When I joined a video slot game company, I thought it would just be like learning any other genre. I'd made games in just about every genre. I thought video slot games were simple because the interaction was limited to pressing a spin button and watching to see if you won. Raph Koster said in his Theory of Fun for Game Design that video slot gameplay was basically broken. Of course he hadn't made any slot games, like me at the time, so of course we were very wrong.
Slot Players are NOT Mindless
Slot players are gamblers. While the physical interaction of hitting a SPIN button is simple, the mental engagement is complex. Gamblers base their decisions on what they perceive as the potential of a math model – in terms of how the game features work and what sort of wins or near-wins they experience. When all the games in a casino have to fall with a certain return to player percentage (RTP), how the players win and how often they win become much more important as differentiators. How else would some games become more popular than others? This concept is expressed in basic slot math terms as VOLATILITY and PAY DISTRIBUTION.
It's the Art, then the Math
There is a saying in slot development, “The art is what attracts the players, but it is the math that keeps them.” It was told to me on my first day on the new job, and it's something I told every new hire or candidate afterward.
Like a storefront with a bad sign or a book with a bad cover, a very good math model can suffer if the art does not attract new customers. Yet what theme attracts one person may be entirely different than what attracts another. So not only is the art important, but it also needs diversity to ensure a wider net is cast. Putting a new art theme on the same math is called a SKIN. A new promising math model might be release on multiple skins.
When I first started, we had a knee-jerk reaction to any time the lead theme tanked. The assumption being that a good math model could withstand poor opinions about art. When we had a dud on an original math game, we'd pull skins in development off that math model and try to fix it or adjust the art to a different math model, which of course ate up time and killed our roadmap goals.
Then one time, a perfectly well executed Chinese-themed game tanked within the first month of its release. A skin of that math done by a completely different team with a unique pre-rendered 3D art style was already in the final stages of deployment. We were loathe to change it because it had just gotten approval by GLI, the industry watch-dog, and marketing and sales were eager to sell it. So we released the skin, and wouldn't you know, it took off. Same exact math, same exact features, rendered differently produced widely different results. We had a hit, and the game was deployed everywhere in banks of slots.
Yet ... as the saying goes, "it's the math that keeps them." Eventually people grew tired of the math. The art couldn't sustain them, all it could do was attract more players. We tried skinning the math some more to give it new life, but none of them faired as well. Lessons learned? Don't ever rely on a single theme to determine the viability of a math model, and don't underestimate the value of art to create a hit.
So why not just skin a good math model over and over again? Some games are skinned a lot to keep them alive, but do it too much and it would be like having a block of various restaurants that all serve the same food or a shelf of books all telling the same story. Players have different tastes and eventually get bored of their choices.
When a player checks out a slot game, they look at the art first, then take in the game features – usually advertised with branded logos and feature text on the top screen or glass or in the splash screens for online. These features are what makes the math different. They are what’s patented. They are what’s branded and trademarked. They are what makes the VOLATILITY and PAY DISTRIBUTION unique. When a feature is successful, it may be re-used in combination with other features or on a different set of reels with a slightly different volatility and pay distribution.
A feature is often a twist on a slot rule, which makes it look like the game is more likely to pay in your favor (it's not, it's all hocus pocus). It could be turning symbols wild , respinning or nudging the reels or paying side bonuses or increasing the number of free spins or picks, etc. Whatever it does, it doesn't change the overall return percentage (RTP), just the perception of your odds of winning.
This is the interaction, the engagement with the gambler, that makes slot games more complex than the simple SPIN button implies.
A player who sees a feature they like or have never experienced may give a new game a chance. That feature may be expressed differently in the game because of how it was combined with other features or unique reel layout. If they determine that it is only a SKIN of a game they played before, they are less likely to try it out, and also less likely to be curious the next time a new game arrives, because players checking out the machine with the NEW GAME banner over it are looking for a new experience.
For this reason, developers prefer to make subtle iterative twists on successful math models rather than doing purely skins. Skins are useful to fill a bank of machines with your already successful math model, but they don't give you a new shot at a home run. Thus each original game, each new twist on a trending feature is a new "at bat".
Development companies like casinos want ALL of their math models to perform excellently. They DON’T. There is NO ONE SINGLE SECRET RECIPE – partly because of player diversity but also because players get bored and they look for something new. A good performing game may pay the bills for a while, but eventually the sales decline. Companies execute a scatter-shot approach – a combination of innovative designs, derivative math models, and twists and skins on older good performing math models. They constantly play the competition, read sales reports, chase down successful trends, dream up and patent new ideas, and clandestinely reverse engineer or steal each other’s math*.
* In no way am I implying we've done this!
