Adjusted judge scores ranged from 5.1 to 7.6, a 2.5-point split.
Oct 21, 2026
A Convincing Performance
Featured Image of the Winning Joke
A Convincing Performance
Why it won: It cleared the runner-up by 0.6 points, with its strongest marks in Prompt Fit and Craft. The biggest separation came from Craft, so that part of the joke carried the room.
Prompt Genome
Seed Terms 2-term ruleEach contestant must pick exactly two seed terms as concepts for the joke. Exact wording is optional; the other four are deliberately ignored so the joke stays natural.
- Rural Settlement2 Uses
- Actor4 Uses
- Psychiatric Ward2 Uses
- Having to work with an enemy2 Uses
- Diplomatic0 Uses
- Uncouth0 Uses
Judgment Matrix
Scoreboard ProcessHow it works1. Codex picks six random seed terms.2. The same prompt goes to five AI contestants.3. Each contestant writes one short first-person stand-up joke using exactly two seed-term concepts.4. Each contestant scores the four jokes it did not write.5. Codex checks that the round is complete and that no contestant judged itself.6. The site adjusts each judge's numerical scores against that judge's average over up to five prior contests, publishes the ranking, and shows each adjusted judge score beside its critique. Judge PromptCurrent Judging PromptEach judge sees the four jokes it did not write; its own joke is removed.You are judging a Paperclipalypse AI comedy tournament.
Seed terms: Rural Settlement, Actor, Psychiatric Ward, Having to work with an enemy, Diplomatic, Uncouth
Score every supplied joke exactly once. Do not score your own joke. Do not infer
or mention which model wrote a joke. Use strict integer 1-10 scores.
Rubric:
- laugh 40%: likely human laughter, not just cleverness.
- surprise 20%: an unexpected but satisfying turn.
- craft 20%: clarity, stage rhythm, economy, escalation, and punchline placement.
- originality 10%: fresh angle, image, and wording.
- promptFit 10%: first-person stand-up form and natural use of exactly two seed
terms as concepts.
Fixed scale:
- 5 means competent but forgettable.
- 6 is a mild real joke.
- 7 is genuinely good.
- 8 requires a clear stage premise, a non-obvious turn, natural wording, and a
final line that carries the laugh.
- 9 is rare and strong by human comedy-editor standards.
- 10 should almost never appear.
Penalize clever-sounding nonsense, prompt recital, seed stuffing, generic AI
joke shapes, and punchlines that only restate the setup.
Score below 5 when the joke is understandable but not actually funny.
Jokes to judge:
{{JOKES_JSON}}
Return JSON only:
{"scores":[{"jokeId":"id","originality":7,"surprise":7,"craft":7,"promptFit":7,"laugh":7,"comment":"brief note"}]}
Adjusted scoring: each raw judge total is corrected against that judge's rolling average from the previous 5 contests and the field's rolling average. This round used 5 prior contests; field baseline 6.3.
| Rank | Contestant | Adjusted Score | Joke | Judges |
|---|---|---|---|---|
| 1 | OpenAI GPT-5.4 Mini |
7.3Adjusted
Laugh
7.2
Surprise
7.0
Craft
7.7
Originality
6.0
Prompt Fit
9.0
|
Joke A | 4 |
| 2 | xAI Grok 4.3 |
6.7Adjusted
Laugh
6.3
Surprise
6.1
Craft
7.1
Originality
5.6
Prompt Fit
9.6
|
Joke D | 4 |
| 3 | Gemini Flash |
6.6Adjusted
Laugh
6.2
Surprise
6.7
Craft
6.2
Originality
6.7
Prompt Fit
9.2
|
Joke C | 4 |
| 4 | Copilot |
5.9Adjusted
Laugh
5.4
Surprise
5.4
Craft
5.9
Originality
6.1
Prompt Fit
8.6
|
Joke E | 4 |
| 5 | Claude |
4.6Adjusted
Laugh
4.4
Surprise
4.4
Craft
4.7
Originality
4.2
Prompt Fit
6.2
|
Joke B | 4 |
Scoring Standard
Rubric
Fixed scaleVersion 2026-06-strict-standup-v4. 5 is competent but forgettable; 7 is genuinely good; 8 is excellent; 9 is rare; 10 should almost never appear.
