It’s no secret that land-based casino gaming is now legal in most U.S. jurisdictions. What may surprise some however is that with this gaming expansion came increases in slot machine hold (the win percentage of all amounts wagered that is retained by the operator), with some states experiencing a 50 percent or more rise in hold rates from 2001 to 2013. In this report we will examine some of the factors that have led to an uptick in slot hold and whether higher hold rates help or harm slot machine growth, profitability and customer satisfaction.
Casino gaming, like any other consumer choice, is a value proposition. In evaluating whether increased hold percentages aid or harm slot machine popularity and revenue, it helps to examine the reasons why slot holds have been on the rise—a list that includes operator response to consumer spending habits, technology improvements that have sped the pace of slot play, higher gaming tax rates and changing consumer entertainment preferences.


While slot hold has continued to increase over the past decade, consumer spending and gambling revenues did not follow the same trend. In 2007 gaming revenues began to trend downward ahead of the Great Recession, with consumer spending and gaming revenues both bottoming out in 2009. But consumer spending recovered and quickly rose past its January 2008 peak, while gambling revenue is just now approaching the high-water revenue mark it experienced in 2007 (see Chart 1). Gaming revenues have not kept pace with the overall growth in consumer spending, and casino companies have struggled to remain viable as consumer discretionary income shrank and the elasticity of demand for casino gaming proved that casino operations are vulnerable to economic conditions.
Meanwhile, since the Great Recession, new casinos have dotted the map in Kansas, Illinois, Ohio, Maryland, Delaware and Pennsylvania, leading to increased competition for the consumers’ gaming dollars. One operator response to this combination of casino expansion combined with stagnant gaming revenue growth has been to raise hold levels, an act that has taken place in several U.S. gaming jurisdictions (see Chart 2).


In addition to operator response to challenging market conditions, evolving slot play technology also plays a role in the rise of slot hold percentages. Coins have largely disappeared from today’s slot player experience; replaced by bill validators and meters that add credits to the game based on the bill accepted and denomination of the machine. Coins no longer fill trays when jackpots are hit, only credits added to the machine. The maximum number of credits can be played by pushing a button and the “wheels” spin automatically when the max bet button is pushed. It’s ironic that slots were once referred to as one-armed bandits because of the handle on the side of the machine; this lever is now rarely used and is no longer an option on many machines today.
These changes in technology certainly have contributed to the speed of play and perception that player’s don’t get as much time on device as they did when the coins were hand fed. While lower denomination slot machines generally have higher holds, some believe that shift in technology has accounted for the overall increase in hold as technology advances create a migration to lower denomination slot machines.

A 1980s era 5-cent Bally Slot Maximum machine had a maximum wager of 25 cents and top payout was $100. Coins were hand fed one at a time into the machine throughout a coin slot and all payouts were immediately paid in the form of nickels dropping in the tray at the bottom of the machine. The speed of play was determined by how fast you could put your nickels into the machine and pull the handle. While the 1980s version only allowed you to wager up to five coins, today’s version allows you to wager 90 credits, or $4.50, at the push of a button—18 times what was wagered on older games.
Adjusted for inflation that maximum 25 cent wager in 1985 has a value of 55 cents in 2015. Yet slot machines today are built to take as much as nine times that 1985 value. Notice the 1980’s 5-cent slot machine had three reels compared with the five reels on the 2015 model. The maximum 25 cent wager returned 400 times your wager for the top payout in the 1980s; today it returns 100 times your top wager of $4.50.

In a recent study by WMS (Englman, Allon 2014 WMS), a major slot machine manufacturer that is now part of Scientific Games, reviewed one of their most popular games in 2004, Super Jackpot Party versus one of their most popular themes in 2014, Raging Rhino. According to this research, a player’s median loss after an hour of play on Super Jackpot Party was $12.50 in 2004. That same hour of play on Raging Rhino in 2014 resulted in a player’s median loss of $48.00, an increase of almost 300 percent in median loss. To further illustrate this point, a survey of 100,000 players monitored in 2004 showed that more than 50 percent of the players were still playing after one hour on a $20 buy-in. In 2014 that same $20 only gets 10 percent of the players an hour of play on Raging Rhino. In 2004 after ten minutes of play, 100% of the players were still engaged in Super Jackpot Party. In 2014, only 30 percent of the players were still engaged after 10 minutes on Raging Rhino. This indicates a major shift in time on device, one of the key factors for customer satisfaction.


