The pick.
The price. The result.

Our slate tiers

For boards first published from October 30, 2026, price tiers use the original saved selection price. Switching books or later price changes keeps the selection in its recorded tier. A tier describes price and default sizing; it does not guarantee a result.

Daily Board

Bench is excluded from our official picks and hit/miss record. Its separately labeled hypothetical tracker uses the corresponding price tier's size. All recent-form L5/L10/L20 views use only current-season appearances; a smaller sample is shown honestly.

Parlays

Same-game is a separate badge and filter across the price tiers. A recommended SGP's final ticket price must be -110 or better. Conditional model targets and fair model odds are not offered sportsbook prices; unquoted ideas receive no stake or tracked payout. Saved marginal-price products remain illustrative and do not price an SGP.

Ry’s personal parlays retain the separate 0.5u default. You can set your own unit dollar value and sizes. Earlier editions retain their original category and default-stake policy, including 2u/Layup and 1u/Lob; they are not reclassified under these new tiers.

Your board, at a glance.

Each compact row shows the pick, saved price, our win estimate, and live or final result. MP means minutes played; it appears beside the result on phones. Expand a row for its projection, estimate status and timestamp, exact terms, research and box-score source. Missing minutes stay unavailable rather than being guessed.

Your filters change the view, not our saved selections or estimates. Hit rates use final wins divided by final wins plus losses; pending picks, pushes and DNPs are excluded. Different lines for the same player can share one underlying performance.

A saved price has a time and a context.

Collected sportsbook observations come through PropLine. Ry may also supply a saved screenshot price. We identify the sportsbook and show the capture time when it is recorded; otherwise the time is unavailable. A saved quote is not a current offer, an executed wager, or proof that a reader could obtain the same price.

The research uses NBA statistical history. ESPN box scores supply statistical results for published selections with verified game and player identities. Public source links support result checking; they are not an endorsement by those sources.

Historical imports can lack an original publication timestamp. We disclose this rather than treating a file creation time or later import as proof that a selection was public before a game. Recorded time, capture time, and result check time are separate facts.

How statistical results are checked.

We match a selection to an explicit game and player identity, then compare the relevant box-score statistic with the exact posted operator and threshold. Combination stats use the named components. Strict whole-number over/under lines can push at equality; a milestone such as 8+ includes eight as a hit.

A live statistic stays unfinished until the game is final. A player who did not play is not assigned a zero-stat miss. Missing statistics, unresolved identities, and source failures remain pending or unavailable. Postponements and cancellations are kept distinct.

A statistical hit or miss does not establish the sportsbook’s payout, void decision, or correction policy. Overtime and settlement terms can be unconfirmed. We therefore do not report verified wagering profit from these records.

A record you can inspect over time.

Sunday and daily editions are dated. The archive keeps the published selections and their available outcomes. New result updates create new saved versions; corrections should state what changed and retain the earlier record. Losing selections belong in that history.

While a picks page is open, available verified box scores refresh about once a minute. Source failures leave the latest known result visible. Refreshing the website does not fetch new sportsbook prices or place wagers. Publication updates and price captures occur separately from statistical result checks.

Interpreting research responsibly.

The displayed record is a selected sample, not an independent audit of the entire algorithm. A short run, multiple related lines, missing historical timestamps, and changing prices all limit performance conclusions. Hits and misses alone do not measure calibration or establish an advantage over the market.

We publish the selections and basic outcomes needed to inspect that record. Final probabilities and projections can be public; the algorithm code and its calculation rules remain private. A projection is an estimated stat total; the win estimate concerns the exact posted proposition. Both are saved estimates, not live recalculations or guarantees. A personal pick without a matched estimate stays unavailable. Stronger performance claims require an explicitly described evaluation with a date range, sample size, baseline, and uncertainty; none is implied by the board.

Your combination, at saved prices.

Use + to add a priced selection to the illustrative parlay calculator. Another line for the same player and stat in the same game replaces the earlier choice. Potential return includes your stake; net profit subtracts it. Same-game or same-player legs can be correlated, and a sportsbook’s combined price may differ.