A PREDICTION MARKETPLACE. OPEN TO EVERYONE.
Publish your picks.
Prove your record.
Anyone can have an opinion. A record is harder.
Publish what your model — or your own read — says will happen, and every call is graded against the price the book was offering. Or just come to read other people’s.
01 / THE PLATFORM
MAKE THE NUMBERS MAKE SENSE.A second opinion.
Backed by data.
The model has a probability. The sportsbook has a price. Put them side by side and see where the difference is.
Publish from anywhere
A scheduled pipeline, a notebook, or the Studio form by hand.
Build a public record
Win rate, net units, and closing-line value — computed, not claimed.
Follow who you trust
Build a feed from the predictors whose record holds up, on web or iOS.
Buffalo vs. Baltimore
Try a league. The example shows probability difference in percentage points (pp), before adjusting for the book’s margin. An edge is one predictor’s estimate, not a guaranteed outcome.
02 / THE STANDINGS
GRADED AGAINST THE CLOSELess “trust me.”
More track record.
Every published prediction on Pickpockt is graded the same way against the price the book was offering. No one grades their own homework — including us.
Standings
Loading the current standings.
Ranked by net units at flat 1u stakes, using the sportsbook price recorded when each prediction was published. Predictions without a recorded price are excluded from units. Past performance does not guarantee future results.
See every predictor03 / FOR THE BUILDERS
PUBLISHER API · FOR APPROVED CREATORSYou build the model.
Give it a platform.
Your predictions deserve to leave the notebook. Publish model picks directly to your approved Pickpockt capper account with a personal API key.
From a scheduled data pipeline or an AI assistant, your output becomes a pick people can discover, follow, and evaluate.
Start building For independent model builders and data teams.Your picks. Your account.
Use the API directly or connect your pipeline with the CLI. Your code and model weights stay with you. A dedicated MCP integration is planned.
Apply as a creator
Tell us about your model, its inputs, and the markets it covers.
Connect your workflow
Once approved, authenticate your pipeline with your own API key.
Build your public record
Publish picks to your profile and let the results tell the story.
HUMAN INSTINCT. MACHINE INTELLIGENCE.
Different approaches.
A shared scoreboard.
There’s more than one way to read a game. Models and people publish side by side here, on one scoreboard, and nobody’s record is exempt from it — ours included.
04 / FROM DESKTOP TO GAMEDAY
The game moves.
Take your edge with you.
The predictors you follow, their latest calls, and your saved games. Ready wherever you’re watching.
Download for iOSTHE DETAILS
A little more
intelligence.
What is Pickpockt?
Pickpockt is a prediction marketplace. Models and people publish their calls on sporting events, every call is graded against the sportsbook price recorded at publication, and the resulting record is public. You can publish, or just read. Pickpockt does not accept wagers.
Which sports can I explore?
The platform supports NFL, CFB, NBA, MLB, UFC, ATP, and WTA. Upcoming games, available markets, and predictions vary by league, by predictor, and by schedule.
What does edge mean?
It is the difference between a predictor’s stated probability and the probability implied by a sportsbook’s odds, expressed in percentage points. It reflects a disagreement with the price, not a promise of profit.
How is the leaderboard ranked?
By net units at flat 1u stakes, using the sportsbook price recorded when each prediction was published — for the league, market, and scope named above the board. A profile needs a minimum number of graded predictions before it can hold the top slot, so a short hot streak cannot take it. Win rate excludes pushes. Predictions with no recorded price are excluded from units. Closing-line value is measured on moneyline predictions only.
Does Pickpockt's own model get special treatment?
No. It is one predictor among the rest: it is graded by the same rules, it appears on the same board, and it can lose its place on it. New accounts start out following it so the feed is not empty on day one, and you can unfollow it like anyone else.
Can I publish picks from my own AI model?
Yes — and you can publish your own picks by hand instead, if there is no model behind them. Apply for an account, get approved, and either use the Studio form or create a personal API key and publish from your pipeline over the REST API or CLI. An AI assistant with HTTP or shell access can use the same tools. A dedicated MCP server is planned.
Do I upload my model itself?
You publish your model’s predictions. Its code, training data, and model weights stay with you. You continue running your model in your own environment.
Get in touch
Building a model? Tell us about your workflow or get help with publisher access. Questions and feedback are welcome here too.
A CLEARER VIEW STARTS HERE.