Near Misses: Why Matching Three Numbers Makes You Buy More Tickets

You matched three numbers. You won a few pounds, or nothing much. And the feeling is not "I lost" — it is "I was close."

That feeling is one of the best-studied effects in gambling research, and it is worth separating carefully into two questions: why the feeling occurs, and whether it means anything. The answers are "we have decent evidence" and "no, nothing at all."

The classic review

R. L. Reid, The Psychology of the Near Miss, Journal of Gambling Behavior, volume 2 (1986), pages 32–39.

Reid's review is the origin point for the modern literature. His argument was that certain gambling outcomes are failures that resemble successes — two jackpot symbols and a third just above the payline, three of six numbers matched — and that these outcomes are treated by the player as evidence of progress rather than as losses.

Reid also noted that near misses can be engineered. In a game where the outcome display is designed rather than dictated by the mechanism, the frequency of near-win displays can be tuned independently of the actual payout probability. That observation drove decades of regulatory attention to reel mapping in slot machines.

Draw games are different in an important way, and it is worth saying plainly: a lottery draw cannot be tuned to produce near misses. The three-matched outcome is a straightforward consequence of the combinatorics, not a design choice. What lotteries do choose is the prize structure — whether matching three pays at all, and how visibly that tier is presented.

The neuroimaging evidence

Luke Clark, Andrew J. Lawrence, Frances Astley-Jones and Nicola Gray, Gambling Near-Misses Enhance Motivation to Gamble and Recruit Win-Related Brain Circuitry, Neuron, volume 61, issue 3 (2009), pages 481–490.

The design used a simplified slot-machine task in an fMRI scanner with healthy, non-problem participants. The findings:

• Near-misses were rated as less pleasant than full misses. Subjectively, they feel worse.
• Near-misses nonetheless increased the desire to continue playing.
• Near-misses recruited overlapping brain circuitry to that engaged by actual wins, despite delivering no money.
• The effect was restricted to trials where the participant had personal control over arranging the gamble — where they had chosen, rather than been assigned.

The last point is the one that matters most for lotteries, and we will come back to it.

Two honest caveats. First, this is a slot-machine task, not a lottery draw; the transfer is plausible rather than demonstrated. Second, it is a laboratory study of a psychological mechanism, not a measurement of how many extra tickets anyone bought. Treat it as evidence about why the pull exists, not as a measured effect on lottery sales.

What a near miss is worth: the arithmetic

Here is the part the feeling gets wrong.

Matching three numbers is not a partial jackpot. It is a different, far more likely event that shares no causal relationship with the jackpot at all.

The UK Lotto figure is straightforward combinatorics. The number of ways to pick exactly three of your six numbers and three of the other 53 is C(6,3) x C(53,3) = 20 x 23,426 = 468,520, out of C(59,6) = 45,057,474 total combinations. That is one in 96.2.

So matching three in UK Lotto happens to roughly one ticket in every hundred. Across a draw selling millions of lines, hundreds of thousands of players have the "I was close" experience every single week. It is the most common non-trivial outcome in the game, and it is engineered to be frequent by the prize structure, not rare by the odds.

Meanwhile the gap between one in 96 and one in 45 million is a factor of 468,520. In no meaningful sense were you close.

Why "close" is the wrong concept

A near miss in the physical world is genuine information. If your arrow lands next to the bullseye, you are a better archer than someone who missed the target, and adjusting your aim slightly makes sense.

A lottery draw has no such geometry. Matching three of six does not indicate that your selection was in some sense nearly right, because there is no continuum along which selections can be more or less right. Every combination has identical probability, and the draw is independent of every previous draw — the point made in cold numbers and the gambler's fallacy.

Nor does having matched three last week change anything about next week. The machine does not remember. The proof of that independence, and of how little a second ticket buys you, is in why two tickets change nothing.

The control finding, applied

Clark and colleagues found the near-miss effect appeared when participants had chosen their gamble, not when it was assigned.

If that transfers — and it is a plausible if unproven transfer — it predicts something specific about lottery play: near misses should bite harder for people who pick their own numbers than for people who take a quick pick. Choosing creates ownership, ownership creates the sense that the outcome was partly yours, and a near miss then reads as feedback on your choice rather than as noise.

It is not feedback. It is noise. But the felt difference is real, and it is closely related to why people will not swap their regular numbers even for a cash premium.

The practical version

Nothing here says stop playing. It says be clear about what a three-match is:

• It is a common outcome, not a rare one — around one ticket in a hundred in UK Lotto.
• It is not partial progress. There is no progress to be partial towards.
• It carries zero information about future draws.
• The urge to buy another ticket after one is documented, expected, and unrelated to your prospects.

If you want to see how the tiers actually distribute, jackpot odds vs any prize odds separates the two, and the odds calculator will produce the full tier table for any game format.

The related pressure — a huge jackpot making the purchase feel more attractive when it has made your odds no better and your expected share smaller — is in why jackpot size drives sales.