If you’ve just started a Blueprint course — or if you otherwise have a reasonable study plan that includes learning a sound strategic approach for each Logical Reasoning question type — you’ve probably just encountered the Soft Must Be True question. You know, the ones that say stuff like “If the above statements are true, which of the following is most strongly supported?”

We love these questions. In a truly nerdy hat-tip, we named this blog after these questions. But don’t let our love or the “Soft” modifier throw you off — these questions can be vexing for a ton of LSAT test takers. And LSAC’s malevolent psychometricians seem to know this, having taken quite the liking to these Soft Must Be True questions on recent exams. In fact, on the most recently published LSAT, 20% of the scored Logical Reasoning questions were Soft Must Be True questions, making your success on these supposedly “soft” questions a “hard” reality to face.

So this post is going to give you some advice to firm up your approach to these Soft Must Be True questions. All it takes a hardened resolve and solid practice, and these can become, more or less, Softball Must Be True questions. Here’s Part 1 of the FAQs for Soft MBTs. Check back later this week for Part 2.

“What is a Soft Must Be True question?”

Let’s start by backing up for a second. What are Soft Must Be True questions? And how do they relate to the regular old Must Be True questions?

Soft Must Be True questions, just like basic Must Be True questions, require you to make a deduction from a set of facts provided in the stimulus. But when you make a deduction for Must Be True question, the deduction must be 100%, without-a-doubt true. And that’s a “hard” burden of proof to bear. Any sliver of a chance that a deduction is not true is enough to render an answer choice incorrect. This is why, incidentally, most correct answers for Must Be True questions are derived from either making a transitive deduction from conditional statements or from simply summarizing the information in the stimulus. These are tried and true ways to make an unassailably true deduction.

On the other hand, we have a “softer” burden of proof on Soft Must Be True questions. These questions ask for the “most supported” or “best illustrated” inference. We don’t have to be 100% certain our deduction is true. We have a little more leeway to make inferences or offer explanations for surprising results. That said, this is still a pretty high burden. We have to be like 90% sure our deduction is true — we can only entertain a small amount of doubt to our deduction’s truth.

Think of it this way: Let’s say your friend regularly drinks margaritas made with (fancy, expensive) Patrón tequila. But when this friend requested that you buy him this drink at the bar, you instead bought him a margarita made with (cheap, plastic-bottled) well tequila. And then your friend drank and enjoyed this cheaper facsimile of a cocktail, never mentioning that it didn’t taste like the fancy margarita he typically orders. What would be the most supported inference we could draw from this? Probably that your friend can’t taste the difference between a margarita made with Patrón and a margarita made with cheap tequila.

Now, we can’t be certain that this is true — there’s a remote possibility that your friend tasted the margarita, and beyond tasting the lime juice, and the Cointreau, and the agave syrup, and the salted rim, your mans tasted the tequila and wasn’t getting the notes of like oak barrel and lemon zest that he associates with Patrón tequila. He just didn’t say anything to be polite. In other words, this inference wouldn’t meet the burden of proof for Must Be True questions. But because this possibility is very unlikely — especially if your friend isn’t a supertaster — it’s going to be good enough for Soft Must Be True questions.

“So where’s the line on what inferences we can draw?”

Let’s take the above example about the margaritas. We need to be careful not to go too far when making a deduction. We can’t engage in any unfounded speculation. We couldn’t say, for instance, that your friend can’t taste at all, since we only have evidence that he can’t seem to distinguish between two types of margaritas and we don’t have any evidence relating to his ability to taste other foods or drinks. Nor could we say that your friend has below-average taste receptors, since we don’t know if anyone else could distinguish between these two types of margaritas. Both of these inferences require us to speculate about information not included in the example.

We also couldn’t say that that there’s no difference between Patrón and cheap tequila. While one person can’t seem to notice the difference when these tequilas are used as an ingredient in a margarita, to say there’s no difference between the two is going way overboard. Maybe most people, unlike your friend, can tell the difference between the two when used in a margarita. Maybe when the two tequilas are consumed straight up, there’s a huge difference.

There are basically three types of inferences we can never make on a Soft Must Be True question. And incorrect answer choices will almost always make at least one of these mistakes.

The first is an inference that deals with outside information. On a Soft Must Be True question, we can only make inferences about the info provided in the stimulus. Any answer choice that brings in new ideas, facts, or hypotheticals should be eliminated.

The second is an inference that is stronger than its support. You can’t make a conclusion that is stronger than the evidence used to support it. If I said, “Most BMW drivers I encounter drive dangerously and selfishly,” we certainly couldn’t conclude, “All BMW drivers drive dangerously and selfishly.” Answer choices do things like this all the time on Soft Must Be True questions, and should be eliminated with impunity.