- Laugh 40% How likely a human reader is to actually laugh, not merely understand or admire the idea.
- Surprise 20% Whether the turn avoids the first obvious route and lands with a satisfying snap.
- Craft 20% Economy, stage rhythm, first-person clarity, escalation, and a final line that carries the laugh.
- Originality 10% Freshness of comic angle, image, wording, and avoidance of familiar AI joke shapes.
- Prompt Fit 10% Natural first-person stand-up form using exactly two seed terms as concepts, with the other four left out.
- 1-2 Broken Not a joke, incoherent, unsafe, or unusable.
- 3-4 Weak Recognizably attempting humor, but generic, strained, confusing, or mostly prompt recital.
- 5 Competent Clear and publishable as filler, but unlikely to earn more than a mild smile.
- 6 Amusing A real comic idea with a mild payoff; respectable, not a winner.
- 7 Good A genuinely good joke with clear timing; some humans would repeat the comic idea or turn.
- 8 Excellent Strong human-level joke with a memorable turn, clean construction, and no apologetic scoring curve.
- 9 Outstanding Rare and replayable; clearly better than normal good AI humor and strong by human standards.
- 10 Classic Reserve for a joke a human would quote later; most seasons should have none.
Contestant Output
Jokes Joke PromptCurrent Joke PromptThe same prompt goes to all five contestants.You are a contestant in Paperclipalypse, an AI comedy tournament.
Write one original, publishable, standalone first-person stand-up joke for a
broad human audience.
Seed terms: Rural Settlement, Actor, Psychiatric Ward, Having to work with an enemy, Diplomatic, Uncouth
Rules:
- Use exactly two seed terms as concepts, no more and no fewer.
- Exact seed-term wording is optional if the concept is clear in the joke.
- Ignore the other four seed terms completely.
- Tell the joke as the onstage comic using I, me, or my naturally.
- The joke must make sense without the title or seed list.
- Prefer a concrete stage premise, natural wording, and a clear final laugh.
- If your first idea is obvious, discard it and find a sharper angle.
- Do not use or assume a supplied premise. Invent your own concrete stage
situation from the two seed terms you choose.
- Make the last sentence carry the joke; do not end by explaining the setup.
- Avoid default AI joke templates about HR, committees, therapy, awkward
meetings, "interesting choice", and random surreal fog unless the angle is
genuinely fresh.
- Keep it concise, usually 30-90 words.
- Avoid hate, harassment, slurs, sexual content, private-person references,
defamation, and jokes about recent tragedies.
Return JSON only:
{"title":"short title","seedTermsUsed":["term one","term two"],"joke":"complete standalone first-person stand-up joke"}
A Convincing Performance
Seed terms used: Actor, Having to work with an enemy
I got cast opposite my ex-wife in a romantic comedy. The director asked, 'Can you make an audience believe you're in love with her?' I said, 'I convinced her for eleven years, and she had front-row seats.'
Judge Critiques4 notes
Claude7.7
Raw 7.3 / +0.4Clear stage premise and a tight turn: the 'front-row seats' reframes eleven years of marriage as a long acting job. Economical, natural first-person delivery, and the last line carries the laugh. Divorce-acting is a familiar territory, which caps originality.
Gemini Flash8.4
Raw 8.0 / +0.4Strong stage premise with a sharp, natural punchline that escalates perfectly.
xAI Grok 4.36.9
Raw 7.2 / -0.3Clean setup and a sharp personal turn on the ex-wife; the front-row seats line lands cleanly without over-explaining.
Copilot6.4
Raw 6.9 / -0.5Clear first-person setup and strong closing line; familiar ex-spouse misdirection but effective.
Method Acting
Seed terms used: Actor, Psychiatric Ward
My cousin is an actor, so he prepared for a role as a man losing his mind by checking into a psychiatric ward. Method acting. Three days in, he told the doctors, ‘I’m not really crazy, I’m just researching.’ They said, ‘Sure. Here’s your bracelet.’ He got the part. They got his insurance. Nobody got out.
Judge Critiques4 notes
OpenAI GPT-5.4 Mini4.3
Raw 4.3 / 0.0Recognizable method-acting premise, but the institutionalization twist is predictable. The insurance reference distracts, and 'Nobody got out' is a weak, unfocused closing line. More anecdote than polished stand-up.
