Methodology and data

Where the numbers come from

• Prizes remaining: the Texas Lottery publishes a CSV every day with, for every active scratch game and every prize tier, how many prizes were printed and how many have been claimed. Each game's official page adds the overall odds and the approximate number of tickets printed. We fetch these once a day and keep every snapshot.
• Claims and retailers: the state open data portal publishes the Winners List of Texas Lottery Prizes (dataset 54pj-3dxy), one row per paid claim with the selling retailer. It is updated weekly and mostly covers prizes of $600 and up, which must be claimed at a lottery office. Small prizes paid at the counter are not in it.

How a game is scored

For a game with price p, tiers with prize a, printed n, claimed c, and about T tickets printed:

• remaining prizes in a tier = n − c
• fraction sold ≈ total claimed ÷ total printed (claims track sales closely for a game with millions of small prizes)
• tickets remaining ≈ T × (1 − fraction sold); when T is not published we use total prizes × overall odds
• payout now = Σ (remaining × a) ÷ tickets remaining ÷ p
• designed payout = Σ (n × a) ÷ T ÷ p
• value delta = payout now − designed payout. Positive means the unclaimed pool is richer than the game was built to be.
• any win = remaining prizes ÷ tickets remaining; break even counts only prizes ≥ p

Every scratch game pays out well under 100%. Texas games are designed between roughly 55% and 80%. Ranking by payout now finds the least-bad ticket on a given day, and the delta finds games that have drifted in your favour because the big prizes have not been hit yet. Nothing here changes the fact that the expected outcome of any purchase is a loss.

Scratch Match

We enumerate every multiset of current price points that sums to your budget (up to 20 tickets), fill each price slot with the best game or two by payout now, and rank the combinations by expected return. The three shown always include a one-ticket option and a multi-ticket option when both exist.

Hot shops

Each retailer within your radius gets a hot score: for each claim of $600 or more in the window, log10(amount) × 0.5^(age in days ÷ 60). Recent wins count more, and the log keeps one jackpot from dominating forever. We also show the average and median amount per claim, the largest win, and the number of $10,000+ wins. Lottery claim centers and promotional accounts are excluded.

Read this honestly: tickets are printed and shipped at random, so a store's past winners do not change the odds on the next ticket it sells. High-volume stores show more winners because they sell more tickets. The map answers "where have people been winning lately", nothing more.

Data status

• Prize snapshots: 1 most recent, latest as of 09/05/2026 (82 games)
• Claims: 106,409 rows, latest paid 08/31/2026; 15,881 active retailers, 13,195 geocoded