Explain the Discrepancy: Both Facts Are TRUE

By EMPOWERgmat on October 13, 2026

Here is a scene I watch play out every single week. A Test Taker reads a Critical Reasoning prompt. Fact one: a city put in brighter streetlights to cut nighttime crashes. Fact two: nighttime crashes went UP. Their brain lights up like a pinball machine. "One of these has to be wrong." So they go hunting through the answers for the one that proves the crash numbers were bogus. They find it. They click it. They are wrong.

That is not a reading problem. It is a stance problem. On a discrepancy question, you are NOT the judge deciding which fact to believe. Both facts are TRUE. Period. Your only job is to find the missing piece that lets them both be true at the same time. Lock in that one rule, and most of the wrong answers expose themselves before you finish reading them.

What you are actually looking at

Critical Reasoning (CR) questions give you a short paragraph, usually an argument or a set of facts, and then ask you to do something with it. On the GMAT Focus Edition, CR lives in the Verbal Reasoning section next to Reading Comprehension: 23 questions in 45 minutes, which works out to a little under 2 minutes per question on average. With Sentence Correction gone, every Verbal question is a reading and thinking question, which is exactly why your reading has to get sharper.

A discrepancy question (some books call it a paradox) gives you two facts that seem to clash. You will know one by its question: "Which of the following, if true, most helps to explain the discrepancy described above?" Sometimes it says "resolve the apparent paradox" or "explain the surprising result." Look at those two little words: IF TRUE. The test is telling you to accept the answer as fact, the same way you accept the paragraph as fact. Nothing on that screen is up for debate.

The one rule, and why it works

Think about what "explain" means in real life. Say your friend tells you, "I went to bed at 9, and I am exhausted." You do not say, "I bet you did not really go to bed at 9." You say, "Did the neighbor's dog bark all night?" You accept both facts and look for the thing that connects them.

That is the whole job. The right answer adds a NEW fact that makes the clash disappear. The wrong answers almost always do one of three things instead:

They attack or cast doubt on one of the two facts.
They support only one side, which makes the clash look even WORSE.
They talk about something next door to the topic but never touch the clash at all.

Now stop reading for a second. Before you look at the answer choices on your next discrepancy question, ask yourself one thing: what would have to be true for BOTH facts to hold? I will bet you can name two possible reasons in under 20 seconds. That prediction is your weapon. It turns five answer choices into one obvious match and four obvious misses.

Two puzzle pieces that clearly do not fit each other, joined neatly by a third piece that slides in between them. — The right answer is the missing piece, not a reason to throw one piece away.

Work one the right way

Try it

Two years ago, the city of Millbrook replaced the streetlights on its main roads with much brighter ones, hoping to reduce nighttime car accidents. In the year after the new lights went in, the number of nighttime accidents on those roads was higher than in the year before.

Which of the following, if true, most helps to explain the discrepancy described above?

Studies in several other cities found that brighter streetlights lowered the rate of nighttime accidents.
Some Millbrook residents doubt the accuracy of the city's accident records.
After the new lights went in, many drivers who used to avoid the main roads after dark began using them at night, so nighttime traffic on those roads rose sharply.
The new streetlights use less electricity than the old ones did.
Drivers on Millbrook's main roads report that the brighter lights make road signs easier to read at night.

Show the answer

Step 1: Notes on paper. Do not keep this in your head. Write it down:

F1: brighter lights (goal: fewer night crashes). TRUE.
F2: NUMBER of night crashes went UP. TRUE.
Q: how can both be true?

Step 2: Predict before you look. Notice the word "number." A count of crashes is not the same as a crash rate. Possible bridges: more cars on those roads at night, drivers going faster because they can see better, or a change in how crashes get counted. Any real fact like that would do it.

Step 3: Test the answers against BOTH facts.

(A) This backs up the idea that brighter lights should help. That makes the clash WORSE. Eliminate.

(B) Here is the trap. It feels like an explanation because it gives you permission to toss out fact two. But doubt is not a fact about the roads, and fact two is TRUE. This answer picks a side. Eliminate.

(C) More cars at night means more chances to crash. The lights can make each trip safer AND the total number of crashes can still rise. Both facts survive. Keep.

(D) Electricity has nothing to do with crashes. Eliminate.

(E) Easier to read signs supports the idea that the lights make driving safer. Same problem as (A). Eliminate.

Step 4: Prove (C) with numbers. Test values. Say the year before had 1,000 night trips and 1 crash for every 100 trips. That is 10 crashes. Now say the year after had 3,000 night trips, and the lights made driving safer: 1 crash for every 200 trips. That is 3,000 divided by 200, or 15 crashes. Each trip got SAFER, and the count still went from 10 to 15. No contradiction.

The answer is (C).

Look at what killed (B). Not some fancy logic. Just the rule. The moment you treat both facts as TRUE, any answer that asks you to doubt one of them is dead on arrival. And look at what killed (A) and (E): they explain why fact one makes sense, but they do nothing for fact two. An answer that helps only one side is not a bridge. It is a wall.

A brightly lit road at night packed with far more cars than the same road under dim lights beside it. — Safer trips, more trips. Both facts live together.

The mistakes I see every week

Picking a side. You decide one fact must be a mistake and go looking for proof.
INSTEAD: Write both facts on your notepad and put TRUE next to each one. It sounds silly. It works, because it physically stops you from voting one of them off the island.
Reading the answers before you have a prediction. Five answers written by professionals will push you around if you show up with no plan.
INSTEAD: Spend 15 seconds naming two possible bridges. Even if neither one shows up, you now know what a right answer LOOKS like.
Choosing an answer that fixes only one fact. It sounds relevant, so it feels right.
INSTEAD: Run every survivor through one test: does this make fact one AND fact two make sense together? If it only touches one side, it is gone.
Bringing in what you "know." You know brighter lights are safer, so you decide the crash count must be off. That is your outside knowledge overruling the prompt, the same mistake people make when they answer from what they know instead of the passage.
INSTEAD: The prompt is the whole world. If it says crashes went up, crashes went up.
Skating past the number words. "Number," "rate," "percent," "total" and "average" are where most of these clashes hide.
INSTEAD: Circle those words in your notes. A count can rise while a rate falls, and that gap is the explanation more often than you would think.

If this feels a lot like Evaluate the Argument questions, where you hunt for the question that would settle things, good. Same family. In both, you stop arguing with the prompt and start asking what missing fact would change the picture.

Your move

So here is the forced choice. What do you want MORE: the comfort of feeling like you caught the test in a lie, or the points? Because you only get one.

Today, pull 5 discrepancy questions you have already done. Cover the answer choices. For each one, write F1, F2 and two possible bridges on paper, in under a minute. THEN uncover the answers and see how many of the four wrong ones your prediction killed on sight. If you could not come up with bridges, that is not a sign you are bad at this. It is the exact drill you need to repeat until it is automatic.

Sources

GMAC: GMAT exam sections, timing and format