Increased corporate gaming taxes also play a key role in ratcheting up slot hold percentages. The tax environment that gaming is required to operate in can become a barrier to growth. While operators keep less and less of their profit, they are forced to try and generate more revenue to maintain operating margins, pay increased wages, incur increasing supplier costs for food and beverage supplies and try to maintain an entertainment offering with the newest and most desirable slot machine products.
When the casino operator is faced with shrinking margins from revenues that are shared at levels of more than 50 percent, they are forced to cut costs and increase the slot hold to try and maintain an attractive return on investment for the shareholders or investors who have taken on the challenge of investing.


Whatever the cause, what impact has this raise in hold had on the slot machine business as a whole? To analyze the effect of hold on slot profitability and growth, we reviewed commercial casino gambling revenue data provided by state gaming agencies (see Chart 3).
In Illinois, revenue data showed that as slot hold increased to an average all-time high of 9.0 percent, coin-in per unit has dropped to an all-time low in 2014 to $1.2 million. The overall decrease in coin-in per unit from 2001 until 2014 was 53.8 percent. While the overall average hold increased from 6.1 percent in 2001 to 9.0 percent in 2014, an increase of 47.5 percent. Win per unit peaked in 2007 at $175,891 and while the average hold was 7.3 percent. As hold has continued to increase to an all-time average high of 9.0 percent in 2014, win per unit has dropped to an all-time low of $109,149. This represents a 23 percent increase in hold from 2007 until 2014 and a 38 percent decrease in the win per unit.

In Indiana, coin-in peaked in 2006 at $1.5 million and the hold was an average 8 percent. As the hold continued to increase to an average of 9.2 percent in 2013, coin-in dropped to an all-time low of $1.1 million. The increase in the average hold from 2006 to 2013 represented an increase of 15 percent and the change in coin-in represented a decrease of 24 percent. The average win per unit peaked in 2007 at $117,595 with an average hold of 8.2 percent. While the hold increased to an all-time high in 2013 of 9.2 percent, the average win per unit dropped to $102,642. Although the average win per unit was not an all-time low in 2013, it represents a change of 13 percent from the 2007 high while the hold increased 12 percent.

For Missouri, coin-in per unit peaked in 2003 at $1 million when the average hold percent was 6.7 percent. When the hold percentage hit its all-time high of 9.4 percent in 2013, average coin-in decreased to $888,934. While the hold represents a 40 percent increase from 2003 until 2013, coin-in dropped 11.5 percent. Since 2001 the average hold has increased from 5.7 percent to 9.4 percent in 2013, an increase of 65 percent. At the same time, average win per unit has increased from $52,986 in 2001 to $83,349 in 2013, an increase of 57 percent. This is unlike other states we examined; one possible explanation for these results may be the repeal of the $500 loss limit (the maximum loss permissible per two hour gaming cruise) in 2008 and the smoking ban implemented in Illinois, which benefitted the St. Louis area casinos.

Pennsylvania is relatively new to commercial casinos and has a short history. In 2008 when gaming was introduced the average slot machine hold was a relatively high 8.7 percent. In 2009 average hold was 8.8 percent and average coin-in was $1.2 million. By 2014 average hold was at 9.9 percent and average coin-in decreased to $883,832. In six years the hold increased 12.5 percent and the average coin-in decreased 25.5 percent. Starting in 2009, average win per unit was $104,212 and hold was 8.8 percent. In 2014, win per unit was at an all-time low of $87,293 and average hold was at an all-time high of 9.9 percent. The change in average hold from 2009 to 2014 was 12.5 percent higher while the average win per unit decreased by 16.2 percent.

The script of increased hold leading to reduced coin-in and win per slot machine was flipped when examining results from Nevada and the Las Vegas Strip. While the win per unit from Nevada seems small by other state standards, it is distributed across approximately 265 casinos in the state. In 2001 the average hold was 5.3 percent and the average win per unit was $31,626. In 2013 the average hold was 6.4 percent and the average win per unit was $43,081. While the average hold increased 21 percent, the average win per unit increased 36 percent.

On the Las Vegas Strip, the average 2001 hold was 6.2 percent and the average win per unit was $38,693. In 2013 the average hold was 7.6 percent and the average win per unit was $68,844. While the average hold increased 22.6 percent, the average win per unit increased 78 percent. The Las Vegas Strip suffered during the Great Recession and is still below its peak of 2007 but has made significant strides toward a full recovery. Nevada and Las Vegas have average hold percentages significantly below the regional casino markets and also benefit from the lowest commercial gaming taxes in the nation.


