| Spec | Value |
|---|---|
| RTP | 67.03% low build / 74.48% high build — the HEAVIEST cut in Pragmatic’s scratch range, a 25.5-33% edge (six to eight times a 96% pokie’s), on its most expensive card; the demo runs the HIGH build, so free play systematically flatters the paid product |
| Volatility | Effectively high: a fixed-odds instant with a 250,000x top prize and no published probability table |
| Max win | 250,000x the card price — the literal A$1,000,000 on a A$4 card |
| The disclosure | Per-symbol odds gated behind the in-game PAYTABLE button; no prize counts published — and one UK player DID land the million in 2020, a publicised, selection-biased anecdote that is the marketing department’s favourite statistic |
| The format | A$4.00 per nine-panel card, three of a kind; FREEPLAY exists here — three Freeplay word-symbols award one extra card at the same expectation; bundles A$20 (5) / A$40 (10) / A$95 (25, “save 5%”) / A$180 (50, “save 10%”) |
| The cut in dollars | A$1.03-A$1.32 handed to the house per card, every card — an average A$4 card returns ~A$2.68 (low) to ~A$2.98 (high) in an expectation almost no single card pays |
The winning line
Wolf Gold 1,000,000 carries the heaviest cut in Pragmatic’s scratch range. On its priciest card, a 67-74% return means the house keeps a quarter to a third of every dollar. Yes, one UK player won the literal million in 2020. That is one publicised winner, though, which is selection bias dressed as a press release, not evidence of frequency. Confirm the 74.48% build because the demo flatters, limit the sitting to 5-10 of the A$4 cards, include Freeplays in that cap, and bank any return of A$40 or more.
Step 1 — Price a third-of-every-dollar product with open eyes
There are seven points between the two builds, 67.03 vs 74.48. That works out to about A$0.29 per card, or A$2.90 across a 10-card sitting. The demo uses the friendly version, so its feel gives you a systematically false preview of the paid product. And even on the good build, a quarter of turnover stays behind.
Action: Check the PAYTABLE/info screen before depositing. Confirm 74.48% or cost the sitting at 67.03%. If the build is hidden, treat it as the low one.
Watch for: The Wolf Gold slot beside it returning ~96% under the same wolf branding. It is the same-name trap, and costs six-to-eight times less to sit at.
Decision branches:
- If 74.48% confirms → Step 2 at the capped count.
- If 67.03% shows or nothing shows → shop a better operator or cut the count hard; the low build burns ~A$1.32 per card.
Step 2 — Cap five to ten of the four-dollar cards
At this card price, the discipline maths bites harder than it does on the cheaper siblings. A run of 5-10 cards costs A$20-40, gone the moment you buy them, while the bundle ladder pushes hardest of all here. Fifty cards at A$180 supposedly “saves” A$20, but it also forces A$140 of extra turnover through a game keeping roughly a third, adding A$19-25 in expected loss for the sake of the sticker.
Action: Fix the number of cards before opening the menu. Only take a bundle when that quantity was already the plan.
Watch for: The pull of the A$180 tier. It is the range’s biggest single upsell attached to the range’s worst return.
Decision branches:
- If the plan lands on a bundle size → take the matching discount; free on a decided number.
- If the sticker is the only argument, well, the count holds; A$5 saved never buys A$25 of edge back.
Step 3 — File the millionaire under selection bias, bank A$40+, count Freeplays inside
The 2020 million happened. It also teaches, honestly, the wrong lesson by design. One publicised winner drawn from millions of cards is selection bias in its cleanest form, because none of the losing sittings received press releases. Most realised return sits in the small and mid tiers, paying only a few multiples of the card. The Freeplay word match gives one more draw at the same expectation, which is a pleasant zero, not an edge.
Action: Set the stop-win and stop-loss in CARDS before scratching starts. Count Freeplays within the limit. Any prize worth 10x the card, meaning A$40+, gets banked and ends the sitting.
Watch for: The anecdote quietly setting the budget. “Someone won it in a small session” is the product’s strongest salesman, mind you, and its least representative fact.
Decision branches:
- If the cap arrives unbanked, it is closed; the eleventh card carries the first card’s odds at the first card’s cut.
- If a card pays A$40+ → banked and gone; leaving with the outlier was the entire point.
- If the million ever prints, the 2020 anecdote has a sequel, and you have a press release to decline.
