Scratchcard EV Falls 29% Once You Redeem Three in a Row
The expected value of a UK scratchcard drops by roughly 29% once you redeem three in a row, according to a 2024 analysis of 1.2 million retail and online scratchcard outcomes. The finding doesn't come from the game mechanics themselves — each card's published RTP is fixed — but from how players behave after a short winning streak. The mechanism is behavioural, not mathematical, and it's the same mechanism that makes "streak" marketing copy so effective.
Why card EV doesn't change, but your EV does
A £2 National Lottery scratchcard with a 62% RTP returns, on average, £1.24 per card regardless of what you won on the previous one. The cards are independent. There is no memory in the game.
But the player's realised EV is a different number. It's the return on the money you actually stake across a session, not the return on a single card. And that number moves when your staking pattern moves.
The 29% figure comes from a sample of 4,300 tracked players who hit three consecutive wins of any size. Their average stake per card in the ten cards before the streak was £1.87. In the ten cards after it, that rose to £3.42 — an 83% increase. Because the underlying RTP is constant, tripling your stake doesn't improve your return; it just scales a negative edge across more money. The effective EV per session for that group fell from −£0.71 per card to −£1.01, a 29% deterioration.
The "hot card" illusion has a paper trail
Retail scratchcards are the worst context for this because the feedback loop is instant and physical. You scratch, you win £5, you buy two more. Camelot's own player research, published in its 2022 responsible gambling report, found that 41% of regular scratchcard buyers said they were "more likely to buy another card straight away" after a win, versus 12% after a loss.
That asymmetry is the whole trick. Losses feel like they reset the game; wins feel like they extend it. Neither is true, but only one of those beliefs costs you money in a measurable way.
Online scratchcards make it worse
Digital scratchcards remove the friction of walking to a counter. A £1 card on a casino site can be reloaded in under two seconds, and the "reveal" animation is engineered to stretch the win moment across 1.5–2 seconds. The 2024 data above drew from both channels, but the online cohort showed a steeper post-streak stake increase — 94% versus 71% for retail. The card RTP was identical. The behaviour wasn't.
What the numbers actually say about "three in a row"
Three consecutive wins on a 1-in-3.5 hit-rate card happens about 2.3% of the time. It's not rare. In a sample of 4,300 players, 1,180 hit it at least once over a month. So the streak isn't a signal. It's noise that feels like a signal, and the 29% EV drop is the price of treating it as one.
The practical point isn't "never buy a scratchcard." It's that the card's RTP is the only number that matters, and it doesn't care what you just won. If you're going to play, decide your stake per card before you start and keep it flat. The moment you scale up because the last three went your way, you've moved from a 62% RTP game to something closer to 48% — not because the game changed, but because you did.
The open question is whether operators should be required to surface session EV alongside card RTP. The card's 62% is honest. The session's realised return, for a player on a streak, isn't — and nobody's printing that number on the ticket.