To determine the impact slot hold was having on player perception of the slot machine experience, we turned to some recent consumer preference studies. In August of 2014 an online survey was conducted with the assistance of co-author Dr. Roger Solano to a sample of 5,188 known casino slot players from a Midwest regional market. An e-mail invitation to take the online survey was sent to each addressee and all respondents submitted their answers online.
One of the primary complaints of the casino customers who were surveyed was the perception that slot machines are too tight, take their money too fast or should provide more and better payoffs. One in four respondents took the time to write their concerns about this particular issue.
In another study that was conducted to establish whether players could tell the difference between the hold of a slot machine or not, (Lucas, A. F., & Singh, A. K., 2011), the researchers found the following regarding the game of Lobstermania, a popular reel slot theme: “In Lobstermania, a player wagering $1.00 per spin would lose, on average, 3.8 cents per spin on the 96.2 percent game [refers to the payback percentage for that version of Lobstermania] and 15 cents per spin of the 85 percent game [a second version of Lobstermania]. Thus, the player loses approximately four times more money per spin on the 85 percent game than on the 96.2 percent game (15 / 3.8 =3.95). A player arriving with a ‘bankroll’ of $10.00 and wagering $1.00 per spin, who gambles until the bankroll is depleted, would make, on average, 263 one dollar wagers on the 96.2 percent game ($10.00 / $0.038 = 263), but only 67 one-dollar wagers on the 85 percent game ($10.00 / $0.15 = 66.7); thus a player with a specific bankroll would have approximately four times more gambling time on the 96.2 percent version versus the 85 percent version (263 / 66.7=3.95).

“As slot hold moves higher, time on device declines as customers lose their money faster. We estimate that in 2004 a player could budget $100 to play slots and gamble for 3.2 hours on a 75 cent-per-spin bet. With higher holds, $100 may only last 2.6 hours, a 22 percent decrease. If we assume the average bet moves from 75 cents to $2 for the most popular games, the average time drops from 3.2 hours to one hour.”
Based on the information above, it would be easy to conclude that the higher the payback percentage, the more likely we are to get more value for our money because it will likely result in more time on device.


There is no question that the entire gaming industry is concerned about the future viability of the slot product. Every month there is new research being undertaken to try and understand the migration away from the casino product by the lower level slot players. Our research shows that there is no one factor that answers that question. It is a series of events that have combined to create the perfect storm. The improvements in technology including the credit meters, bill validators and coinless play have all increased the speed of the game. The challenges of operating profitably in tax environments where many states collect more than 50 percent of the casinos revenues are a factor. Customers are also starting to question the entertainment value of their gaming experience and whether they can have more fun with those dollars in some other form of entertainment.
Revenues across the board in all states have seen the Win per Unit decline and overall gaming revenues decline as the hold percentage continued to climb. A player is sensitive to the value equation when they see their entertainment dollars disappear all too quickly. Slot machine manufacturers ponder why their new product isn’t selling—shrinking revenues can impact the ability of the operator to provide the latest themes in slot machines to the floor. Our slot customer survey indicates the perception of a feeling that players don’t win very often and their money goes too fast. When penny machines play like dollar machines they should represent a hold percentage consistent with a dollar game. Even today, operators place the highest pay tables on the penny slot even though the player can wager over $2 per spin.
When we consider the tax rates, technology, customer perceptions, manufacturer’s challenges and the need for business to operate profitably, we can only conclude that slot machine hold is defining the entire experience, and in the end, the customer will speak with their discretionary income and tell us whether we are meeting their expectations or not. SM&M


Rob Guthrie began his career in gaming as a dealer in 1985 and has worked in casino operations in Nevada, Indiana and Pennsylvania, serving in positions of increasing responsibility throughout his career. He is currently employed as the director of table games at Rivers Casino in Pittsburgh, Pa.

Dr. Roger Solano, PhD., professor at Slippery Rock University, served as the academic advisor and co-author of this project which served to fulfill the requirements for Guthrie’s MBA, awarded in 2015 at Slippery Rock University, Slippery Rock, Pa.


• Englman, Allon (2014) Keeping Players in the Game, Breaking the Cycle, presented by WMS, a Scientific Games company.
• Lucas, A. F., & Singh, A. K. (2011). Estimating the Ability of Gamblers to Detect Differences in the Payback Percentages of Reel Slot Machines: A Closer Look at the Slot Player Experience. UNLV Gaming Research & Review Journal, 15(1)