How we score value.

Each agency publishes, for every instant game, some slice of its prize tiers — how many of each were printed and how many are still unclaimed. That’s the raw material. The Value Score reads it every morning and asks one question: are the big prizes disproportionately still out there?

Not every agency discloses the same slice, though — so Lottizen uses three different formulas depending on what a given province actually publishes, and labels which one applies on every ranking page. See the three methods below for the full breakdown.

Three scoring methods

A province’s data-completeness badge (shown on its ranking board) tells you which of these applies before you even open a ticket:

1 · Retention — Ontario, British Columbia, Quebec (“Full tier data”)

OLG, BCLC, and Loto-Québec all publish a real printed total and a real remaining count for every valued prize tier. This is the richest data, so it gets the full retention formula — see the step-by-step walkthrough below.

2 · Remaining Value Index — Western Canada / WCLC (“Remaining counts only”)

WCLC (covering Alberta, Saskatchewan, and Manitoba) never publishes how many prizes a ticket printed with — only how many of each tier (≥$100) are still unclaimed today. Retention is mathematically impossible without a printed total, so Western Canada instead uses:

— the dollar value of remaining prizes, disclosed tiers only, per dollar of ticket price. This is a raw CAD figure, not a 0-100ish score, and it is not comparable to Ontario/BC/Quebec’s retention scores or to Atlantic Canada’s top-prize percentage — only to other Western Canada games.

3 · Top Prize Remaining % — Atlantic Canada / ALC (“Top prizes only”)

ALC (covering New Brunswick, Nova Scotia, PEI, and Newfoundland & Labrador) publishes a real printed total and remaining count — but only for a game’s own headline top prize tier(s), never any lower tier. So Atlantic Canada’s score is simply:

— summed over the top tier(s) only. It happens to land on a similar 0-100 scale to the retention score by coincidence of units, but it is measuring something different (one tier, not the whole game) and isn’t directly comparable either.

Data-source limits — and what we’re still digging for

WCLC and ALC’s narrower formulas above reflect what their public feeds actually contain today, not a shortcut we chose. Lottizen has looked (and keeps looking) for any regulatory disclosure, annual-report appendix, or provincial gaming-commission filing where either agency might publish fuller printed-total data — if one turns up, these two provinces will move to the full retention formula. Until then, this is the honest ceiling of what can be measured from public data.

Loto-Québec is a special case worth naming: its printed totals are published on a live “état de réclamation des lots” page (paired total + claimed columns per tier), which Lottizen uses as Quebec’s sole source. During this build we specifically looked for a separate downloadable PDF re-publishing the same per-game totals — to cross-check for drift — and could not locate one distinct from the live page; the one prize-related PDF discoverable from Loto-Québec’s own rules page turned out to be a generic draw-game formula table, unrelated to scratch tickets. If a genuine second source surfaces, reconciling it is a candidate follow-up.

Retention, step by step (Ontario / BC / Quebec)

1 · Add up prizes, two ways

Across a game’s valued prize tiers:

2 · Two fractions left

f is the “remaining prize ratio” — a proxy for how much of the print run is still unsold. g weights that by dollars.

3 · Retention = value vs count

• retention > 1 — big prizes are still disproportionately unclaimed. Better value.
• retention = 1 — prizes draining evenly with sales. Baseline.
• retention < 1 — jackpots mostly gone, small prizes left. Worse value.

4 · Scale by the game’s payout rate

The payout rate is the share of a game’s sales it is printed to return as prizes. Where the agency publishes it for each game, Lottizen uses that figure; where it doesn’t, Lottizen uses an assumed 62%, a rough instant-game average, and says so below. The score is still not “cents back per dollar” or an expected return: it describes how the game is built and how much of its prize money is left, never the odds of any ticket.

How each province is ranked

Each agency publishes different data, so each province’s ranking uses the method its data supports. Scores are only comparable within a province.

Using a published payout rate changes the order within a province, because rates differ by game (OLG’s run from about 63% on $2 tickets to about 75% on its $50 and $100 tickets). An assumed rate is the same for every game, so it doesn’t change the order; it only sets the scale. See the Ontario price guide for the published rates by price.

What the score is — and isn’t

• It is a like-for-like way to compare which games still have the most prize value left today, straight from each agency’s own numbers — within one province and scoring method.
• It isn’t a prediction, a system, or a way to beat the odds. The house edge is baked in and unchanged.
• It works from disclosed prize tiers. Each agency’s feed lists a game’s notable prizes, not every $2 win (and for WCLC/ALC, only a subset of tiers at all), so the score reflects the prizes worth chasing. Treat close scores as ties.
• It isn’t comparable across scoring methods. Ontario/BC/Quebec’s retention scores, Western Canada’s Remaining Value Index, and Atlantic Canada’s Top Prize Remaining % are three different measurements — only compare games within the same province.

Draw results: sources and gaps

Winning numbers come only from each operator’s own published results. When two official sources disagree about a draw and no third official source settles it, we don’t pick one: the draw is left out of the history and the statistics, and its date is listed on that year’s results page as “Official source records disagree; not recorded.”

Western Canada’s older draws are the one place where a single source remains. Before October 2025, WCLC’s results for the Western Extra, and for some Western Pick 2/3/4 dates, survive only in its “since inception” PDFs. A draw printed in two or more of those PDFs is stored when they agree. A draw printed in just one PDF, with no second official source to check it, is still shown, but it is marked “PDF only” on its year’s results page, and the statistics page says its data range includes such draws. That covers 727 Western Extra draws (2007–2024) and 48 draws of each Western Pick game (2019–2024). The PDFs are not error-free:

• Western Extra: Where WCLC’s PDFs can be checked, they disagree with each other on 68 of 4,043 dates (1.7%), and 6 of 1,264 PDF rows since October 2025 differ from WCLC’s results pages (0.5%).
• Western Pick 2/3/4: Where the Pick PDF can be checked, it disagrees with WCLC’s per-draw prize pages and results pages on 48 of 9,087 values (0.53%).

OLG’s daily games (Pick 2/3/4, Encore, Daily Keno, Poker Lotto) from 2014 to 2019 come from OLG’s published draw-result news releases and can’t be cross-checked against OLG’s results API, which only reaches back about 16 months. Later releases matched the API on all 1,064 draws where both exist.

Draws left out because official sources disagree: 67 Western Extra dates (2005–2025) and 42 Western Pick dates (9 Pick 2, 15 Pick 3, 18 Pick 4). One misprint was corrected rather than left out: WCLC’s PDF prints the Lotto Max bonus number for January 25, 2022 as “374”; Loto-Québec and BCLC both publish 37, which is what we show.

Draws refilled from BCLC’s per-draw records on October 27, 2026, where the other archives lacked them: seven Daily Grand draws (2021–2025), each matching Atlantic Lottery’s results exactly, and four Lotto 6/49 draws (May 2, 1998; Aug. 15, 18 and Sept. 15, 2012), for which no second official source we read covers those years.

Download the full datasets on Kaggle (CC BY 4.0, updated weekly):

• Canada Lottery Draw History: every Canadian draw game Lottizen tracks, one CSV per game, with each missing or left-out draw and its reason
• Canada Lottery Jackpot History: Lotto Max, Lotto 6/49, Daily Grand and Lottario top prizes per draw, where official sources agree

Freshness

Rankings for all 5 provinces are rebuilt every morning from each agency’s public data, on independent schedules per agency (so one agency’s outage never delays another’s). This page’s data was generated October 27, 2026. Prizes can be claimed at any time — always confirm the current numbers with the official lottery operator (OLG, BCLC, WCLC, ALC, or Loto-Québec) before you buy.