Logic Games: Deductions Or Not? Forum

[patricksouthern]

I may open a veritable can of worms with this question, but I believe it merits asking anyway.

First, the background: I didn't decide to take the LSAT until about a month and a half ago. I'm taking the June test anyway, because of a few factors. I'm traditionally a strong standardized tester, plus I'm getting married in late July and my work schedule becomes far more difficult to deal with when August/September roll around. For all those reasons, the June test seemed to make more sense to me than putting it off until fall.

So far, I've fared reasonably well. My initial diagnostic was a 163, and I've PT'd as high as 175 (that came on Sunday, so I'd say I'm peaking at the right time). I'm quite strong at LR, thanks largely to the methods and practice I got as a result of taking Blueprint's online course. If I miss more than 2 in any LR section, it's a surprise at this point. Because of the limited study time I had (and the fact that I contracted a rather nasty case of food poisoning that basically left me incapable of much other than sleep about 2 weeks into my prep), I didn't truly work through any methods for attacking RC, but I've consistently scored about a -3 there, which I'm OK with.

My biggest problem has been LG. My initial diagnostic was a train wreck there (-10), and progress has only been incremental ever since. Some may recall I posted about a week ago concerning my issues with identifying key deductions, and that is still my biggest issue and fear. On the PT where I scored 175, I had easily my best games section ever (-1). But those performances are few and far between, and I'm quite apt to get stumped by difficult games. I struggle with speed, as I invariably have to work through several hypotheticals in some questions to get to an answer. This often means I'm left to guess at the last few questions.

The crux of it all is this: Blueprint's method relies critically on making key deductions and "scenarios" and while I've gotten a bit better in those areas, I'm hardly ready to walk in and tackle a full LG section and feel confident I'll score 20+ consistently because I am not consistently good at identifying those inferences.

To try to address my weakness, I began working through PowerScore's LG Bible in recent days. I'm still using it and the explanations do seem to help a bit, but the methods are substantively the same -- swap out the word "Scenarios" for "Templates" and you have largely the same mode of attack.

My mind was blown when I read through the entire 44-page Dave Hall thread on here (yes, all of it) and he repeatedly said he didn't use inferences at all to attack LG. Intrigued, I went to Velocity's site and watched his explanations for the June 2007 Games section. Indeed, his methods seem very, very different and a bit more straightforward for someone like me who has struggled to make deductions.

If you've read through all that, sorry for my long-windedness, but here are my ultimate questions: do you believe a strategy that doesn't rely crucially on inferences can be as effective as one that does? With less than a week to go before my LSAT, is it even worth giving another way of attacking LG a shot? Perhaps worth the $250-something it might cost to get Velocity's Games videos? Has anyone made a switch like this "midstream" and had any sort of positive or negative experience to share?

[Easy-E]

Are you PTing where you want to be right now?

[patricksouthern]

Are you PTing where you want to be right now?

I'd take that 175, for sure. Anything below 170 I'd be disappointed with. Low 170s I'd begrudgingly accept. If I could truly master LG, I honestly think 174-175 is my worst-case scenario.

[Easy-E]

Are you PTing where you want to be right now?

I'd take that 175, for sure. Anything below 170 I'd be disappointed with. Low 170s I'd begrudgingly accept. If I could truly master LG, I honestly think 174-175 is my worst-case scenario.

What were your last, say, 5 scores on PTs? And were they recent?

[patricksouthern]

Are you PTing where you want to be right now?

I'd take that 175, for sure. Anything below 170 I'd be disappointed with. Low 170s I'd begrudgingly accept. If I could truly master LG, I honestly think 174-175 is my worst-case scenario.

What were your last, say, 5 scores on PTs? And were they recent?

Blueprint only has you take six PTs during the course, and I just finished the course on Sunday. So the scores show a rather direct progression and represent growth over the course of about a month. After my 163 diagnostic, my PTs have been 169 (LG -9), 168 (LG -6), 170 (LG -5), 172 (LG -3) and 175 (LG -1). I'm planning to take another 3 four-section PTs between now and Monday.