Thus an important metric for all slot companies is how many original math models they produce. The big boy slot game companies aim for around 50 to 100 new slot games a year, and easily one third to a half of them have new original math or twists on successful math models.
The Arms Race
There are new math features dreamed up every day from innovative designs to unique combinations and twists that add up to a new slot experience … or maybe not always new enough. These ideas are patented, implemented and defended in an extremely litigious and competitive market. IGT created a Vegas strip hit with “Wheel of Fortune”, and Bally acquired a wheel bonus patent and sued them (and lost), invalidating the patent and opening it up for every other manufacturer to put out a wheel game. Aristocrat had huge success with the “Reel Power” feature, just about everyone copied them, and they sued (and also lost), since it was really a variation of scatter pay, a common feature. Cadillac Jack (now AGS) put out Streaming Stacks, a random symbol stack feature, and got sued by Konami for violation of their patent, and it was settled out of court and spurred Cadillac Jack to produce more patents. IGT (not GTech) and WMS (now SciGames) constantly sued each other, often exchanging patent licenses as a settlement. These companies all have a war chest of game design and hardware patents to protect themselves.
I was shocked when I discovered how these slot manufacturers tried to protect their designs with patents. Just imagine if Westwood Studios had patented the RTS base-building and unit construction mechanic. Westwood (and eventually EA) could have sued Microsoft over Age of Empires and just about every other RTS game publisher. Imagine if ID had patented the first-person shooter to keep Epic, Valve, Activision, Ubisoft and EA from ever producing their hits. Imagine if Origin patented the MMORPG. We might never have seen Everquest or Worlf of Warcraft. This was what the slot manufactures were effectively doing, or trying to do with their patents. The climate that I stepped into was HIGHLY litigious, but also devious in how they worked around these patents. Thus I spent many of my 6 years patenting ideas, my own and others, just to give us some elbow room at the table.
Pacing in a Slot Game
There are a lot of great articles on Gamasutra about the importance of pacing. They all boil down to how a player needs to have periods of low action to build up tension followed by bursts of intensity. This puts players on an emotional roller coaster. A poorly paced shooter might throw enemies at a player constantly, making each engagement blend into the next - making for a very intense but ultimately boring experience.
Pacing in a good slot game is similarly driven by a roller coaster of emotion, which maps to the VOLATILITY of the math model. A player has to lose as much if not more often than they win to make the winning feel better. These spaced-out wins also tend to be bigger.
Games that constantly award small wins deaden the experience. These are called DRIBBLERS. There are players who prefer these kinds of games, though casino operators don't because they lead to very long TIME ON DEVICE, or SEAT TIME on the same twenty or forty bucks they cashed in with. Players who like DRIBBLERS can get a whole evening's entertainment for the price of going to the movies, so it's not surprising that these tend to be popular at the bars and local casinos.
GAMBLER math models are higher volatile. These games can suffer from long stretches of zero or very meager wins before hitting a very big win or triggering a bonus, where players can get their money back. Time it right and with a little luck you could beat the odds and hit the big wins or bonuses back to back and walk away with stacks of cash.
Of course, most players tend to just put the money back into the slot machine, like it's free money, completely forgetting the fact that they may have lost hundreds in the previous trips to the casino. This is how the casino makes money (and how my job was so stable).
Presenting the Win
A huge mistake is to oversell a small win. Similar to pacing, if a player is constantly getting grand ring-ups, symbol animations and bells & whistles, then they make every win feel the same. Players will speed past huge wins and not even notice. For this reason, a trend in slot games is to not even animate the winning symbols if it is a low win below 2X (twice your bet), then have a huge fireworks show of coin and gem fountains when you win 20x or more.
Just triggering the bonus, something players will wait pateintly for because that is where the math models often pay out, will often get a special presentation. If it takes 3 symbols to trigger a bonus, games will present special effects during the spin, playing a bonus marker animation and sound on the first symbol, another louder one on the second, then throwing a special animation on the subsequent reels, highlighting and speeding them up and adjusting the timing of the reel stops to really build up the anticpation. Then if it hits, a bell will ring - a piercing sound that is common to all games when a special win is presented to a player.
Like Pavlov's dogs, players learn to listen for these sounds and look for these effects in anticipation of a great award. They'll keep coming back to game to recreate that experience in hopes or repeating their win or getting a much better win.
Other players will hear the bell, and look around to see which game is paying out. They may pause at what they're doing to see the credits ring up on the top screen (there indeed for the bigger audience), and they'll make a point of playing that game to try to recreate that experience for themselves.
In this way, games are constantly vying to sell their winning experience, and presenting the wins in proportion with the size of the win is critical to a game's success.