The third is an inference that proposes a relationship between two things in the stimulus that the stimulus doesn’t support. You’ll commonly see answer choices that propose causal, comparative, or proportional relationships, and these answer choices almost always incorrect.

“OK. How do I find the right answer on these?”

This is where most test takers get into trouble on Soft Must Be True questions. At Blueprint, we’re very fond of anticipating the right answer to questions before looking at the answer choices. Ideally, you don’t want to rely on “discovering” the right answer in the answer choices. The answer choices are not your friends. They are laden with all kinds of tricks and traps designed to entice a certain percentage of test takers into selecting the wrong answer. To avoid these “sucker” choices, we recommend anticipating what the right answer will look like, and looking for that in the answer choices.

A lot of the early concepts we cover in the Blueprint course are very involved methods of anticipating the right answer. These frequently require students to put pencil to paper and actually write things out. Like diagramming conditional statements. Or writing out the primary structure for Reading Comprehension.

There’s not really an analog to these processes on Soft Must Be True questions. Few Soft Must Be True questions involve diagramming, so there isn’t a pressing need to press pencil to paper on these questions. Consequently, many students fail to do any anticipation on Soft Must Be True questions, rushing headlong into the answer choices like lemmings off a cliff.

However, there are several steps you can do to help anticipate the right answer to Soft Must Be True questions. The first thing you should always do is summarize the stimulus to yourself. If it’s a shorter stimulus, you can usually make a short, one-sentence summary. If it’s longer, you may have to make a few bullet points. Make sure you’re not repeating the frequently confusing (obfuscating, abstruse, recherché) language the stimulus uses. Use your own words. Doing so will ensure that you understand what the stimulus is about and — just as importantly — what the stimulus isn’t about. This will help you avoid answer choices that bring in outside information.

Then, you want to look for strong pieces of information in the stimulus. Strong pieces of information are great for making deductions. In order to infer something about a subject, you need to know a lot about that the subject. If you can find a really strong statement, you want to think about how that strong statement relates to the other statements in the stimulus. By thinking about how that strong statement affects the other statements, you should be able to anticipate a deduction that will appear in the answer choices. Even if you can’t anticipate the exact deduction they’re looking for, you at least know the right answer will probably relate to that strong statement.

For instance, take this Soft Must Be True question …

“Modern science is built on the process of posing hypotheses and testing them against observations– in essence, attempting to show that the hypotheses are incorrect. Nothing brings more recognition than overthrowing conventional wisdom. It is accordingly unsurprising that some scientists are skeptical of the widely accepted predictions of global warming. What is instead remarkable is that with hundreds of researchers striving to make breakthroughs in climatology, very few find evidence that global warming is unlikely.”

First, we should all take a moment to appreciate an instance of anti-scientist bias on the LSAT. The former philosophy majors who write this exam love to criticize the more pragmatic field of scientific research, so here’s one of many questions describing how scientists are craven and fame-hungry and overall not that cool. Cool.

Then, we should of course summarize this information. Basically, we can whittle this stimulus down to: “Scientists try to disprove hypotheses. Disproving conventional wisdom brings the most recognition, so scientists try to disprove hypotheses relating to global warming, although very few can find evidence to disprove those.” Or, if you are really into the whole brevity thing, “Very few scientists who try to undermine global warming hypotheses find evidence to do so.”

Next, we should look to a strong piece of information to try to make a deduction. And the strongest statement made here is “Nothing brings more recognition than overthrowing conventional wisdom.” That is crazy strong. Nothing could possibly bring more recognition to a scientist that overthrowing conventional wisdom. So if we had two scientists — one who cured cancer, and one who overthrew the conventional wisdom that the common cold is caused by being chilly — we would apparently recognize the latter’s accomplishment more. That’s a strong take, and we need to pay attention to that.

Finally, we should think about how that strong piece of information relates to the rest of the stimulus. Well, if overthrowing conventional wisdom brings a ton of recognition to scientists, and if global warming is “widely accepted,” scientists have a pretty big incentive to overthrow the conventional wisdom relating to global warming. Doing so would bring a scientist a ton of recognition, presumably ensuring him or her a lucrative second career as a talking head on every Fox News show and a representative of fossil fuel companies’ interests. Even if we can’t make that specific deduction, we should be looking for an answer choice that relates to the benefit scientists accrue by overthrowing conventional wisdom.

Either way, that should lead us to the answer choice that says, “Most researchers in climatology have substantial motive to find evidence that would discredit the global warming hypothesis.” This is totally consistent with the deduction that we anticipated and, importantly, it’s not stronger than anything we summarized and doesn’t bring in information that wasn’t included in our summary. Voila, it is the “most strongly supported” inference.

So whenever we get a very strong statement in the stimulus of a Soft Must Be True question, we should take these steps. But what if there isn’t a strong statement? … Stay tuned for Part 2 later this week