This post was first published on my blog where I plan to regularly discuss game design and monetization topics.

As aspiring mobile developers are looking at the app stores for inspiration, there is no way any of them missed the Finish miracle Supercell. The company has dominated the strategy tycoon top grossing charts for over two years with two quite similar and yet profoundly different strategy tycoon titles: Clash of Clans (soft-launched in June 2012) and Boom Beach (soft-launched in November 2013). In this series, I will juxtapose the two games and attempt to reverse engineer the reasoning and thought process that led Supercell to implement changes in their second endeavor on the strategy tycoon front - Boom Beach.

Following the HHH framework outlined in my recently published book - Freemium Mobile Games: Design & Monetization, I will first compare the designs of the hook, habit and hobby phases of both games and only then examine what the new mechanics and game design elements actually mean for the user experience and the game's retention and monetization. In an attempt to keep even the casual readers involved, I will split this rather detailed investigation into four parts. The first three will examine each game phase separately, detail all the changes and Supercell's rationale for implementing them. Then in the last summarizing part I will criticize Clash of Clans' mistakes, discuss how some of those mistakes were partially repaired in Boom Beach and exhibit the most brilliant and unfathomably dull designs in both titles.

The Hook

Building on all their Clash of Clans user behavior data, the backbone of the first time flow of Boom Beach tightly follows the Clash of Clans first time flow: getting attacked, building a base, building an army and attacking in return to conclude the tutorial, but the devil lies in the details and there are several seemingly minor differences that make Boom Beach's first time flow better polished, an effect which is unfortunately countered by Supercell's flawed decision to remove a crucial for the hook phase element - the high conversion item. Let us start at the very beginning…

The entry point has been changed from green goblins that attack you in Clash of Clans to a name-yourself window in Boom Beach. This is important because the ritual of naming a virtual city or character inevitably gets players emotionally invested into a game even if they have not played that game yet, as in the case of Boom Beach. Keeping the very same backbone of actions and the name-yourself window as a part of the initial flow tells us that the trick of naming the virtual character / base for emotional attachment definitely worked for Supercell and they have kept it in order to improve the game's early churn, retention and monetization. Switching the name-yourself action from much later in the first time flow (second half of the Clash of Clans' tutorial), towards the very first action (as done in Boom Beach) tells us that they too have observed, similarly to my own experience, increased churn at that action. In discussions with colleagues that deal with very different game genres and very different audiences, this seems to be a common sighting experienced by all titles that include the naming process as a part of the tutorial. Although I have heard many theories regarding the spike in churn when players are faced with the name-yourself window, the most likely cause is that bringing the keyboard on screen breaks the game flow and forces users out of the virtual domain the game had enveloped them in, essentially providing them with a clear-cut exit point. Couple this with the fact that naming your virtual self in a game requires mental effort to recall a suitable name and a physical effort to type it and the user behavior bottleneck at that point becomes much clearer. The solution chosen by the Boom Beach team is as simple and as effective as they come. By switching the naming process to the very beginning of the first time flow, Supercell simultaneously attaches players emotionally to the game and minimizes the probability of their exit as at that point they do have the cognitive capacity and curiosity to complete the task. A brilliant move, especially for a company that has already made a name for itself and the process of typing a name will not stop players from testing the third big title it releases.
 

There is a big difference in terms of tutorial length between the two games. While the Boom Beach tutorial is usually finished in less than 2 minutes, the Clash of Clans tutorial lasts for more than 6 minutes. This shift in tutorial length can mostly be explained with the maturation of the audience between the release of the two titles and one should really appreciate the adaptive capabilities of Supercell. While at the launch of Clans of Clans some of the game's mechanics constituted a novelty for the mobile market, at the end of 2013 when Boom Beach first launched the mobile gaming market was much more developed and the audience was a lot more sophisticated. Back in mid-2012 the casual audience was considered the main revenue stream for any title, but later on Game of War and Clash of Clans itself showed that making big money from mid-core players is a sustainable business venture. This new perspective (and probably unsatisfactory results from the lengthy Clash of Clans' tutorial as of late) inspired a major shift in Supercell's appropriate introduction to the mid-core players and brought about a solid decrease in tutorial length achieved via three key changes in Boom Beach.

