How to Test New Jokes (and Know If They're Actually Funny)

A new joke isn’t funny because it makes you laugh. It isn’t funny because the friend you tested it on laughed. It isn’t even funny because the open mic crowd laughed once. A new joke is funny when it lands at three different open mics in front of three different audiences with no other variable changing.

One laugh is luck. Three is signal. Comics who skip the testing phase ship jokes that only worked in a specific room and wonder why those jokes die at every booked spot after.

The 3-room rule

A joke needs to survive three different open mics before you trust it. Different rooms have different audience compositions, different acoustics, different time slots, different energy. A joke that kills at your home Tuesday might die at a Saturday late show in a different neighborhood.

The 3-room rule isn’t scientific — it’s a sanity check. If a joke works at three independent rooms, you’ve probably written something real. If it works at one, you’ve probably written something that worked because the room was warm.

Three tests new material has to pass

Test 1: Can you say it out loud without cringing?

Read the joke out loud, alone, in your living room. If you cringe at any line — the setup feels long, the punchline feels lazy, the premise feels borrowed — that line is the problem. Cut or rewrite before you take it on stage.

This sounds basic but it’s skipped constantly. Most comics show up to the open mic with material they’ve never said out loud. The room hears the same hesitation you would have heard in your kitchen.

Test 2: Will it survive 3 different open mics?

Run the joke at three rooms in three different conditions. Don’t put it in the same slot twice in a row. The position in the set matters — a joke that bombs in slot 2 might kill in slot 6, because by slot 6 the audience knows what you sound like.

Test 3: Does it work on a cold crowd?

Open mics are warm rooms. Other comics laugh at almost anything because they understand the form. The real test is a paying audience that didn’t come to support comedy — they came because their friend dragged them. If the joke holds up there, it holds up.

You won’t get a cold-crowd test until you’re booking real spots. Until then, the 3-room rule is your best approximation.

Why your first laugh isn’t real

The first laugh on a new joke is almost always corrupted by:

• Host laugh. The host running the room laughs at most things to keep the energy up. It’s a courtesy laugh, not signal.
• Comic laugh. Other comics in the back laugh at jokes they recognize as well-constructed, not necessarily at jokes that landed for the audience.
• Room laugh. Some rooms are just generous. The Tuesday at your home bar will laugh at almost anything.

Watch (or listen to) the recording. Was it a laugh from the audience, or from the back? The audience laugh is the one that matters; the back-of-room laugh is data about the bit’s craft, not its effect.

Where to slot new material

Test new jokes in the middle of your set, sandwiched between your strongest material. The opener and closer carry too much risk to use as test slots:

• Lose the opener and you lose the room’s attention for the rest of the set.
• Lose the closer and you lose the chance of getting booked again.

The middle is where weakness can hide. If a new joke dies in the middle, your strong closer can still save the set. If it kills, you’ve learned something without risking anything.

Track the data, not your feelings

After every set, mark each joke with a simple system:

• ✓ — killed
• — — landed (got the laugh you expected)
• ✗ — bombed
• ? — ambiguous (laughter but not where you wrote it)

Do it within 5 minutes of leaving the stage, before the bias of memory sets in. Comics consistently misremember which jokes worked — the brain rewrites the set into a flatter version of itself once you’re not standing on stage anymore.

Stand-Up Writer’s per-joke ratings work this way: tap a star next to each joke after the set, and the data accumulates across rooms. After 5–10 sets, patterns emerge that you can’t see from a single performance.

When to give up on a joke

If a joke bombs three times in three different rooms, the joke is the problem — not the rooms. At that point you have two options:

1. Rewrite from scratch if the premise still excites you. The premise is the part worth fighting for; the setup and punchline can always be redone.
2. Kill it if you can’t remember why you wrote it in the first place. Holding onto a dead joke costs you the slot it’s taking up in your set.

The third option is to leave it on the bench — not in your set, but not deleted. Some jokes need to age. The premise that doesn’t work for you at 24 might land perfectly at 28.

When a joke is “done”

A joke is done when you’ve tweaked it three times in a row and the laugh hasn’t gotten bigger. At that point, it’s as good as it’s going to get in this version. Move on. Trying to optimize past the ceiling burns time you could spend writing the next joke.

For more on cutting fat from individual jokes, see setup, punchline, tag — joke structure 101. For figuring out which jokes deserve the bench versus the bin, see what to do when your set bombs.