the most likely spots before the guy can hand you a map with all the exact locations marked.


edit: just to be clear, because I'm not sure we really are arguing entirely different methods of attack, here's what I would do for a typical game.

PT23G2 is the interviews and hirees game. There are 6 given rules which can certainly be manipulated and inferences can certainly be made about them. I would write the rules as I read them, create a quick initial diagram with what is known, and then go right into the questions, no more than 40-60 seconds after starting the game.

If I understand you correctly, you would advocate spending time manipulating these rules to create new deductions or inferences about the game. Here's a list of all the inferences that can be made versus the ones the explicitly stated (and diagrammed) rules:

Image

I don't want to distort your own argument, but if I understand you correctly, you would advocate for making some or all of these written deductions before going into the questions. I just can't in good faith support that assertion (and please explain an example of what you mean if this is not accurate). The ONLY things I would do with these rules is combine the first and second one as I write them (since that is a blatantly obvious combination) and then mark F directly into my main diagram.

bp shinners
Posts: 3091
Joined: Wed Mar 16, 2011 7:05 pm

Re: New logic games = plug and chug?

Postby bp shinners » Tue Sep 25, 2012 9:03 pm

cahwc12 wrote:By simply writing rules and going to the questions, I consistently not only -0 games sections, but finish with 7-10 minutes leftover to double check most of my answers. This time is saved by not bothering to make non-obvious deductions that the game will elucidate as needed. I can look at the rules and pretty quickly see how they fit together, but I don't take the time to write out scenarios and make guesswork. Instead, I let the questions tell me exactly where those deductions are (and they always will), which is a guaranteed hit versus a more luck-based ("I hope I find it if I do this!") chance hit.


While I don't doubt your assertion, I must say that I think you'd be surprised how ineffective that method is for the vast majority of people out there. I've sat down with plenty of students, and when they write out the rules and go straight to the questions, they take about 3 minutes/question to get through it, and they don't have very solid accuracy.

And to completely use the same argumentative strategy against you, I consistently score -0 on games with 15-20 minutes left to double check everything by making deductions and drawing out scenarios before I go to the questions.

So, in short, I'm not saying that making these deductions is necessary to doing well and doing so quickly. However, for most people, it's a better way to go about the game than just going straight to the questions. Being able to do that tells me that you're naturally good at games, which allows you to make these deductions over and over again throughout the section without error. That's not the usual situation for people taking the LSAT.

You say, "I can look at the rules and pretty quickly see how they fit together" which is great for you, but most people in my classes would give you a dirty look. They don't see how they fit together, and they need to sit there and work with them for a few seconds before they can start to intuit how they work together.

I also think you don't understand what I mean by making deductions (as is the case with most people who use your method). There's no guesswork involved. There's no luck involved. And while the questions tell you where those deductions are, so do the rules - and using the rules to find them is much more efficient than testing out answer choices. The entire process of making deductions/scenarios takes people with whom I work maybe 45 seconds. If you come up with 1 major deduction, that's going to save you 2-3 minutes throughout the game. And if you don't make a deduction, then it's going to take you a whole lot less time than 45 seconds (maybe 15). Either way, you end up with a better understanding of the game. Most people who don't like making deductions/scenarios think that there's something more complex and less obvious going on than there actually is. All I'm trying to do is combine a few rules, or see if there's a rule that's so strong it limits me to 1-2 worlds.

As to the game you point out, I didn't quite follow everything you were saying, doing, but here's what I'd do:
H?:
IN:
________________
OUT:

1) G->J
~J->G

2) J->L
~L->~J

3) Write into setup

4) ~K->~F(h)
F(h)->K

5) ~M->~K(h)
K(h)->M

6)M(h) and L->O(h)
~O(h)->~M(h) or ~L
______________________________
DEDUCTIONS
I look for any term that shows up in more than 1 rule: here, that list is J, K, M, and L. I can combine the J rules. I can't combine the other ones because they also talk about being hired/are part of disjunctions.

G->J->L
~L->~J->~G
______________________________
SCENARIOS
I look for any rule that's strong enough to make scenarios. For a grouping game, that list is (1) Must Be Together/biconditionals, (2) at least one (~A->B) rules, (2a) the same term that shows up in the sufficient condition of two rules with opposite negation/team assignment (then I have a scenario where that player is IN/Team A, and one where he's OUT/Team B), and (3) weird rules (you know, the ones that say something like exactly two of ABC must be chosen). I don't have that here. Question time. And, glancing at them quickly, I think I've probably got everything because they're pretty much all conditional (one elimination - the first one - that you don't have to do any work whatsoever to do ever, and one absolute, but it's a could be true that's asking about group sizes) - if there were deductions, I'd expect more absolute questions.