1) Tycoon is almost non-present in Boom Beach's tutorial. While in Clash of Clans, half of the tutorial deals with building resource generators, resource storages and barracks, in Boom Beach more time is devoted to the story (which surprisingly for a freemium game, somewhat makes sense) than to its tycoon. Most of the player's buildings present at the end of the tutorial are pre-built at start and there is no mention of any resource buildings at all. Talking about collectors and storages is not only unnecessary for newcomers, as the vast majority of players today have a perfect understanding of all the elements that constitute an economy, but it is also harmful to the game as the more sophisticated audience today might perceive a hand-in-pocket attitude as early as the first time flow and this will inevitably cause a churn spike right then and there.

2) Waiting times are removed from the Boom Beach tutorial. The total waiting time during the tutorial in Boom Beach is 16 seconds and the speed up of even these few seconds is free of charge, i.e. it costs 0 diamonds to finish buildings and army construction instantly. Compare that to Clash of Clans' tutorial where more than 40 minutes of waiting time and high speed up prices are lurking and once again Supercell's brand new philosophy for avoiding a hand-in-pocket impression becomes obvious. Instead of letting the sophisticated audience know that a lot of waiting times lie ahead of them and a lot of premium currency will be spent on it, Boom Beach focuses on delivering an ultra-short, deeply satisfying tutorial impression that emphasizes the game's action phase.

3) Time spent in action phase during the tutorial has been halved. In Boom Beach's tutorial, around 40 seconds are spent inside the game's action phase in total versus close to 80 seconds in Clash of Clans. The longest tutorial battle in Clash of Clans takes around 40 seconds and the one in Boom Beach takes around 30 seconds. This obviously is misleading, as I would estimate an average Boom Beach battle takes at least 50% longer than an average Clash of Clans' battle and the maximum allowed action phase time - 4 minutes in Boom Beach versus 3 minutes in Clash of Clans, clearly supports this notion. Boom Beach fighting simply requires a lot more interaction and strategic decision-making which takes more time to execute, but to keep the tutorial brief Supercell has had to somewhat mislead users.
 

20+ workers on screen and not one of them needs the extra pay!

The session end is implemented in a novel way that accommodates both user preferences and freemium laws. Once the games' tutorials are over, one major difference between the two titles becomes immediately obvious - Boom Beach has a clear session end, a point at which Supercell tells the user it's time to stop playing, while Clash of Clans does not. The successful practice of insanely long game sessions from Clash of Clans has been retained and rightfully nurtured - it is still possible to protect your Boom Beach base from enemy raids by staying online for days or even weeks, but going from one action phase to another throughout such long game sessions has been made impossible in Boom Beach in a brand new way. While in Clash of Clans the only obstacle to jumping from one action phase to the next is the time and cost of training new troops, this is also the number one complaint of Supercell's fans and this has been addressed accordingly - all surviving armies in Boom Beach are immediately available to fight again. As the ironclad rules of freemium dictate though, players must be stopped from binging on a game and the way this is done in Boom Beach is noteworthy - Supercell has introduced a complete monopoly over the enemies players can engage as all opponents are automatically generated / chosen server-side when the player is not looking at the map. The only way to get new opponents once all the old ones are defeated, is to improve the radar building which requires to upgrade the headquarters first, which requires to upgrade the resource storages before that, etc. This is why the point at which the enemies on the map are over is the point at which there is practically nothing else left to do in Boom Beach. The only way to continue playing is to start paying with hundreds and thousands of diamonds for upgrades in order to move the radar along with the whole base forward. With this new game mechanic Supercell revolutionize freemium game design once again - they simultaneously address their fans' biggest concern by making units transferable from one action phase to the next and manage to prevent players from binging by providing an unseen before session limiter.