[Swimp]

OP-- LG are my albatross as well. Also, like you, I was interested to see Dave Hall discuss his technique in that other thread. I tried it out a little yesterday and came to the conclusion that I don't do any worse using his diagram style than I do with the PS diagram style. I haven't tried the "Hall Method" on any timed sections yet, but my current plan is to stick with it for a week or two and see where I am. I suspect that either method will work with enough practice, but since Dave's style involves more scribbling and less elegance, I'm optimistic that the learning curve will be milder.

Here's hoping!

[humbugger]

LG is 100% experience and confidence. Once you realize that every game has definite solutions and you know you can answer the questions, they're pretty straightforward. If you're not mentally prepared, you might panic during the actual test and really blow it.

Get your comfort level and take the October test. It only takes half a day.

[Helicio]

For a while I did Powerscore's method. Then I just started making maybe 2 or 3 inferences and going straight to the questions. This works far better. Hypotheticals also make it less likely that you will get questions wrong, I have found, because you have something concrete and verifiable to work with.

I routinely score -0 to -1 to at most -2 on LG and use hypothetical-heavy strats all the time. I just did the dino game for the first time today in 8 min with my method and dominated it.

Honestly, hypotheticals are (to me) the way to go. You have nothing to worry about.

Let's take this test Monday and own it.

[Br3v]

For a while I did Powerscore's method. Then I just started making maybe 2 or 3 inferences and going straight to the questions. This works far better.

+1

[Helicio]

If you are having trouble knowing when to stop making inferences, think of it like you think of making popcorn. As long as the corn's popping fast, keep going. But if it slows down too much, if only 1 or two are popping within 30 seconds, you know it is time to move on.

Good luck, and if I were you I would take it Monday. 175 means that at this point it is going to come down to how you feel on test day/factors like that.

I've been PTing around 175 and am taking Monday.

[rglifberg]

I bought the Velocity LG section and it has worked really well for me, especially on the recent sections where making key deductions really isn't necessary. I do a combination of Velocity and Powerscore's method, because sometime making simple deductions makes the game a lot easier, especially for games with a lot of conditional statements.

[TERS]

If you are having trouble knowing when to stop making inferences, think of it like you think of making popcorn. As long as the corn's popping fast, keep going. But if it slows down too much, if only 1 or two are popping within 30 seconds, you know it is time to move on.

Nice!

[flem]

For a while I did Powerscore's method. Then I just started making maybe 2 or 3 inferences and going straight to the questions. This works far better.

+1

I'd like to +1 this as well. PowerScore's method is pretty good but it relies too much on the "stare at it and think really hard" method, where Dave Hall and bp shinners seem to just look for overlap, try and make a deduction, and just let the rules guide the game. Draw out a hypothetical or two and get going.

Also look for a global "Must be True" question, because if you missed a major deduction those questions will help you out.

[Easy-E]

For a while I did Powerscore's method. Then I just started making maybe 2 or 3 inferences and going straight to the questions. This works far better.

+1

I'd like to +1 this as well. PowerScore's method is pretty good but it relies too much on the "stare at it and think really hard" method, where Dave Hall and bp shinners seem to just look for overlap, try and make a deduction, and just let the rules guide the game. Draw out a hypothetical or two and get going.

Also look for a global "Must be True" question, because if you missed a major deduction those questions will help you out.

When you say draw out a hypothetical or two, do you mean like open frames, or just two possible ways the game could go, completely filled?

And I second the notion on non-conditional "Must be True". Pretty much a bonus rule.

[flem]

When you say draw out a hypothetical or two, do you mean like open frames, or just two possible ways the game could go, completely filled?

And I second the notion on non-conditional "Must be True". Pretty much a bonus rule.

Just draw one or two workable scenarios in addition to your diagram. For me it internalizes the rules more and it sometimes pays off with local questions.

[patricksouthern]

I appreciate the guidance and advice of those who have replied. I think, at the very least, I'm going to start tackling the conditional (i.e. "local") questions in every game immediately after doing the "elimination" (rule-checking) question. That, in addition to drawing out one or two complete hypotheticals, should provide me with a lot of possible outcomes heading into tougher global questions that I might otherwise sink a ton of time on without having made deductions.

I'm planning on taking a four-section PT this afternoon, so I'll be interested to see how this method works for me.

