Blackjack Table Simulator

The common belief is that a bad player at your table hurts your odds. This simulator tests it directly, by dealing thousands of shoes to a full table and measuring your seat under both conditions. No money, no accounts, no stored data.

How the table simulator works

You set the table size, from one to seven seats, and the strategy of every seat other than your own. Your seat always plays the published basic strategy chart for the ruleset selected. Each round is dealt to every seat and the dealer, every seat acts in turn, the dealer plays to house rules, and the shoe is reshuffled at the penetration point you set.

The same configuration is run twice, once with every other seat playing optimally and once with every other seat playing typically, so the comparison is the headline rather than a single run. Where you set a mix of optimal and typical seats, that mix is run as a third column.

  • Your win rate and the house edge actually realised at your seat
  • Your bankroll curve across every round simulated
  • Hands per hour, so more seats visibly means fewer hands for you
  • Variance and standard deviation of your per round result
  • Rounds dealt per shoe and cards used per round, under each condition

What a typical player does

Typical is not a random error rate. It is a fixed list of published deviations, applied consistently every time the hand comes up, so the profile can be inspected rather than trusted.

HandBasic strategyTypical playPublished costSource
Pair of aces against a dealer 10SplitHit as a soft 122.48 per 10 stakedBlackjackInfo
Soft 17, ace and six, against a dealer 7Double or hitStand1.59 per 10 stakedBlackjackInfo
Pair of eights against a dealer aceSplitStand1.28 per 10 stakedBlackjackInfo
Soft 18, ace and seven, against a dealer 9HitStand0.85 per 10 stakedBlackjackInfo
Hard 11 against a dealer 10DoubleHit0.60 per 10 stakedBlackjackInfo
Pair of eights against a dealer 10SplitStand0.52 per 10 stakedBlackjackInfo
Hard 12 against a dealer 2HitStand0.44 per 10 stakedBlackjackInfo
Soft 18, ace and seven, against a dealer 10HitStand0.37 per 10 stakedBlackjackInfo
Hard 12 against a dealer 3HitStand0.24 per 10 stakedBlackjackInfo
Any soft hand the chart says to doubleDoubleHit, or stand on a soft 18Part of the 0.53 percent totalWizard of Odds
Any hand the chart says to surrenderSurrenderHitPart of the 0.53 percent totalWizard of Odds

Items one to nine are the playable hands from the BlackjackInfo list of the ten most common basic strategy mistakes, with the published cost per ten units staked in a six deck game where the dealer stands on soft 17 and doubling after a split is allowed. The tenth item on that list, taking even money on a blackjack, is an insurance decision, and this simulator offers no insurance, so it is left out rather than approximated. The last two entries are the simplifications published in the Wizard of Odds simple strategy, whose total cost the Wizard of Odds records as 0.53 percent of turnover.

Why other players do not change your odds

In a freshly shuffled shoe the order of the cards is already fixed and unknown. Another player taking or refusing a card changes which unknown card reaches you, but every unseen card is equally likely to help or hurt, so the expected value of your own hand against the dealer is unchanged. This holds whether the other player is following the chart or ignoring it.

The one real exception is shoe depletion. Different play at other seats consumes cards at a different rate, which changes how far into the shoe you get before the shuffle and which cards remain when you get there. At a low deck count with deep penetration that effect is measurable, and the simulator reports its size from its own output for the deck count and penetration you set rather than asserting a figure.

Where the strategy chart comes from

Optimal seats read the same structured strategy data as the Blackjack Trainer, generated from the Wizard of Odds basic strategy calculator and the BlackjackInfo basic strategy engine and compared cell by cell. The two sources agree everywhere except a pair of eights against a dealer ace in a two deck game with the dealer hitting soft 17 and late surrender available. Both answers are accepted for that cell rather than one being picked silently.

Frequently asked questions

Does a bad player at the table cost me money?

Not in any way this simulator can measure at normal shoe games. Over thousands of shoes the win rate at your seat moves by a fraction of a percentage point between an all optimal table and an all typical table, which is inside the sampling error of the run. What does change is how many hands you play per hour and how fast the shoe runs down.

So there is no effect at all?

There is one honest exception. Other seats consume cards at a different rate, so at a low deck count with deep penetration they change which cards remain before the shuffle. The simulator quantifies that as a difference in rounds dealt per shoe and cards used per round, so the size of the effect is visible rather than dismissed.

Why does the third base seat get blamed?

The seat acting last is the most visible when a hand goes wrong, so its choices are remembered. The mathematics does not single that seat out. A card taken at any seat has the same effect on what reaches you as a card taken at the last seat.

Does a fuller table cost me anything?

It costs you hands, not edge. Each occupied seat adds dealing and decision time to the round, so a seven seat table deals far fewer hands per hour than a heads up game. The simulator states the timing assumption used and reports hands per hour for the table size you set.

Is any money involved?

No. Every seat is a bot, there is no input during a run, results are expressed in betting units, and nothing is stored or uploaded.

Related glossary terms

Related guides

Related pages