By Adam Alsen, Julian Runge, Anders Drachen and Daniel Klapper

Is social gaming merely a marketing tactic? The point has been made (e.g. Liew 2009) and was welcomed by parts of the community where long standing gamers like(d) to see casual social games (CSGs) as crippled social media versions of authentic, deeply immersive games. We posit that this account is incomplete and that CSGs are, more often than not, characterized by engaging social gameplay that can achieve a lot more than simply act as a catalyzer for the viral spread of a game.

Before we get to empirical evidence for this notion, we want to briefly take a retrospective look at CSGs. The advent of social networks such as Facebook in the 2000s established a wholly new platform for marketing that many industries now make ample use of. An industry that reached a new level of widespread success in this new medium is gaming. Previously, games were confined to dedicated websites or consoles where the target audience had been limited to “gamers”. Facebook changed that – suddenly everyone was logged into the same platform, which also happened to be a platform that allowed the distribution of games. It was possible to achieve tremendous viral spread among gamers while also targeting an entirely new audience that hadn’t previously been involved with (online) gaming. Social gaming was born (see figure 1).

Figure 1: Google Trends data (i.e. online search interest over time) for “facebook” and “social gaming”; “social network” and “social network game” are not included because they show biasing kinks for the release of the Social Network movie; source: Google Trends, accessed at https://www.google.com/trends/explore on January 5th 2016

New companies achieved massive success building CSGs, with the most prolific example being Zynga which achieved annual revenues in the hundreds of millions thanks to games games such as Farmville and Mafia Wars on Facebook. King, with their mega-hit Candy Crush Saga, initially launched on Facebook and later ported onto mobile platforms, reached $1.33 billion in annual revenue by 2014. A plethora of companies followed suit: Wooga, Socialpoint, SGN, Big Fish Games – to name but a few.

The unprecedented viral spread and casual nature of CSGs has fueled their image of being business tactics rather than actual games. In this view, the social features included in these games, e.g. comparing your progress with your friends, sending and receiving in-game items as gifts, or competing against other players in tournaments and leaderboards, exist only to promote viral spread of the game. To understand if this account of social gameplay mechanics in CSGs is correct, we investigated experimental data from the introduction of a new social feature in Diamond Dash; a highly casual puzzle game developed by Wooga and one of the early stars of social gaming. If social features merely fuel viral spread, they should not afford meaningful social interaction to players, hence leaving engagement with the game largely unaffected. Let’s see if this is actually the case.

The new feature introduced was Team Battles. Team Battles let players form teams to compete against other teams, fostering both cooperation and competition. The impact from this feature on both mobile (App Store) and browser (Facebook) is larger than anything we have seen from other changes to the game. Revenue went up by roughly 80%, active players by 5 and 10% respectively and sessions per player by roughly 10%. Most importantly, these effects were persistent, too. Beyond its novelty, the new feature caused a structural change to the way players interacted with the game and with each other.

Figure 2: Revenue per day in Control (without Team Battles) and Test (with Team Battles) group; data from Diamond Dash on Apple’s App Store (further charts in Alsen et al. 2016)

In essence, our results show that social features in CSGs are far from mere tactics for viral growth - they can greatly impact the way players engage with a game. In fact, the impact from this particular social feature was larger than that of any other feature that had been added to the game in the past.

In light of the tremendous revenue impact of added social gameplay it seems that developers should strive for more social interaction in and through their games. In fact, they do. Virtually all top 100 grossing games include Facebook Connect and King, the arguably most successful  maker of CSGs, is building its own social network inside its game portfolio. The games research community is well aware of the trend: Malaby (2007), Shumaker (2011). The importance of social functionality is likely to gain further importance as CSGs expand to new platforms and technologies (wearables, VR and AR). On the game analytics front, we need more Social Game Analytics to guide the way.

Questions abound: What are the effects of adding social features to casual games? How does Facebook Connect affect your player base (we previously wrote on that)? How should we measure engagement with social features? And how do these in turn shape organic and viral spread of a game?

Surely you want to know more about the analysis underlying this post. Luckily, we summarized it in a brief paper that was presented and discussed as part of the Player Analytics workshop during the AAAI Artificial Intelligence in Interactive Digital Entertainment 2016 conference. You can find the paper here. Your comments are most welcome. Let’s get the Social Game Analytics ball rolling.