The Dial Behind the Near Miss
How Games Decide How Close You Came
"You only needed 2 more jellies" is a message a game chooses to show. Near misses raise the urge to keep playing, and the studios that build casual games have documented, in papers and patents, how they set difficulty and price the way out. The same mechanics have moved into trading apps. The trust problem is not the dial itself; it is hiding the dial while selling the escape.
When you run out of moves in Candy Crush a few short of the goal, the game says so: “Out of moves! You only needed 2 more jellies.” When you were nowhere close, it just says “out of moves.” Chanel Larche, Natalia Musielak and Mike Dixon had 60 regular players play for half an hour while their heart rate and skin conductance were recorded, and found that these near misses raised frustration, arousal and the urge to keep playing compared with ordinary losses (Larche et al. 2017). The study measured feelings, not purchases, and its effects sat close to the edge of statistical significance. It still showed something worth naming: closeness is information the game chooses to display.
Gambling regulators noticed this decades ago. In 1986 R. L. Reid described slot machines built so the jackpot symbol would come to rest just above or below the payline, and a British Royal Commission in 1978 had already criticized instant lottery tickets sold on their “heart stoppers,” tickets that give the illusion of nearly winning a big prize (Reid 1986). The near miss is an old tool. What’s new is how precisely it can be set.
The dial is documented
Studios have published how they tune difficulty. In a paper from Electronic Arts, Su Xue and colleagues described a live match-3 game in which the random starting board was the difficulty lever. On one level, the easiest board won up to 75% of the time and the hardest only 15%. Choosing boards per player raised rounds played by between 4.4% and 7.9%, with no significant change in revenue, and the authors wrote that the adjustment should be invisible to players so they don’t feel coddled or abused (Xue et al. 2017).
One level, two starting boards
Share of attempts that win
- Hardest starting board15%
- Easiest starting board75%
The patents say the same thing in legal language. An EA patent on dynamic difficulty adjustment states that “the difficulty adjustment is undetectable by a user” (EA patent). A King patent describes offering a player extra moves on a level, with the number depending on how many times that player has already played the level and failed (King patent), and an earlier King filing describes levels with “a true fail, to enable monetization through purchasing extra moves” (King filing 2014). A patent shows what a company filed, not what it runs in a given game, and nothing here shows that Candy Crush adjusts difficulty per player. The documents do show the design logic: set how close the player comes, and sell the way past it.
What casinos are not allowed to do
Nevada’s rules for gaming devices already draw a line here. A machine’s outcome must not depend on previous results, the amount wagered, or “the style or method of play,” and once an outcome is chosen, the device “must not make a variable secondary decision which affects the result shown to the player” (Nevada Regulation 14.040). No equivalent rule covers free-to-play games, which can tune outcomes to the player and show the closeness however they like.
The same mechanics in finance
Trading apps borrowed the family. In 2021 the US Securities and Exchange Commission asked for comment on platforms using “slot-machine style interactive graphics, interactive wheels of fortune, or virtual ‘scratch-off’ lottery tickets” (SEC 2021). In 2024 Massachusetts settled with Robinhood for a $7.5 million fine; its order describes a free-stock reward revealed by mimicking the motion of scratching off a lottery ticket, advertising stocks like “Microsoft, Visa, or Apple” despite a low chance of receiving them (Massachusetts consent order). Robinhood neither admitted nor denied the allegations. And in a randomized trial with 2,430 participants trading play money, the Ontario Securities Commission found that people given points for buying and selling made 39% more trades than a control group, though the points had no real value (OSC 2022).
These are cousins of the near miss rather than the same thing: what’s documented in finance is lottery-style reveals and points, not engineered almost-wins. The shared mechanism is a designed feeling placed right where the next action is sold.
At the desk
For a crypto or fintech project, the useful line is about disclosure, not mechanics. Adjusting difficulty, sequencing rewards or showing progress isn’t the problem. The trouble starts when the closeness is set by the operator, hidden from the user, and paired with a paid way out. A points program, a quest system or an airdrop campaign that tells people how close they are should be able to say how that closeness is decided, and keep the “almost there” message away from the checkout. Whether people would mind the dial if they could see it is an open question; what enforcement has targeted so far is the hidden version.
How this could be wrong
The evidence that near misses change behavior, rather than feelings, is weaker than the evidence for the feelings. The Candy Crush study was small and measured urge, not spending, and EA’s own test found no revenue change from adjusting difficulty. Patents show intent, not practice. If studies found that visible, disclosed difficulty tuning drives the same purchases as the hidden kind, the trust argument here would lose most of its force.
References
- Reid, R. L. "The Psychology of the Near Miss." Journal of Gambling Behavior 2, no. 1 (1986): 32–39. · doi:10.1007/BF01019932
- Larche, Chanel J., Natalia Musielak, and Mike J. Dixon. "The Candy Crush Sweet Tooth: How 'Near-Misses' in Candy Crush Increase Frustration, and the Urge to Continue Gameplay." Journal of Gambling Studies 33, no. 2 (2017): 599–615. · doi:10.1007/s10899-016-9633-7
- Xue, Su, et al. "Dynamic Difficulty Adjustment for Maximized Engagement in Digital Games." WWW '17 Companion (2017): 465–471. · doi:10.1145/3041021.3054170
- Aghdaie, Navid, et al. (Electronic Arts Inc.). "Dynamic Difficulty Adjustment." US Patent 10,286,323 B2, May 14, 2019.
- King.com Ltd. "Modifying Gameplay Parameters." US Patent 11,666,826 B2, June 6, 2023.
- King.com Ltd. Patent application US 2014/0080600 A1, published March 20, 2014.
- Nevada Gaming Commission. "Regulation 14.040, Minimum Standards for Gaming Devices." Adopted April 22, 2010.
- U.S. Securities and Exchange Commission. "Request for Information and Comments on Broker-Dealer and Investment Adviser Digital Engagement Practices." Release No. 34-92766, August 27, 2021.
- Massachusetts Securities Division. "Consent Order, In the Matter of Robinhood Financial LLC." January 2024.
- Ontario Securities Commission. "OSC Study Finds Gamification Techniques Influence Retail Investors." November 17, 2022.