Skip to content

Step 4: Eliminate, defer, or Select

Read exactly what the answer says. Don't add or subtract anything. Understand what it means.

Read all the answers. But you don't have to read all of every answer. As soon as they're wrong, they're gone.

When you find an answer you like, you can be extra skeptical of the rest. (Maybe you could be that skeptical always?)

When you get turned around by the answers (this will happen), take a second to reset. Remind yourself of your job and your prediction.

Generally, on Reasoning, you should eliminate first, then defer, and select.

Eliminate answers you know are wrong

In puzzle practice you have time to explain what makes the bad answers wrong. Avoid giving vague reasons like “it feels out of scope.” Find a specific word or phrase that poisons the answer. Articulating reasons will help improve your intuitions and speed.

During race practice, you will not always take the time articulate a clear reason.

Defer on weirdos

You defer in order to not get stuck.

“Huh” and “Maybe” are both great reasons to defer on an answer.

A confusing answer might be confused. It's not worth figuring out.

Or maybe you don't get it. In which case, come back later, if you need to.

A defer can be positive or negative. Notice how you feel about it. Trust and verify those feelings when you revisit it later.

Select answers you like

Not merely "like" but "like like." You're marrying this answer. Choosing it will bias you against the others. This will help you speed through them. That's a good thing, mostly. But you might speed past the actually correct one.

If you don't love it, notice that you like it and defer. When there's nothing better you can choose it and move on.

Skill practice suggestion: Rank the answers

Give each answer a rank from 1-6.

  1. Love it
  2. Like it
  3. Unsure-Positive
  4. Unsure-Negative
  5. Dislike it
  6. Hate it

Then pick the highest ranking answer.

As you practice, try to figure out what it would take to have more ½s and ⅚s rather than ¾s:

  • Do you just need to be more confident?
  • Do you need more useful (either more flexible or more specific) predictions?
  • Do you need to read the answers more generously? Or do you need to be more skeptical?