So for this one, very little difference between what you'd get and what I'd get. However, this game doesn't really have any major deductions or good scenarios. If you look at a game like the scientist panel (botanists, zoologists, chemists), however, I'd argue that it's damn near impossible for most people without working ahead of time on some deductions. And ask the people who had the mauve dinosaur game how far they got with plug-and-chug.

Also, for most people, having the contrapositive written out will help - you have no idea how common it is to completely miss the application of, or misapply, the contrapositive without writing it out.

So, in short, this doesn't work for everyone, and some people will plug-and-chug quite well. However, from personal experience, most people don't have that talent. And I couldn't even begin to approach the speed with which I do games if I didn't spend the time up front to 'solve' the game before moving on.

-edit- For copying something down wrong.
Last edited by bp shinners on Tue Sep 25, 2012 9:07 pm, edited 1 time in total.

bp shinners
Posts: 3091
Joined: Wed Mar 16, 2011 7:05 pm

Re: New logic games = plug and chug?

Postby bp shinners » Tue Sep 25, 2012 9:05 pm

cahwc12 wrote:The ONLY things I would do with these rules is combine the first and second one as I write them (since that is a blatantly obvious combination) and then mark F directly into my main diagram.


And just a follow-up, that's really the only thing you can combine. So yea, for this game, not a whole lot of difference between our methods. If you pick a game that's all absolute questions, I guarantee we'll diverge quite a bit.

User avatar
Br3v
Posts: 4174
Joined: Mon Jun 13, 2011 7:18 pm

Re: New logic games = plug and chug?

Postby Br3v » Tue Sep 25, 2012 9:22 pm

Newer games seem to have less major inferences and therefore time consuming hypos have to be made on virtually each question which is time consuming--but it is especially time consuming if you spent a nice chunk of time creating a great diagram that subsequently doesn't help you.

User avatar
NoodleyOne
Posts: 2358
Joined: Fri May 25, 2012 7:32 pm

Re: New logic games = plug and chug?

Postby NoodleyOne » Wed Sep 26, 2012 6:28 am

Agree with shinners. For the record I'm a -0 LGer but this over emphasis on hypos vs inferences is misguided. I think tlsers are so focused on finding a pattern that they are seeing ones that don't exist. Efficient hypos have always been a key skill, but so have efficient inferences. This has not changed in the most recent tests.

User avatar
cahwc12
Posts: 941
Joined: Mon Jul 02, 2012 4:49 pm

Re: New logic games = plug and chug?

Postby cahwc12 » Wed Sep 26, 2012 10:07 am

To be clear, by "I can see how they fit together" I mean that I can see that they fit together in some way. Exactly how I don't know, because I don't spend that time playing with the rules to figure it out. I recognize that there are variables common to multiple rules, but I only combine them when it is obvious (as in the earlier example). The entire point of my approach is that you intentionally forego this time because the questions will always give you the most efficient way to figure the game out.

I didn't remember this scientist panel game so I dug it up and re-did it. Maybe you did it faster than me (close to 5 minutes), but even if you did I contend you'd have been even faster had you used my approach. I really feel that this plays to the strengths of my method and not yours, and here's how I did it:

Image

The element of luck I've been referring to is how quickly you arrive at key deductions or scenarios while playing with the rules, not whether or not you are able to make them. In this game, after trying to see what scenarios or inferences I can pull out of the game, I realize that you can create the hypo for Q2 beforehand. But this doesn't help you because in Q2 you would be prompted to produce this exact diagram anyway. Either you draw it twice, losing time, or you drew it beforehand, thus preempting the need to draw it, still not saving time. But to produce this hypo at all originally, you must look at all the rules, think to yourself what gives you the most information (producing a scenario from either K or M since they are most frequent, then deciding to go with M, then producing the diagram by applying the rules.)

I suppose you could additionally create a diagram that shows 2+ of B included and then 2/3 out for Z, or 2+ Z and then 2/3 out for B, but that seems highly unhelpful when you're using the rules to arrive at those same deductions while doing the questions--the difference being you didn't create scenarios beforehand that aren't directly useful.

Note that in each of the 5 questions (I didn't include Q1 for obvious reasons), all I did was apply the rules. There's no magic in my approach just as there's no magic in yours. The difference is that I don't spend time digging for treasure before being given the map with locations of the treasure. Each of these four hypo questions is solved by a single hypo and application of each of the rules.

I feel like massaging the game and rules beforehand is like a safety net or crutch for test-takers that, after a certain point, you shouldn't need and will hamper your score. I will also agree that for most test-takers it represents an improvement in their games-solving ability. But your method should not be an end, but rather a means to an end. I wasn't good at games when I started prepping for the LSAT. This game through tons of practice on my own and not anything else.

