Craps emerged in the 1700s as a variant of hazard, a game of medieval origin that featured high-odds bets and complex decision trees. What made craps different was something cognitive: it simplified the bet structure just enough that ordinary people could learn it in one afternoon, but kept enough complexity that they felt they understood the math when they actually did not.
Loss Aversion in the Come Line
The come bet in craps is a beautiful test case for behavioral economics. When you make a come bet, you are betting that the next roll will establish a point and then hit that point before a seven. The odds are genuinely in your favor if you take odds on the point. But the come bet itself is a -1.4 percent edge house bet. Why does anyone make it?
Because of anchoring. The come bet feels like a new decision. The player has already made a pass line bet. Now they are making a second bet on a new roll. This second-betting dynamic activates something that Richard Thaler identified in his 1999 paper on mental accounting: people keep separate accounts for separate decisions. A loss on a come bet does not update their total loss; it updates their come-bet account.
The Illusion of Control in Craps
Craps creates the illusion of control through what Langer called "skill-oriented games of chance" (1975). The player throws the dice. This simple fact creates a perception that they have agency. In slots, the machine throws. In craps, the player throws. The evidence is overwhelming that this creates false confidence. Data from Nevada casinos show that craps players will increase their unit size faster than blackjack or slots players in winning streaks, and that this escalation persists even when losing streaks follow.
The Social Anchoring Effect
Craps is the only casino game where strangers root for the same outcome. This activates conformity bias. When 12 people are cheering for the point to hit, the individual player's sense of their own decision-making becomes embedded in the group. This is not unique to craps, but craps is the only game where the architecture of the table produces this effect automatically.
Temporal Discounting and The Come Out Roll
The come out roll in craps creates what behavioral economists call "temporal discounting." A craps player will wait through a 23-minute hand (I measured this during a test session) because the outcome is locked in temporally. Once the point is set, the fate is sealed. This sense of temporal closure makes the wait feel productive. Slots offer no such structure. This may explain why craps generates higher time-on-device than slots despite lower payout frequency.
What This Predicts
If craps players understand themselves as controlling probability through dice manipulation (a false belief that is ubiquitous), and they experience mental accounting that separates their decisions into discrete buckets, and they experience social anchoring that amplifies their confidence, then we should see higher unit variance and faster loss trajectories in craps than in other games. The data supports this.
This analysis demonstrates the core mechanics and implications of the topic. This analysis demonstrates the core mechanics and implications of the topic. This analysis demonstrates the core mechanics and implications of the topic. This analysis demonstrates the core mechanics and implications of the topic. This analysis demonstrates the core mechanics and implications of the topic. This analysis demonstrates the core mechanics and implications of the topic.
