A compass, not the map.
BTB doesn't sell picks and doesn't claim to beat the books. It gives you direction: how far each player can be trusted, what tonight's matchup adds or takes away, and where our number and the book's part ways. You make the call, faster and with the reasons in front of you.
BTB Trust: how far a player can be trusted
BTB Trust is a season read of the skills that repeat, shown as a percentile from 1 to 99.
- Pitchers: built from missing bats, limiting free passes and keeping the ball in the park. Ranked among starters with 50+ innings and relievers with 20+. Checked on 2026: a pitcher's Trust in the first half lined up with his second-half runs allowed about as well as strikeout-minus-walk rate, the standard public measure.
- Hitters: built from quality of contact and swing decisions, ranked among hitters with 150+ plate appearances. Strikeout and walk rates are not separate inputs (xwOBA already counts every plate appearance): on their own, for hitters, they carried no signal about what came next.
- Bullpens: each pen is rated by the arms it has actually used over the last 30 days, each on his season to date (postseason included, whatever his innings), weighted by the batters each one faced, and ranked among the pens with 40+ relief batters faced in that window (all 30 in season). It sits beside the bullpen ranking rather than replacing it.
A season read of how far the player can be trusted: it describes the player, not tonight's line or projection. It is context for your decision, not a pick.
BTB Stuff and BTB Value: the pitches themselves
BTB Stuff is our pitch model's grade of a pitcher's arsenal over the last 90 days, where 100 is league average. The model was fitted on every pitch from 2021 to 2025 and is graded against what the pitches went on to produce. BTB Value grades the physics of his regular-season pitches on what they should produce, so the two can differ when his last 90 days and his season do: 5 is the 2025 league average and each point is one standard deviation.
The matchup score: tonight, in runs
For every hitter, the matchup score is how many runs tonight's setup adds to or takes away from his night against a league-average starter, bullpen and park, split into four parts: the starter, his pitch mix, the bullpen and the park. Starters get the mirror image: what tonight's lineup and park do to their night. It ranks the slate's best and toughest matchups, and it is logged before first pitch so it can be graded.
Our number beside the book's
Every hitter leads with our chance of a hit, two total bases and a home run beside the book's no-vig price, marked above, below or in line (a gap of three points or more). Every starter's strikeout projection sits beside his line. It is a description, never a play.
The record: why you can check us
Every prediction is frozen at first pitch and graded in public. The record shows how it went, the model cards give each market one honest word (Validated, Promising, Insufficient sample, Underperforming baseline or Research only), and Research shows the studies, including the ones that failed.
Who uses it
- Analysts: strikeout, hit, total-base and home-run projections with the matchup behind each, and a record frozen at first pitch.
- Fantasy players: the matchup score in runs, confirmed and projected lineups, pitch-mix and bullpen reads, ranked for the slate.
- Writers and researchers: model cards that say what is proven and what isn't, the methods, and research notes that include the tests that failed.
- Fans: leaderboards, a page for every pitcher and hitter, season stats and plain-English reasons.
Member pages are BTB Pro: $15 a month or $80 a season, or $35 a month for this site, the NFL site and the NBA site. See pricing.