The price columnis no longer PENDING.
Live moneyline, run line and total prices for every game on the slate, de-vigged into consensus fair numbers, then inverted against your +6% EV gate so each market shows the exact model probability it would take to become an official play. Starter projections and park factors now drive a run distribution behind the moneyline, run line and total columns — logged as a provisional lean, never promoted to an official play on its own.
Every market, priced against the gate
Fair is the average no-vig probability across the books quoting that market — not a house guess. Model prob needed is the break-even probability at the best available price plus your +6% margin, rounded against the bettor. Drop your event probability next to it: higher clears, lower passes.
| ET | Game | Market | Selection | Best | Book | Fair % | Model % | EV price | EV model | Needed | Conv | Status |
|---|
Convergence = 100 − (spread between book no-vig probabilities, in points × 20). 100 means the books agree to the decimal; low scores flag a market still finding its number.
Full board with probable starters
| ET | Game | Probable starters | Venue | Runs env | HR env | Model tot | Best ML | Fair ML | Total |
|---|
Projected starters, park-adjusted context
| Game | Tm | Pitcher | Hand | Opp | IP | BF | R | H | HR | K | BB | Runs env | HR env |
|---|
Runs env and HR env are the park-and-weather run environment for that game, not an adjustment applied on top of the projection. A starter with no projection published shows Pending rather than a substituted league average.
Ten hitters most likely to leave the yard
Ranked by the published probability that the hitter records at least one home run, taken from the day's BallparkPal Hit a HR export and read by hand from that export rather than pulled from a feed — the number is reproduced exactly as published, never rounded into a range or re-derived. The best price column is the top posted Over 0.5 price among the books screened, and shows Pending for any hitter no book has put up yet, which is normal the night before. Edge is the published probability minus the price's implied probability and only appears where a real price exists. Every row is Context — a home-run row is not an official play and does not enter the results log.
| # | Hitter | Tm | Game | HR chance | Fair | Best price | Edge | Park HR env | Opp starter | HR allowed |
|---|
Fair price is the export's own fair American number for that probability, not a sportsbook quote. Park HR env is the park-and-weather home-run environment for that game from the day's park export; a park export whose own date does not match the slate is dropped rather than reused. HR allowed per start is the opposing starter's projected home runs allowed, shown only when that starter's surname matches the projection row — otherwise Pending.
What the sportsbooks themselves imply
Ranked by the market's own de-vigged probability of at least one home run, taken from the only sportsbook posting both sides of the 0.5 line. Books that post the Over alone cannot be stripped of hold, so they contribute a price comparison and never a fair number. This is a separate view from the top-ten board above and is never merged with it: one number is the market's, the other is a model's. Every row here is Context — home-run rows are not official plays and are not tracked.
| # | Hitter | Tm | Game | HR chance | Best price | Book | Park HR env | Opposing starter | HR allowed / start |
|---|
Park HR env is the park-and-weather home-run environment for that game from the day's park export, not an adjustment layered onto the price. HR allowed per start is the opposing starter's projected home runs allowed for this outing. A starter with no published projection shows Pending rather than a substituted league average.
Where the projection sits above the posted line
Ranked by projected strikeouts minus the line the market itself agrees on. Books do not all hang the same number, so each pitcher is priced at the line with the most two-sided quotes and only the books posting that exact line are averaged — never a blend across different lines. Every row is Context — a strikeout prop has no published event probability behind it, so it cannot clear the +6% gate and is never tracked in the results log.
| # | Pitcher | Tm | Game | Line | Projected K | Edge | Market over | Best over | Book | Proj. innings |
|---|
Projected K and projected innings come from the day's starter export. Edge is projected K minus the posted line — a raw projection gap, not an adjusted price. Market over is the de-vigged chance of going over that line, averaged across the books quoting both sides of it. A starter with no published projection shows Pending and is listed after the ranked rows rather than being given a substituted league average.
Every tracked selection, graded
| Slate | Game | Market | Selection | Price | Book | EV at entry | Gate | Tier | Result | Units |
|---|
Graded the morning after each slate from official settlement, at a flat 1 unit per tracked selection, with official plays and provisional model leans kept in separate records. Settlement wording is preserved exactly — Won, Lost, Refunded, Half Won, Half Lost. A slate with no qualifying play is logged as No action, not skipped.
How this page is built
Consensus de-vig, not a single book
Each book's two-sided market is stripped of hold with p = (1/d) / Σ(1/d), then averaged
across every book quoting that line. Average hold on tonight's board is printed in the KPI strip.
+6% EV, rounded against you
Required probability is 1.06 / decimal price. Plus-money requirements round up, minus-money
magnitudes round down — the same single rounding rule the report uses everywhere.
Starters and park, nothing more
Each side's expected runs are the opposing starter's projected runs allowed over his projected innings, plus a relief baseline across the innings he does not cover, with the park's run environment applied to the relief portion only — the projections already price the park. Both sides become negative-binomial run distributions, convolved into a joint margin and total. There is no team-offence, bullpen-quality or defence term, so rows it likes are logged as Lean, never promoted.
What this page will not invent
Event probabilities come from the model inputs — a BallparkPal export, starter projections and park factors. Any market with no event probability shows Pending and can never become an official play. The home-run board is priced straight off the market and carries no event probability, so it stays context-only. Strikeout props are not posted yet for this slate.
Mean absolute deviation is measured against the consensus de-vigged probability. A market where the model is 6 points away on average cannot produce a trustworthy +6% gate on its own — which is why the model tier is provisional and tracked separately in the results log.