[Easy-E]

I appreciate the guidance and advice of those who have replied. I think, at the very least, I'm going to start tackling the conditional (i.e. "local") questions in every game immediately after doing the "elimination" (rule-checking) question. That, in addition to drawing out one or two complete hypotheticals, should provide me with a lot of possible outcomes heading into tougher global questions that I might otherwise sink a ton of time on without having made deductions.

I'm planning on taking a four-section PT this afternoon, so I'll be interested to see how this method works for me.

Oh I support this approach 100%. Do the global question first (possible ordering...), then the conditionals. The hypotheticals you create from these questions can help you immensely in the "must be true" and "could be true" questions. Nothing better than having a could be true and just going "oh there it is" and moving on.

[Easy-E]

When you say draw out a hypothetical or two, do you mean like open frames, or just two possible ways the game could go, completely filled?

And I second the notion on non-conditional "Must be True". Pretty much a bonus rule.

Just draw one or two workable scenarios in addition to your diagram. For me it internalizes the rules more and it sometimes pays off with local questions.

Yeah, seems like it could be helpful in drawing an inference or two, or at least understanding how the elements are really restricted or not. Do you do this for most games, or just ones where your initial diagram is pretty open?

[flem]

Yeah, seems like it could be helpful in drawing an inference or two, or at least understanding how the elements are really restricted or not. Do you do this for most games, or just ones where your initial diagram is pretty open?

I try and always do it, because it's beneficial either way. If there's a limited amount of workable solutions uncovered by huge points of restriction in hypotheticals, you can wreck that game's shit in no time. If it's open ended, then you're just learning some more by doing so.

I appreciate the guidance and advice of those who have replied. I think, at the very least, I'm going to start tackling the conditional (i.e. "local") questions in every game immediately after doing the "elimination" (rule-checking) question. That, in addition to drawing out one or two complete hypotheticals, should provide me with a lot of possible outcomes heading into tougher global questions that I might otherwise sink a ton of time on without having made deductions.

This is a good strategy, good luck.

[Easy-E]

Yeah, seems like it could be helpful in drawing an inference or two, or at least understanding how the elements are really restricted or not. Do you do this for most games, or just ones where your initial diagram is pretty open?

I try and always do it, because it's beneficial either way. If there's a limited amount of workable solutions uncovered by huge points of restriction in hypotheticals, you can wreck that game's shit in no time. If it's open ended, then you're just learning some more by doing so.

I'll try it out on my timed sections today, thanks. Late in the game for new techniques I suppose, but can't hurt to try.

[jigglebottom]

how about knocking off wrong answer choices in the rule check question at the beginning as you are drawing out the main diagram and rule set? I do this... probably saves about a minute. After each rule I just go straight to the first question and try to knock wrong ones out.

[Easy-E]

how about knocking off wrong answer choices in the rule check question at the beginning as you are drawing out the main diagram and rule set? I do this... probably saves about a minute. After each rule I just go straight to the first question and try to knock wrong ones out.

I've started doing this too, assuming the first question is a complete rule check question (not just partial ordering or whatever). Definitely saves time. If I can, I'll try to see if any of my hypotheticals, including that first, have any variables I can switch around without breaking a rule and give them a quick mark.

[flem]

how about knocking off wrong answer choices in the rule check question at the beginning as you are drawing out the main diagram and rule set? I do this... probably saves about a minute. After each rule I just go straight to the first question and try to knock wrong ones out.

I'm going to start doing this to save time, thank you

[patricksouthern]

how about knocking off wrong answer choices in the rule check question at the beginning as you are drawing out the main diagram and rule set? I do this... probably saves about a minute. After each rule I just go straight to the first question and try to knock wrong ones out.

I've started doing this too, assuming the first question is a complete rule check question (not just partial ordering or whatever). Definitely saves time. If I can, I'll try to see if any of my hypotheticals, including that first, have any variables I can switch around without breaking a rule and give them a quick mark.

Why had I never thought of this? I'm definitely going to try this on my PT today. If it saves time, I'm definitely for it at this point, as speed on LG is probably my biggest issue of anything on this entire exam. Thanks for sharing.

[glucose101]

Use Velocity LSAT. You rarely have to make outright deductions like I used to have to do. Frees up so much mental capacity.