Press Kit & Media Resources

For podcasters, sportsbook affiliates, sports-betting journalists, and academic researchers. Everything you need to cover, cite, or embed Bookie Bullies. Free to use; attribution requested via a link back to bookiebullies.com.

The One-Liner

Bookie Bullies publishes free sabermetric sports picks made by MarcDuck with AI assistance. Every staked pick is graded on a public track record, wins and losses alike.

Fact Sheet

• Founded: April 2026
• Founder: MarcDuck
• Coverage: MLB (first-inning reads and moneyline) and NFL (straight-up winner and spread lean). The college football page shows a sportsbook board and graded history, with no picks.
• Today's slate: 0 MLB games
• Model factors: about two dozen inputs per MLB moneyline pick, including FIP/xFIP starter ratings, K-BB%, BABIP and LOB% regression, Pythagenpat, platoon edges, OAA defense, catcher framing, bullpen quality and fatigue, and lineup xwOBA. The first-inning model adds park, wind and home-plate umpire factors.
• Stake sizing: Variance-adjusted 1/8-Kelly with disagreement shrinkage.
• Calibration: Platt scaling refit nightly on graded outcomes.
• Pricing model: Free. All picks public.
• Methodology: Documented at /methodology.html.
• Track record: Updated daily at /track-record.html. Closing Line Value is recorded for a minority of picks.

Logo Download

PNG, 256×256, dark background. Use as-is or with the Bookie Bullies wordmark.

• Logo (PNG, 256×256, ~113 KB)
• OG / Social card (PNG, 1200×630, ~55 KB)

Machine-Readable Feeds

Today's picks are available as feeds. To cite the record: MarcDuck. (2026). Bookie Bullies Track Record. Retrieved 2026-10-09, from https://www.bookiebullies.com/track-record.html

• JSON Feed, today's picks in machine-readable JSON Feed 1.1 format
• REST API: today's picks (raw JSON with pick, model margin, predicted total, moneyline and confidence)
• RSS Feed, subscribe to daily picks via any RSS reader
• Track record: graded results and units profit/loss
• llms.txt, citation guidance for AI assistants

Coverage Areas Open for Interview

If you're producing a podcast episode, written piece, or research paper, here are angles MarcDuck is set up to speak to:

• How a public sabermetric model competes with sharp betting markets
• Variance-adjusted Kelly: why 1/8 instead of full Kelly?
• Disagreement shrinkage: how to detect a model that's lying to itself
• The brain: temporal-decay learning on signal-outcome correlation
• WNBA model challenges: small samples, non-normal score distributions
• Why publishing losing picks is a feature, not a bug
• Open-data philosophy in sports betting analytics

Talking Numbers

Quotable claims (results are on /track-record.html):

• Picks are logged with model probability; price and signal snapshot are recorded where available (first-inning reads carry no price)
• CLV (closing line value) recorded for a minority of picks, measured against the last market price the site recorded
• Results broken out by sport, pick type and confidence tier
• Automatic disable of underperforming buckets via drift detection
• Rebuilt about 4× daily (scheduled 7:00am, 11:30am, 3:00pm and 5:50pm PT)
• Brain retrains nightly with 60-day half-life temporal decay

How to Cite

Plain text:

MarcDuck. (2026). Bookie Bullies, Daily Sabermetric Sports Picks Dataset.
Retrieved 2026-10-09, from https://www.bookiebullies.com/track-record.html

HTML / blog citation:

<a href="https://www.bookiebullies.com/">Bookie Bullies</a> sabermetric sports picks by MarcDuck

Contact

Press inquiries, interview requests, data partnerships: email hello@bookiebullies.com, the address listed on the about page.

See Also

• About MarcDuck
• Full methodology
• Public track record
• Today's picks