Writing out contrapositives is another crutch. It certainly will be helpful for most people, but once you can train yourself to read a conditional rule backwards for the contrapositive, it makes you faster and stronger at dealing with conditionals. It took some practice, but I'm able to do it effortlessly now, and it saves time without cutting accuracy. The other reason I don't like contrapositives is for the same primary reason I don't like pre-question hypotheticals: redundancy. At best, when you create a scenario before going to the question, it gives you a better sense of the game and you do the work that ends up solving a question before you get to it. At worst, you've confused yourself by playing with all the rules and running a dozen questions in your head and have created a dearth of work that is not relevant to any question in the game.

I also realize that potentially this game is not a great example for you because it has mostly hypothetical questions, whereas some games have more global questions (where the strength of pre-emptive inferences would seem to shine more). In a global (absolute) question, I read through the question and apply the rules to the answer choices, usually eliminating 1-3 ACs. Then I do something else--I move on. Again, I don't waste time making extraneous scenarios beforehand. Each hypothetical question in a game is a scenario with which to practice the rules and understand the game better. Once I have finished all those questions, I can use that work to eliminate additional answer choices in those questions. Then I am typically left with one or two ACs, either solving the question for me, or allowing me to create a single hypo and solve it that way.


I really, truly, honestly believe this is bar-none the best approach for solving games--not just for me, but for anyone. I would possibly recommend your approach to beginning test-takers until they gain a certain comfort level with games, and I would only do it with the presumption that they would ultimately graduate from using this approach later on to one that shifts the focus to solving the questions and allowing the questions to elucidate those inferences. The only time I would advocate for pre-question inference hunting would be if the game had extremely limiting rules and a dearth of global questions, and there hasn't been a game like that in probably 15 years, certainly the last 5.

Also, please don't misunderstand--I was overly harsh in saying it's poor advice that you've given. It is good advice--I just think mine in this regard is better. I think with your approach one could consistently solve games and score just as well. The difference is that by diving into the questions as soon as possible (and checking only for very, very obvious inferences beforehand), you would be just as accurate but faster.

User avatar
LSAT Hacks (Graeme)
Posts: 371
Joined: Wed May 30, 2012 9:18 pm

Re: New logic games = plug and chug?

Postby LSAT Hacks (Graeme) » Wed Sep 26, 2012 12:19 pm

I agree with BP shinners. Inferences aren't some magic thing. They (can) come fairly naturally once you've seen how the rules work together on a variety of games. You learn to spot patterns. For example, if 2-3 rules each mention the same variable, or concept --> probably there is an inference.

I went from teaching exclusively 29-38 (It was what Canadian students used) to 52-61, and I have noticed a shift towards *fewer* inferences. But they're still there, and if they're there, the game is much easier if you make them.

*Sometimes* I won't get an inference until I do the first 1-2 questions. But if the questions seem hard, it usually means that I missed something in the setup, not that the questions are actually hard per se.

I've noticed that newer games tend to test your command of the rules. They can be simplified by adding certain rules to the diagram ("not" rules, for example), and making sure you *know* the rest of the rules.

bp shinners
Posts: 3091
Joined: Wed Mar 16, 2011 7:05 pm

Re: New logic games = plug and chug?

Postby bp shinners » Wed Sep 26, 2012 3:24 pm

cahwc12 wrote:The only time I would advocate for pre-question inference hunting would be if the game had extremely limiting rules and a dearth of global questions,


I think we're just going to have to agree to disagree on this one. We have seen different results with different methods, so we're both pretty set in our ways. I've tried your approach and found it slower, and I've found most of my students aren't able to use it to great effect. However, I'm more than willing to admit that my statistics might be biased by having your method taught by someone who thinks a different method is better.

The one thing I think is funny, though, is that the one time you would advocate my method is the one time I wouldn't advocate it. When there are a lot of global questions (we call them absolute questions), you absolutely should find those inferences. When there is a dearth of them, there aren't any inferences to be found. A global/absolute question is essentially asking, "Did you find the inference while making your setup?" or from your perspective, "Can you find the inference now that I'm giving you a list of possible ones?" To me, it will always be easier to prove something up front using the rules than it would be to go through the ACs and try to prove each one until you find the one that works. I advocate the latter method as a failsafe in case students don't find the inference up front (as the answer to an absolute, must be true question is an inference), but I find it much faster and more accurate to answer it before going to the question.

So, in short, I'm not saying you're wrong. I'm just saying that I think you're wrong ;-).




Return to “LSAT Prep and Discussion Forum”