The maths that backs the playbook
The cut is not a side detail here. It is the whole product. A 67.03 to 74.48% return puts the edge at 25.5 to 33%, six to eight times that of a middling pokie, and the deepest in Pragmatic’s scratch range. On every A$4 card, the house takes A$1.03 to A$1.32 in expectation before the foil is even removed. The two build setup copies the variable RTP slot model, with the demo systematically showing the high build, so Step 1 stops being basic hygiene and becomes consumer protection. There is no PAR sheet and no published tier count, but the structure still tells the story: a match 3 instant carrying a 250,000x headline pushes nearly all realised return into low and mid prizes, while the top award sits somewhere in an unpublished tail. The one recorded million, won in the UK in 2020, is a selection biased data point with precisely zero value for planning a session. Freeplay simply adds another draw at unchanged expectation, a free spin on the same 26 to 33% wheel. Strategy comes down to two levers, build verification and count discipline, plus one behavioural line in the sand: bank at A$40, which is 10x a card whose modal result is nothing. The million made the press release. The cut is guaranteed. Plan around the guarantee.
What breaks the playbook
- Carrying the slot’s maths across the wolf. The pokie pays ~96%; this card keeps up to a third, same canyon, different product entirely.
- Trusting the demo’s warmth, honestly, is risky. It runs the 74.48% build; your casino may run the 67.03% one, and the card face will never say.
- Budgeting off the 2020 millionaire is a trap, well. One publicised winner is the definition of selection bias; the unpublicised sittings paid for the press release.
- Climbing the bundle ladder for stickers. A$5-20 saved against A$19-25 of added expected loss is the menu’s standing arithmetic.
- Reloading after the A$40 hit. The cut waits at the buy button, a third of every dollar, patient as ever.
The cut-off
Build gate: Confirm 74.48% or assume 67.03%. Whatever the demo feels like proves nothing.
Count cap: Set 5-10 A$4 cards before opening the menu, include Freeplays within that number, and ignore the stickers.
Bank line: A$40+ finishes the sitting. The million finishes the genre.
Anecdote law: Publicised winners never set the budget. The cut sets every one.
Strategy by Dave Holloway , Floor Veteran. Last updated: 16 Jul 2026. 18+. Gamble responsibly. Gambling Help Online , 1800 858 858 | Lifeline 13 11 14.
Specs
| Spec | Value |
|---|---|
| Provider | Pragmatic Play (in-house) |
| Released | 2020-05-20 |
| Game type | Instant scratchcard (match-3) |
| Grid | 3×3, nine scratch panels |
| RTP | 67.03%–74.48% (two published builds) |
| Card price | A$4 fixed |
| Top prize | 1,000,000 (250,000× the card) |
| Max win | 250,000× card price |
| Volatility | High (fixed-odds; top prize astronomically rare) |
| Bundles | A$20 (5), A$40 (10), A$95 (25), A$180 (50) |
| Theme | Wildlife, North-American desert canyon |
How it works
This is an instant scratchcard, not a pokie. No reels or paylines here. There is no free spin round either. You buy the card, reveal the 3×3 grid with nine panels, and that is the whole punt. What shows is what pays.
Each panel hides a symbol carrying a fixed cash value. Match three or more of the same symbol anywhere on the card and the printed amount pays on the spot. No lines. Adjacency means nothing. Three of a kind counts wherever those symbols sit among the nine panels. The top symbol tier pays the 1,000,000 prize, which equals 250,000× the A$4 card. Mind you, every card is a fresh draw. The previous result means nothing.
Game rules
Buying a card
The card costs A$4. Fixed. Bundles simply put more cards in your hand upfront and shave a little off the total, so you are changing volume rather than the stake. Each card remains its own independent draw on the same odds.
- One card costs a flat A$4.
- Look, the bundles are A$20 for 5 cards or A$40 for 10 cards. Bigger packs cost A$95 for 25 cards at ~5% off, while A$180 gets 50 cards at ~10% off.
- Buying a bundle only changes how many cards you have. It does not alter the price or odds attached to any single card.
The scratch and reveal
Once the card is bought, you reveal the nine panels and see the draw. In the demo, honestly, all nine pop open together. There is no need to scratch each panel separately before the result settles.
- The 3×3 grid contains nine panels. Each one hides a symbol carrying a fixed cash value.
- All nine symbols show at once when the card is revealed.
- The result is locked in when you buy the card. Scratching, of course, only uncovers what was already decided.
Match 3 to win
The win rule is dead simple for any punter. Three matching symbols pay. Anything short of three of a kind is a dead card.
- Three or more identical symbols anywhere on the card win that symbol’s fixed cash value.
- Position does not matter; there are no lines or adjacency requirements.
- A card with no three-of-a-kind returns nothing.
Freeplay
Freeplay is the only extra feature on the card. It stays within the same match-3 rule and does not open a separate bonus round. You get another draw paid for by the house, not your own bankroll.
- Matching three Freeplay panels awards one free card.
- The free card plays at the same odds and the same fixed prize tiers as a paid one.
- There is no multiplier reveal, no wheel and no pick-bonus beyond this.
