Predictions for the September 2019 LSAT

BPPaaron-lsat-blog-december-lsat-predictions

The September 2019 LSAT is approaching — perhaps a bit too quickly for comfort, for some. While the upcoming September test promises a lot of last-minute cramming and practice exams and bouts of panic-induced mania for those signed up to take it, it promises a different time-honored tradition for us at Most Strongly Supported: the predictions post.

Before each LSAT, our crack-team of bloggers makes hyper-specific predictions for the LSAT, which can range from the outlandish and unlikely to the ridiculously outlandish and incredibly unlikely. The let’s-charitably-call-it-a-“joke” underlying these posts was that the LSAT always tests the same concepts and skills, so predicting what’s going to be on the test is kind of besides the point. So we might as well have some fun and predict that the third Reading Comprehension passage is going to be about the cultural history of Norwegian death metal, or whatever.

But this year, we want to provide some more serious prognostication. Even if we’ve joked in the past, we have been tracking, very closely, which types of games, passages, and Logical Reasoning question types have been recurring most frequently on recent exams. So for this September predictions post, I’m going to put on my best Nate Silver cosplay and use data to make some honest predictions. That way, you can at least have a sense of what’s the most likely to appear on the September test, and can focus on those games, passages, and LR questions in the waning days before the September exam.

That said, our usual caveats still apply. We have no inside knowledge of what the test writers are going to put on the September exam. On a personal level, I’m wrong, a lot (just this year, for example, I saw a can of White Claw and thought, “Pfft, hard seltzer? No way that’s gonna catch on.”). But we pay attention to the LSAT about as closely as anyone could, and we think can offer some insight into what’s likely to be on the September test. So let’s get to prognosticating …

Logical Reasoning:

Here’s how I think the questions are going to be distributed on the September test:

Must Be True: 2 questions (since December 2013, an average of 1.59/exam)
Soft Must Be True: 4 questions (average of 5.18/exam)
Must Be False: 0 questions (average of 0.59/exam)

Main Point: 1 (average of 2.18/exam) ,
Describe: 1 (average of 1.24/exam)
Flaw: 8 (average of 7.65/exam)
Parallel: 2 (average of 2/exam)
Parallel Flaw: 2 (average of 2/exam)
Role: 1 (average of 1.94/exam)
Disagree: 4 (average 1.94/exam)
Agree: 1 (average of 0.29/exam)

Strengthen: 10 (average of 8/exam)
Weaken: 3 (average 3.53/exam)
Crux: 0 (average of 0.77/exam)
Sufficient: 3 (average of 2.76/exam)
Necessary: 6 (average of 5.12, exam)
Resolve/Explain: 3 (average of 3.93/exam)

What we call the “Implication family” of LR questions — which includes Must Be True, Soft Must Be True, and Must Be False questions — has been trending down for years now, so I don’t think there will be more than six of these on the entire test. There’s been a slight uptick of Must Be True questions recently, so I think they’re going to slightly exceed their average prevalence. I’d expect one or both of those to be diagrammable. There’s almost always between four and six Soft Must Be True questions, but the average amount of them has been trending down recently, so I expect that number in September is on the low-end of the range. Must Be False questions — a personal favorite of mine — have been more-or-less dying on the vine recently, only showing up once per testing year. Since there was a Must Be False question on the June 2019 exam, I don’t expect to see one here.

Moving over to the “Characterization family” of questions — those that ask you describe some feature of an argument — I don’t think there’s going to be more than one Main Point question. There have been way more than the average number of those on recent tests — with three appearing last June and four appearing last November — so I think the number is going to regress a bit on this exam. Likewise, there were four Role questions in June 2019, way more than the average. Respect to the ‘Bama fans out there, but I expect the Role tide to recede in September. On the other hand, I expect the number of Flaw questions to exceed its average. There has been between six and nine Flaw questions on a given exam since time immemorial. June was at the low-end of that range with six, so I’m slightly bullish on those for September. There are always exactly two Parallel and two Parallel Flaw questions per exam, so I’m feeling pretty confident with that prediction. Finally, I think there are going to be a lot of Disagree and Agree questions; this year feels like 2017 to me, when there was just a total proliferation of those question types.

Finally, to the “Operation family” — those questions that ask you to find the answer choice that changes the argument some way. The one unmistakable trend of Logical Reasoning is the increasing prominence of the Strengthen question. Since December 2017, there’s been between eight and eleven Strengthen questions on each exam. I think there are going to ten in September, and I’m guessing that three or four of them will be Strengthen Principle questions. That said, I’m only predicting three of the similar Weaken questions, as it’s rare to get more than thirteen Strengthen and Weaken questions in an exam. Crux questions are rare, and there was one in June, so I don’t think we’ll see one on test day. And I’m predicting a lot of Sufficient and Necessary assumption questions. For Necessary questions, in particular, there are almost always between four and six, but there were only three in June. I expect the test writers to dig a bit deeper into the bag of Necessary questions for this exam.

I’m not super eager to predict anything beyond this in Logical Reasoning. I went out on a limb with a prediction in June 2019 and got burned, after all. But old habits die hard, so here’s one of our patented hyper-specific predictions: look out for a question — probably a Flaw, Strengthen, Weaken, or Necessary question — about archaeological evidence. The author of the argument will find some artifact and use it to make some conclusions about the ancient people who presumably used that artifact. That author will be assuming that the artifact was always in the location it was found, or that the artifact has the same uses in ancient times that it would have today. This flaw has been super common on the LSAT of late — check out question 21 in the first Logical Reasoning section of the December 2017 test if you want a good example of such a question.

