Skip to content
The format

Turn skill into
a live event.

Create the challenge. Let your audience call the outcome. Perform it live. Let PLLAY verify what happened.

Anything provable can become playable.

Creator ChallengesPredict whether the creator can do it. Not yet open to creators.

Start to finish

Five steps, one broadcast.

This is what the creator does. Underneath it there is nothing new to build — a challenge is a pool opened on the creator's own run, which is the same engine that already runs on gameplay.

01

Create it

Choose a template or describe the challenge to PLLAY AI.

02

Open it

PLLAY creates the live participation experience.

03

Perform it

Do the challenge live.

04

Prove it

AI Referee or another supported verification method determines what happened.

05

Settle it

PLLAY resolves the prediction and updates participants.

A clean read settles on its own. Settle-or-refund, never a guess. An unclear read comes back to the creator to confirm rather than being guessed at.

One challenge, end to end

What the audience sees.

The walkthrough below runs a basketball challenge because that is the clearest example of the shape. It is an illustration: nothing verifies basketball today, and the demo says so in every frame.

Concept demoAn illustration of the format — not a recorded stream, and not a capability. PLLAY has no shot-detection model; outcome reading runs on video-game scoreboards today.
  1. PREDICTIONS OPEN
  2. PREDICTIONS CLOSING
  3. LOCKED
  4. CHALLENGE LIVE
  5. AI REFEREE
  6. VERIFIED
  7. RESULT

Prediction state — SETTLED

20 FREE THROWS / 90 SECONDS

Final

21 valid makes

Target 20 — verified ✓

Yes was the call. Participants update.

Concept demo — a walkthrough of the format, not a recorded stream. Basketball challenges are on the roadmap and nothing verifies them today. Credits, not dollars; PLLAY is pre-launch and cash rails are off.
Available today

One category is verified.

Outcome reading is per-title, and every title PLLAY supports is a video game. The reader looks at an on-screen scoreboard — a score region, a clock, a final banner. That is the whole of what exists.

GamingVerified

Can they close out the round without going down?

Per-title adapters read the on-screen scoreboard. Benchmarked on recorded video; it has never run on a live stream.

The supported titles are listed at Games, and every bench trial is published at Vision AI.

Roadmap — nothing here is built

Where the same mechanism could point.

These are examples, not features. Each one names the specific model that does not exist yet, because a roadmap entry without a ceiling is a promise. None of them has a date.

  • FitnessRoadmap

    Forty pushups in sixty seconds?

    No rep-counting model exists. Nothing in the adapter catalog watches a body.

  • BasketballRoadmap

    Ten free throws in a row?

    No shot-detection model. The catalog reads game scoreboards, not real courts.

  • FootballRoadmap

    Can they hit the crossbar from the forty?

    No ball-tracking model. The football adapters read a video-game broadcast overlay.

  • SoccerRoadmap

    Five penalties, five corners?

    No goal-detection model, and no way to read a physical pitch.

  • GolfRoadmap

    Inside ten feet from a hundred yards?

    No ball-flight or distance model of any kind.

  • BaseballRoadmap

    Can they take five straight pitches deep?

    No pitch or contact model. The baseball adapters read a video-game scoreboard.

  • RunningRoadmap

    A mile under six minutes?

    No timing or GPS ingest. Nothing reads a wearable or a track.

  • PrecisionRoadmap

    Ten darts, ten triple-twenties?

    No target or landing-position model.

  • ReactionRoadmap

    Can they beat their own reaction time?

    No timing harness a creator could run on a broadcast.

  • MusicRoadmap

    Can they play it clean at full tempo?

    No audio-analysis model. Music sets today carry audience formats, not judged performances.

  • FoodRoadmap

    Twelve wings before the timer?

    No counting or consumption model.

  • TriviaRoadmap

    Fifteen straight without a miss?

    Resolves against a stored answer rather than a video read, but no settlement path for it is built.

12 of 13 categories are roadmap. No shipped-on date until the verification is real.

Three shapes

Solo, head-to-head, and the condition.

Solo

Challenge yourself.

The creator sets the condition on their own run and the audience calls whether it holds.

Head-to-head

Challenge another creator.

Two creators, one condition. Each side's audience calls its own.

Eventually anything

If it can be objectively measured and verified, PLLAY can potentially make it playable.

That is a condition, not a schedule. Today it is met by one category.

Head-to-head here is two creators on the same condition, each audience calling its own side. For two creators competing directly against each other, see Creator Battles.

Next

Run one on your next broadcast.

Gaming challenges run on the shipped engine today. Applying is how a creator gets access; there is no self-serve challenge tool yet.