Understanding the Demographic
This isn't new to game developers, but what is new to them is that they might not be the target demographic. Slot games, particularly at brick & mortar casinos rather than online or on mobile social slot games, tend to be played by older, female players (50+). A median player would be a 53 year old woman. So they're not going to necessarily be interested in the latest trends in game console graphics or CG or attracted to the same sort of themes that you as a developer would like to work on.
When I first joined my company I worked with a bunch of artists an animators who would love to be working on a Pixar movie or console shooter, but the reality was that their average age (mid 20's) was not the average age of slot players. Is it any wonder that the cartoony games and the young-adult steampunk themes didn't fair well?
I pushed for 3D graphics, because that was the background I came from, but of course that wasn't necessarily what made a 50 year old grandmother tick. I'd had some success with 3D slots ("Goddess of the Woods"), but not with all 3D games, and given that they took twice as long and often three times more effort to produce, it was hard to convince management that pure 3D made fiscal sense. They'd rather see two or three games ("at bats") of a simple 2D illustrated theme than a single 3D rendered game. They saved the 3D for the more premium or licensed titles.
This point was driven home by another game of mine ("Gold Dragon Red Dragon"), done in 2D with After Effects and Photoshop that was produced and tested in under a month and sold way more than games that took much longer to develop.
Of course, dragons are almost always popular because they represent fantasy as well as wealth (for the dragon horde), and we put out A LOT more dragon games after Gold Dragon Red Dragon.
Then there are license titles. Companies like IGT, Aristocrat and WMS pumped out games with film, television and celebrity tie-ins. Yet if they don't overlap the target demographic like "Wheel of Fortune", "Sex in the City" or "Monopoly" did, then they don't do well. Even though it had a very famous dragon, the "Dragon's Lair" game theme I produced, though popular in the 80's arcade era, didn't translate well to the 50+ female demographic, and despite being placed on a premium math model and later as a skin on our most successful math model at the time, it just didn't measure up to expectations. We had over-estimated its popularily with slot players.
Of course, preferences evolve and fads come and go. One great math model on a theme can bring new life to that theme. Asian themes have come and gone and come back again, until there are so many that they become boring again. High 5 Games with IGT pumped out themes with air-brushed realistic characters, often photo-touch-ups, with top screens that looked more like romance novel covers than slot games. Given the popularily of the romance genre, should we be surprised that they did well with the 50+ ladies? All the other companies starting doing the same thing, once again flooding the market, so that the odd cartoony or sci fi theme now stand out from the rest. In this way, themes and art styles come and go, but always with a finger on the pulse of the latest trends.
The Importance of Research and Feel
Playing competitor games is imporant. This is true for regular video game designers. While they can get accused of producing derivative games, they must all learn to be players first. But just like video game designers, slot game designers, while heavily skilled in math, still have to get a good feel of what players want and are thinking. Any good mathematician can make the game hit its intended payback percentages, but a good slot game designer will put in the subtely, the near-misses, the balance of wins vs losses and the frequency of the features and bonuses to produce a winner.
Feel is closely tied to a gambler's expectations. If it feels like the game has the potential to pay a lot, and it occasionally does, players will keep playing and coming back to that machine to repeat that experience.
How brutal math model feels can be expressed by its volatility. Players can feel if the math is draining your credits away, but if it does so with a few near misses on a feature or bonus mixed in with a few decent 5x and 10x wins, then it feels like the stars might align one day and leave you ahead - cashing out with your next car payment or rent check covered. If you never see any of those wins, because it is such a highly volatile math model that only 1 in 5 players (or sessions) experience the wins, then it becomes a very brutal, very short TIME ON DEVICE game.
Casinos tend to like these brutal games because they get your money faster, but they get far fewer return players. This is perfect for a destination casino like on the Vegas Strip or Atlantic City where players only visit once or twice a year, versus the local casinos and bars with lots of return traffic week after week which prefer less volatile math. Manufactures will often offer different volatilities and payback percentages of the same game to different casinos to account for this. When you are a slot designer, you need to play the games at the casinos you hope to sell to get a better feel of what their players are looking for.
Social Casinos & Conclusion
As my career has gone from brick & mortar, and online real money casino games to Facebook and mobile social casino games, I try to bring these lessons learned. Yet I have found some significant differences.
Mobile players are a different demographic. They tend to be younger and more male. They tend to play more frequently but for shorter sessions. They play while watching TV after dinner, while on break at work or while sitting on the toilet.
Paying real money to win and lose fake money still amazes me, but some players are willing to do it, so it fit very well into the Free-To-Play micro-transaction and Freemium models that rely on whales of 1 to 2% of your paying players to pay for all the rest. As it turned out, the math models that make casinos rich also gets players to lose chips and penny up for some more.
Players LIKE winning, even if it's of no real world value. That's proof that a slot player isn't broken, like Raph Koster said in his book. Slot players get something out of winning, even if it's just a result of hitting SPIN.