Reading Comprehension:

Reading Comp — the unloved and overlooked middle-child of the LSAT — will be a slog. That much I can predict, proverbial hat firmly in hand. These passages have only become more difficult-to-understand, and the questions have only become more challenging, as the years go by. I don’t expect that trend to abate in September. It’s unfortunate, but also undeniable.

That said, the topics of these passages have become, if anything, more standardized in recent years. There will definitely be a science passage, as there are 1.11 science passages per exam since December 2013. These have tended to oscillate in recent years between more pop-science topics like fish farms, nutrition, multiverses, and dowsing and more hard-science topics like the Big Bang, plate tectonics, anatomical changes ushered in by cooking, and brain scans. I have a feeling that this science passage will err on the harder side, likely about either neurology or cosmology.

There will almost certainly be a legal passage, as well. These passages about the law have appeared on 89% of recent exams. Lately, they’ve discussed either problems with witness testimony or problems with judicial decision making. Since the last few law-based passages have dealt with witness testimony, let’s say this one is going to be about judges.

Over the last few years, the comparative passage is almost always about either the arts or the law. I think the legal passage is going to a traditional, non-comparative passage, so I expect September’s comparative passage will discuss the arts, which shows up as a topic on 67% of all exams. We’ve seen recent comparative passages delve into music, opera, and literature appear; maybe this one will be about film?

Finally, I think there will be a culture/history passage to round things out. This is a topic that appears in 40% of all Reading Comp sections. I have a totally unsubstantiated feeling it’s going to be about unions.

Logic Games:

Finally, let’s talk Logic Games. The Logic Games trend you absolutely must be aware of is the increased importance of making scenarios. Games that benefit from the use of scenarios are abundant on the last few LSATs. On the June 2019, November 2018, and September 2018 exams, I thought making scenarios made all four of the games easier. On the June 2018 test, I thought this was true for all but one game. On December 2017, all games.

There’s a reason for this. On older games, scenarios were useful less frequently. On those games, you’d usually get a ton of rules that interacted with each other in pretty obvious ways. These rules which would severely constrain your players and lead to important deductions. To find these deductions, you really just had to pay attention to elements that showed up in more than one rule.

On almost all recent games, however, there are fewer rules and, consequently, fewer elements that show up in more than one rule. On their face, recent games appear much more open-ended and unconstrained than older games were. To figure out how these recent games work and to get a little head start on the questions, it’s become increasingly important to figure out a way to divide the game into a few different “scenarios” — the few ways the game could actually shake out. Making scenarios usually involves constraining one or more part of your set-up. When you do that, you’ll make it much easier to make the important deductions in each scenario — deductions that would have been incredibly difficult to make without resorting to scenarios.

I have every reason to think, then, that the September LSAT will continue this trend. I believe all four games on this test will be easier with scenarios So, as I predict these games, I don’t only want to predict what kinds of games they’ll be, but how you’ll probably make scenarios on that game.

I think the first game will be a 1 to 1 ordering game. 1.06 1 to 1 ordering games appear on each exam — this is the one game that shows up on literally every exam. For this 1 to 1 ordering game, watch out for a block (a rule that says players have to be ordered an exact distance apart from each other) that can only fit in a few places in your set-up. This is one of the most common ways to make scenarios on the LSAT. Pay also watch out fo an “or … but not both” ordering rule (e.g., “A must be ordered before B or after C, but not both”). Such rules have been a frequent cause for scenarios and (for those who are deft with scenarios) celebration alike, showing up most recently on the first game of the June 2019 LSAT or on the second game of the December 2017 exam.

For the second game, I’m seeing an unstable grouping game. These appear on roughly 65% of recent exams, and it’s been a few LSATs since they’ve shown up. I’m expecting you’ll be able to make scenarios based on a “must be together relationship” — a rule that claims that two players much be selected for the same group. Those rules are the bread-and-butter of scenarios on grouping games.

I’m going to guess the third game is going to be another grouping game — but this time, a “stable” one. That just means that, unlike their “unstable” counterparts, you’ll know exactly how many players need to join each group. These games show up on about 41% of all LSATs, and there have been a number of them in recent years. I think there’ll be a catch with this one, however: it’s going to be underbooked. There are going to be more spaces to fill than players available to fill them, so certain players will have to join more than one group.

I also think that there will be deductions or scenarios to be made, based on the limited number of players available to join each group. For instance, let’s say you have five players to fill a three-player group. Once you know that two of those five players can’t fill that group, you’ll know exactly who must join that group. This type of deduction has been really common recently — June 2019, Game 4; November 2018, Game 1; June 2018, Game 4; December 2017, Game 1; and June 2017, Game 4 all feature similar deductions. So make sure to keep track of who can’t join a group, and use that to determine who must join that group.

Finally, I think the fourth game will be a “tiered” ordering game. These games appear on over 70% of LSAT; when the appear, they’re usually the hardest game, and the hardest game is usually saved for last. Sounds intimidating … if only there was a blog post outlining all the common ways to make scenarios or deductions on these games. If it’s not one of those common deductions, I’m expecting scenarios could be made based on a constrained player — a player who can go in only a few places and whose placement will affect the surrounding slots. It also wouldn’t surprise me if the deductions within these scenarios were based on conditional logic (as in, “if B goes third, then C and D cannot proceed him …”).

————

Based on the data, this is our best guesses about what the September 2019 LSAT will look like. Of course, no one outside LSAC’s inner-circle of test writers and administrators knows exactly what the September test will hold. So it’s obviously best to walk into the exam on September 21st with a very well-rounded set of skills. But because there are a finite number of days before the test, we believe that focusing on these can calm your nerves and set you up for success.

Leave a Reply

Your email will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>