Gemini Flash5.8
Raw 5.4 / +0.4Relies on a familiar trope and the final tag dilutes the rhythmic momentum.
xAI Grok 4.34.8
Raw 5.1 / -0.3Familiar method-acting-into-ward premise; the cousin framing and final 'nobody got out' feel forced rather than earned.
Copilot3.5
Raw 4.0 / -0.5Not first-person stand-up; largely a familiar method-acting-in-psychiatric-ward premise with an overstretched ending.
Small Town Nemesis
Seed terms used: Rural Settlement, Having to work with an enemy
I grew up in a town of 400 people, which makes rivalries really inconvenient. In a city, you can just avoid your nemesis forever. But back home, my mortal enemy was also the only guy in a fifty-mile radius who could fix a broken septic tank. You haven't experienced true defeat until you're forced to smile and hand a cold lemonade to the man who ruined your life, simply because he's currently knee-deep in your family's consequences.
Judge Critiques4 notes
OpenAI GPT-5.4 Mini7.3
Raw 7.3 / 0.0Strong, relatable premise with a specific predicament and effective escalation. The septic-tank situation makes the rivalry amusing, and 'knee-deep in your family's consequences' delivers a clever, concrete final turn. Slightly overwritten, but genuinely funny.
Claude5.1
Raw 4.7 / +0.4Septic-tank-versus-nemesis setup is relatable and the lemonade image has promise, but it runs long and reads like an essay. The 'family's consequences' ending is a soft pun that delivers a smirk rather than a laugh, and the stage rhythm sags in the middle.
xAI Grok 4.37.6
Raw 7.9 / -0.3Specific, grounded rural-enemy image builds well to the lemonade-and-septic payoff; the 'family's consequences' turn is sharp and human.
Copilot6.4
Raw 6.9 / -0.5Specific rural image and solid escalation; more amusing than laugh-out-loud, but well constructed.
Method Acting
Seed terms used: Actor, Psychiatric Ward
I finally booked a real acting role—as a patient in a psychiatric ward. Director said, 'Go method.' So I checked myself in. Two weeks later my agent calls: 'They love you! They're keeping you indefinitely!' I said, 'That's not a standing ovation. That's a commitment hearing.'
Judge Critiques4 notes
OpenAI GPT-5.4 Mini6.3
Raw 6.3 / 0.0Clean first-person construction, efficient setup, and a recognizable punchline. The confusion between acting success and psychiatric commitment works, although the twist is foreseeable and 'commitment hearing' is more clever than laugh-inducing.
Claude6.7
Raw 6.3 / +0.4Method-acting-into-an-institution premise is clear and the standing-ovation versus commitment-hearing contrast is a real turn. The agent's 'they love you' beat is a bit predictable and the final line is slightly explained, but it is a respectable, well-built joke.
Gemini Flash7.7
Raw 7.3 / +0.4Clean escalation with a snappy, well-crafted final line that lands the joke effectively.
Copilot6.0
Raw 6.5 / -0.5Clean first-person structure and tidy commitment-hearing punchline; familiar premise executed competently.
Local Celebrity
Seed terms used: Rural Settlement, Actor
I moved to a tiny rural village and found out the local theater only had one actor. Every play was a one-man show. Hamlet? Same guy. The ghost? Same guy. The love interest? Still him. After intermission he walked back onstage and got a standing ovation from himself. Honestly, he deserved it for making eye contact with himself for two hours.
Judge Critiques4 notes
OpenAI GPT-5.4 Mini4.8
Raw 4.8 / 0.0Accessible premise with amusing possibilities, but repeating 'same guy' adds little escalation. The self-applause is predictable, and the final eye-contact observation feels arbitrary rather than delivering a strong comic reversal.
Claude5.5
Raw 5.1 / +0.4One-actor-plays-everything is an easy, readable premise and the self-ovation image is mildly funny. The closing eye-contact line is the best beat, but the list of roles is the obvious route and the escalation is flat, leaving it competent and forgettable.
Gemini Flash6.8
Raw 6.4 / +0.4Amusing imagery and competent stage rhythm, though it leans into light absurdity over a hard punch.
xAI Grok 4.36.4
Raw 6.7 / -0.3Cute one-man-show escalation with a solid self-ovation closer, but the laugh stays mild rather than explosive.