Blackjack is a casino card game that involves trying to create a hand total higher than the dealer’s but not exceeding 21. The player may hit, or ask for an additional card, until they are satisfied with their hand or bust. The dealer will then reveal their cards and either stand or draw until they also have a blackjack, or bust.

If the player’s hand-total beats the dealer’s, they win one times their bet. If the dealer has a better hand-total than the player, they sweep the player’s bet and play another round.

The underlying strategy in Blackjack is simple, and it can be learned by playing for free or for real money. This makes the game popular with mathematicians and other intellectuals who enjoy a chance to beat the house edge. However, the game’s popularity has waned in recent years, due to increased competition from other casino games and the rise of high-roller gambling alternatives like baccarat.

Aside from its complexity, another feature of blackjack that sets it apart from other casino games is that the dealer’s actions must conform to a prescribed strategy that is known to all players in advance. This makes it possible to study the effects of expected regret and omission bias in an experimental setting where the dealer’s behavior is under observation.

In the study described here, we analyzed video recordings of actual gameplay at a Blackjack table to investigate the impact of errors associated with omission and expected regret. In addition, we measured the effectiveness of a variety of techniques for reducing these errors. Our findings suggest that some of these techniques, such as counting cards, can significantly improve a player’s winning chances.

To record the video footage, we sat in a booth behind the dealer and watched as the players placed their bets. We then filmed the dealer’s action, including the dealing of the cards and their movement around the table. We used the same camera to record the resulting data on an Excel spreadsheet.

Our results show that, on average, players make about 1 error per 100 hands when using a basic strategy. This error amounts to a 5.5 percent loss of money over an extended period of time. Our analysis also shows that the number of errors made increases with a player’s confidence level in their ability to learn the game. This is consistent with the idea that unjustified confidence in knowledge leads to greater risk taking and decreased use of hints to improve their play.

The most common mistake in Blackjack is to take an ace as 11 when it should be valued at 1. This occurs frequently, and has the effect of increasing the player’s losses. The other major error is to double down when it should not be done. We found that these mistakes are related to a combination of factors, including unjustified confidence, outcome expectations, and anxiety. This research has implications for designing educational interventions to reduce these errors.