Methodology
Every metric on this site derives from the two per-tier facts state lotteries publish: printed odds ("1 in N tickets") and prizes remaining of total printed. No proprietary inputs, no modeling beyond what those two facts support, recomputed daily from the sources listed below.
Tickets printed
For any tier, tickets ≈ tier total × tier odds. We take the median of that product across tiers with published odds to damp rounding noise. Texas states its print runs outright and New York's overall odds encode them, so for those states the estimator recovers the published figure exactly.
% sold
Tiers with thousands of prizes are a serviceable random sample of the print run, so Σremaining ÷ Σtotal over high-count tiers (≥5,000 prizes) estimates the fraction of tickets still unsold. Low-count tiers are excluded from this estimator — five top prizes carry no information about sales.
EV per $1
Σ(prize × remaining) ÷ estimated remaining tickets ÷ ticket price: the expected return per dollar on the pool as it exists today, not as printed. It is always below $1 — every scratch-off is a negative-EV bet. The ranking is least-bad, never good.
Luck z
Under a uniform-shuffle null, top prizes remaining ~ Binomial(n, fraction unsold); z measures how far the observed count sits from that expectation. With a handful of top prizes the raw remaining/expected ratio is mostly noise ("3 of 5 left at 62% sold" is z ≈ 1.0, i.e. chance), so hot/cold badges and the score key on z (|z| ≥ 1.5), never the ratio. Standing caveat: states post claims with a lag of days to weeks, so every count here trails reality and a hot board may already have been hit.
Annuity discounting
Top prizes paid as multi-decade annuities are counted at ≈60% of face value (a rough 25-year cash equivalent at ~4.5%), so a $5M-over-25-years game can't outrank a $5M-cash game on sticker price alone.
Score
A 0–100 composite for the default sort: EV dominates; statistically significant luck (capped z) adds or subtracts a little; boards with no top prize left, or with $10k+ tiers depleted well below what sales predict, sink hard. It ranks games against each other today; it is not a probability of anything.
What we cannot know
Which ticket in a roll wins. Nobody can — not us, not the retailer, not any paid system. Winners are distributed through the print run precisely so that no public data identifies them. Everything here is an estimate of a game's remaining pool; nothing is a prediction about your next ticket, and nothing improves your odds.
Sources and data quality
- California — daily crawl of the California Lottery's Scratchers® game pages (calottery.com). Full per-tier data: printed odds and remaining-of-total for every tier.
- Texas — texaslottery.com game pages: full per-tier tables (printed vs. claimed) plus the exact tickets-printed figure, stated outright. The highest-fidelity source we ingest.
- New York — the official open-data API (data.ny.gov, "Scratch-Off Game Daily Prize Status Report") for per-tier remaining/total, plus nylottery.ny.gov for price, overall odds, and top prize. NY publishes per-game overall odds only; per-tier odds are derived from the published totals, so every input remains NY-published.
Integrity and corrections
We have no relationship with any lottery or ticket seller, and nobody pays to be ranked. If a figure here disagrees with the state's own page, that's a bug — report it and the correction ships fast.