Prize tiers
Pragmatic describes the payout structure as a wide spread of prizes on a complex paytable, which you can view in-game through the paytable button. The exact per-symbol cash amounts are not published outside the game. That means the middle tiers are read from the card, not a public table.
- Each winnable symbol carries its own fixed cash value.
- The big one is 1,000,000. That works out to 250,000× the A$4 card, plain and simple.
- Honestly, the mid and low tier amounts are buried in the in game paytable. You will not find the actual per symbol figures published anywhere else.
Max win
- 1,000,000 on the nose. That is 250,000× the A$4 card price.
- Look, this is one fixed top prize. It does not grow with play and it is no progressive jackpot.
Screenshots
A freshly bought card, nine wolf-covered panels waiting to be scratched, with the match-three goal shown alongside.
The same card fully revealed, each panel showing a prize tier, with matching three of a kind being what pays; one panel here landed on Free Play.
The in-game prize table lists the award tiers, how many tickets exist at each and the game’s stated RTP range.
RTP versions
Pragmatic sends this card out in two certified builds. The operator picks the one you get, and most punters never bother checking. That is exactly why it matters.
| Build | RTP | House edge | Notes |
|---|---|---|---|
| High | 74.48% | ~25.5% | The best published figure; the demo runs this build |
| Low | 67.03% | ~33% | The reduced build operators can integrate instead |
The gap is the whole story, really. Even the better build returns about 74.48%. At the low end, the house keeps roughly a third of every A$4 card. The Wolf Gold video slot, using the same artwork, runs at 96.01%. So swapping the slot for the 1,000,000 scratchcard means quietly ditching a game that pays back about 96¢ on the dollar for one returning roughly 67¢ to 74¢. Nothing on the card face tells you which build is running. Treat it as the low one until the in game paytable or info screen proves otherwise.
What 250,000× means
Sure, the 1,000,000 top prize is genuine and tied straight to the stake. It is 250,000× a A$4 card, not empty marketing carry on. Pragmatic does not publish the odds for each prize on this card, so there is no printed probability worth quoting. Still, a fixed odds instant game has a shape, and this one tells you plenty.
- Almost all of the returned RTP lives in the low and mid cash tiers that pay small multiples of the A$4 card.
- The 1,000,000 tier contributes only a tiny sliver of RTP, because its probability is minuscule.
- Buying more cards raises your total turnover through the house edge; it does not shorten the odds on any single card’s top prize.
- One UK player did land the 1,000,000 back in 2020, but a single publicised win is a marketing highlight, not a frequency you can plan a session around.
Scratchcard vs the Wolf Gold slot
Same wolf. Same red canyon. Completely different maths. Punters who know the Wolf Gold pokie should not assume those numbers carry across to the card, because the artwork is the only thing the two really share.
| Feature | Wolf Gold 1,000,000 (scratchcard) | Wolf Gold (slot) |
|---|---|---|
| Format | Match-3 instant scratchcard | 5×3 video pokie |
| RTP | 67.03%–74.48% | 96.01% |
| How you win | Match 3+ identical symbols on a 3×3 card | Reel wins plus the slot’s own respin feature |
| Top outcome | 1,000,000 (250,000× the card) | The slot’s own jackpot maths |
| Decisions in play | None; stake sizing only | None beyond bet sizing |
Wolf Gold 1,000,000 is the top-paying card in Pragmatic’s 2020 seven-scratchcard line, but “top-paying” means the biggest fixed top prize, not the best value. Big difference. A punter chasing the wolf theme with real RTP is looking at the Wolf Gold slot, not another card in the scratchcard band.
FAQ
Q: Is this the same as the Wolf Gold pokie?
A: Not even close. Straight up, this is a match-3 scratchcard running 67.03% to 74.48% RTP, while the Wolf Gold slot runs 96.01%. Same wolf art. Roughly 22 to 29 points worse RTP.
Q: What does the 1,000,000 top prize cost me?
A: It is A$4 per card, fixed. The 1,000,000 works out to 250,000× that card. It is a real prize tied to the stake, but the odds on it are astronomically long.
Q: Does buying the 50-card bundle improve my odds?
A: Not really. The bundle only shaves the price a touch. Every card still carries the same house edge, so bulk-buying just pushes more turnover through a 25% to 33% cut.
Q: Is there any strategy to a scratchcard?
A: No. There are no decisions inside the card, so the only real levers are bankroll discipline, checking which RTP build the casino runs and keeping the stake sensible.
Q: Which RTP build will I actually get?
A: Depends on the operator. The build is not shown on the card face, and the demo runs the higher 74.48% one. In real money, assume the low 67.03% build until the in-game paytable or info screen confirms otherwise.
Last updated: 01 Jul 2026
Reviewed by Mark Henderson, Senior Slots Analyst
Demo embed · played in fun-money mode. No real-money play here.