Dozens of performance metrics are used in managing a casino. These metrics are intended to help managers identify what they are doing well and what they are doing not so well that needs attention. Unfortunately, one of the most widely used of these metrics is the least understood and likely misused of all—table game “hold.”
Simply enough, hold is the percentage derived by dividing the game’s win by the game’s drop. A player first buys chips and then plays. The buys end up in the drop box, and the play generates the game win/loss. We can only guess how this metric evolved. Obviously, some analyst decades ago thought that dividing the little number (win) by the big number (drop) would actually provide some insight into the management of the casino. And the table games department has suffered for it since.
Historically, management has always equated high hold with good management and low hold as a sign of poor management or theft. Senior management believes operations can control hold, and this belief often wreaks havoc on operations. When hold is perceived as low, everyone is worried and carries a long face. If it continues, management panics. Shifts are rotated and executives may be replaced. Hold comes back up … it worked! Without question, more good managers have been terminated and bad managers retained due to this belief that management can actually control hold. Ironically, management has done its best to perpetuate this myth. When hold is high, management is eager to take credit. When hold is low, the players got lucky.
I am often asked, “What is a game supposed to hold?” But there is no “supposed to” when it comes to table game hold. Hold was never fully understood by management before the explosion of Indian gaming and, with tribal councils now reviewing their casino’s performance, this lack of understanding is only exacerbated.
Management’s objective should be to maximize profit, not to maximize this percentage called hold. High hold does not always equate to high profits, and neither does low hold to low profits. Although the examples in this article will draw from blackjack, the principles apply to all table games.
The mechanics of hold are simple enough: The player buys into the game, the player plays, and usually, the player leaves the game with something less than what he arrived with. This difference is casino win and, when shown as a percentage of what the player came with, this difference is hold. However, what actually affects the win portion of the formula? Every player visiting a casino has exit criteria. Some criteria are known to the player while others are not. For example, the player may exit at a certain time to meet a dinner engagement, catch a flight, or attend a show, etc. The player also has a loss exit criterion, which is generally known. Players should know how much they can afford to lose. Finally, the player has a win exit criteria. It varies from player to player, and is largely unknown until it is reached. All of these exit criteria affect hold, and management has little control over them.
Since the player is always constrained by time, the determinant affecting hold the most is table utilization or occupancy. That is, the average number of players at the game. The more players at the table, the fewer hands played by each player. Remember, the player first creates the drop and then management must deal as many hands as possible during this window of opportunity that the player has afforded the casino. A player playing head-up in blackjack will be dealt about four times the number of hands if the game is full. The more hands played, the larger the win and the higher the hold.
One of my favorite examples of this is that of 28 players who are each betting $100 per hand and buying in for $2,000. If you maximize labor, you want seven players on each of four games. If you maximize hold, you want one player on each of 28 games. (See Chart 1.)
The table includes (in bold) the five determinants of hold: players per table, house advantage, bet per player, rounds per hour and length of play.
Let’s first discuss the house advantage. The player chooses where to bet and, if it is a skill game like blackjack, how to play. In dice, there are more than 24 different bets with varying casino advantages. In blackjack, the player’s skill determines his disadvantage. If the house advantage goes up, so does hold. The casino has limited control over the house advantage.
Next, let’s look at the bet per player. What’s particularly important here is the bet-to-buy-in ratio. For this example, I am assuming the average player bets 5 percent of his buy-in. If that percentage is higher, hold will be higher. If the average bet is a smaller percentage, hold will be less. Management has no control over the bet-to-buy-in ratio.
Next is game speed or rounds per hour. If you had a faster dealer or some technology where a head-up player could play 220 hands per hour versus 209 hands per hour, hold would increase. The casino does have control over the game speed. However, this is a double-edged sword. If the dealer deals so fast that the player cannot keep up, the player will leave the game.
The next determinant is length of play. If the average player plays for two hours, the hold will be less than if he played three or more hours. The casino has little direct control over length of play other than keeping the customer happy.
Finally, let’s go back to game occupancy. The casino can control the average game occupancy by spreading fewer games. But here is the traditional hold dilemma: We want to minimize payroll, but we also want to maximize hold. What do we do? We increase the average game occupancy so there are fewer games open to service the same number of players. What happens? Hold goes down and departmental margin goes up. But, more importantly (and most unfortunately), profit goes down.
Management is further confused by the mistaken beliefs that:
1. High hold is always good.
2. Low labor cost is always good.
3. High table utilization is always good, because this means a high return on labor.
4. High departmental margin (as a percent) is always good.
As the 28 versus four tables example illustrates, these variables do not actually travel together. If management tries to maximize all of the individual variables listed, it only leads to confusion, and it certainly does not maximize profitability.
Sometimes maximizing return on labor is paramount, other times it is not. For example, if all the bettors were betting $2, management should focus on maximizing win per table. This would be accomplished by maximizing occupancy or the number of players per table. The casino should welcome nine players if they can fit comfortably. However, hold will suffer—and that’s OK. But if the players were betting $1,000 per hand, the casino should focus on maximizing win per player, which means minimizing players per table. Hold will be higher (also OK), and so will profit. Sometimes high hold is good. Sometimes it is not. Management should focus on maximizing profit, not maximizing hold. Often management interprets every game being full as good. That is only true if every game is open. If not every game open, but every open game is full, then profit was not maximized. If management focuses on minimizing labor, they certainly are not maximizing profit or hold. They are only maximizing return on labor.
The skill of the player also affects hold. A basic strategy player in blackjack will lose less, and consequently hold less, than a poor player playing the same amounts of money and time. Whereas the player can affect his loss through skill, so too can the casino affect the loss of the player by the game rules. For example, a blackjack game where the house hits soft 17 will earn more than the same game where the house stands on soft 17 if both had the same level of play. The house hitting soft 17 would hold more than the game where the house stands on soft 17.
The “trick” to a high hold would be to keep money out of the drop box. What if the casino had a chip cart continually circling the pit with the attendant calling “buy your chips here?” The hold percentage is going to be huge but the actual win unaffected. Some casinos drop their foreign chips in the drop box, while others keep the foreign chips in the float (tray). A policy that drops the foreign chips results in a lower hold than keeping the chips in the float. Management that focuses on maintaining a high hold is only encouraging the use of tricks.
Sometimes a low hold maximizes profit, and sometimes a high hold minimizes profit. If management would keep their focus on maximizing profit and keeping the player satisfied and playing, hold would take care of itself.