The latest Clan Wars update dropped for Clash Royale this month adding a new competitive mode that pits clan vs clan in a 2 day competitive event.  Clash Royale, the poster child of innovative mobile battle arena gameplay has been losing engagement and viewer across YouTube and other streaming platforms. In fact, Clash Royale’s quarterly revenues are around a half of what they were just a year ago.  The games industry is a fickle place where gamers quickly switch between titles to play whatever is hottest at the time. Clan Wars is Supercell’s attempt to lure loyal fans back and give them a new reason to play the game?

Clan Wars – What is it?

A clan war is split into two phases. The first is a 24h period called “collection” in which each clan member has the ability to play 2-3 different match types against a random selection of players.  These match types will be familiar to anyone who has been following Clash Royale for the past year or so in that they mirror the common challenges that are sometimes used for events. So far I’ve seen, double elixir, draft and 2v2 and in each case you can pick from your own cards or draft although they use Tournament level stats.  If you win your match you win a Clan Chest that opens into a central “clan deck” of cards that the clan can then level up and use to build a deck for Battle Day. The more players that compete in the collection phase the better the card selection and the higher levelled the cards.

The second day is called Battle Day and now players may only select from the cards they have won at the level that each card is currently levelled up to.  This creates a tactical discussion session where deck crafting from the cards that have dropped can be aided by the clan themselves. You only have one chance to succeed, with each clan member having a single battle.  If they win, they add a sword to their leaderboard and of the 5 competing clan, one clan will be the overall winner and gets a larger chest. Each clan war lasts around 2 days.

Clan Wars – Why do it?

Clash Royale is a mature product and with it there is a mature audience. Most of the players with over 3000 crowns or more will have been playing Clash Royale for 1 or more years. They will probably have somewhere between 80-100% of the cards unlocked and will be grinding the top levels of those cards. This can take months per card. At this point there is a large amount of player fatigue and moving people from new or different challenge keep people entertained. As I spoke about in my GDC talk a game team wants to establish a Lord of the Rings Metric (one KPI to rule them all). An update at this stage should focus on driving the most interaction with that metric as by improving this all other metrics will tend to trend upwards. For Clash Royale I would argue it is Battles per DAU which leads to more engagement in all their other systems, spenders will most likely have a high battle count.

As well as trying to improve your internal KPIs, the mobile marketing is a constantly changing battleground. External factors can have huge effects on your games bottom line and although unpredictable the one thing you do know is that changes will happen.  Just before this update released – Fornites launch has disrupted and captured 80-90% of the market share relative to the previously steady state.

Problems it could solve

Clash Royale is a fantastic mobile game. It creates subtle depth in it’s characters along complex strategic decisions which requiring a low number of active touch inputs but precise timing. This is what makes it so playable on mobile.  However when a game is designed to for many years there are often issues that long term players get stuck in that are not apparent in the short or mid term. Clans Wars has the potential to alleviate some of the more complex issues that most long term players of the game might be familiar with these are:

1. The Archetype Problem
2. The Yoyo Effect
3. Clan Engagement

1. The Archetype Problem

In all competitive games balance becomes one of the major discussion areas for players. Balance issues usually involve overpowers stats of a single gameplay element, a broken combination of 2 or more elements or a game-breaking unforseen bug when using the element.  In each case the collective of gamers dissect and discuss how to gain micro-competitive advantages with their setups – we call this the metagame (the game outside of the game).

An Archetype is usually a combination of cards that all work well together and therefore commonly get picked and combined together.  They are the most powerful implementation of cards and so by using them you are more likely to beat your opponent. Some decks that a Clash Royale player might recognise would be Lavaloon, Mortar-Hog, Bridgespam etc. When you’re a new player you strive to be able to make certain decks that contain legendary cards that you might not own and this drives you to engage more fully.  In each case the decks usually feature 6 core cards and then rotate 2 cards depending on the players taste. However from a long term players perspective, the issue is the boredom of playing the same archetype again and again.  

The challenge for the designer is to provide enough viable archetypes that during a play session of 10 or more rounds you are unlikely to meet the same deck. Currently in Clash Royale I would argue this is not the case, there are around 5-10 common archetypes and at the top end of the game, you see these decks almost every round. This is not an issue only for Clash Royale, but all online PvP games feature this to some extent.

Providing interesting tools to the community to enable quick counters and evolving strategies. This common archetype problem is exacerbated in Clash Royale because of the very short game sessions, meaning you encounter more decks in the same period of time. The Archetype problem isn’t a direct gameplay issue because it provides interesting content for players to talk about, this is one of the reasons why Clash Royale has been so popular on YouTube and Twitch. However, without enough variety in rulesets you can grind yourself out of enjoying the experience.

2. The YoYo Effect

Any person who plays competitive 1 on 1 games will know the YoYo effect. When placed in a competitive ladder where your opponents are matched by a ranking formula, such as ELO, the YoYo effect occurs. You yoyo between a high and low point on the ladder as you win one, lose one in a repeating sequence. In my case I can’t break the 4000 cap and I yoyo between 3600-4000.  This becomes immensely frustrating as no matter what I do I can’t maintain my progress. It also means that if I am at 4000 trophies I don’t want to play more as I fear a loss more than I want a win.

Levelling up a single card has no real difference to your position post 3000 trophies and so rank becomes a measure of skill. If this we’re the only mode in Clash Royale it is likely that I would have churned out a long time ago, but what kept me in was the events. Events applied new rules, new deck combinations and restrictions on the gameplay. This often levelled the playing field. The challenge was then to try to beat 12 individuals in a row with only 3 lives. This removes yoyoing entirely as now it’s simply a challenge to continue your winning streak. A new event happened every week and each week it would engage me to compete.  I personally drifted from the ladder climbing and focussed on perfecting the events with 2 or 3 that I got all 12 wins. Clan Wars could provide a more detailed structure for events, and pit not just player vs player, but clan vs clan in a clan winning streak.

3. Clan Engagement

I would estimate a very large proportion of players in clash royale are in a clan. The clan provides a group of people who share your passion and you can talk tactics and share replays. The most used feature within the clan would be the Clan requests where you can ask for cards from your clanmates. Within this interface currently there is a large stagnation of content. In my clan I would say there is a core of 8-10 people that actively chat or engage in conversations and discussion, with the majority of people including myself who simply make card request.

The more active and engaged a community the more likely they are to stick, spend and promote your game to their friends. From a long tail perspective providing engagement, challenge and motivation to these players is what sustains longer term revenues. However, without an evolving or changing metagame, a variety of weekly challenges or a global event the clan chat can become dull and boring. I feel that the clan structure within Clash Royale provides all the tools adequate to really helping people to share and strategise but I felt the strategy of the game had tailed off due to the archetype problem in recent months.

Clan Wars – The Update

Let’s think about each of the problems above and how the update attempted to solve them. Whether it was a success or how it could be improved.

1. Archetype Solution

Clan Wars made a valiant attempt to remove the archetype problem from the discussion. As the war is split into a collection day vs a battle day.  The collection day is an ability for players to find cards and then level them up for the clan. This is a beautiful solution as it tries to solve the Archetype problem in 3 ways

Firstly, it adds chance of which cards you find. You no longer have a reliable card collection, each Clan War forces you to look at the cards your clan has found and create a well rounded deck.

Secondly the cards you do find level up depending on the number of wins. This creates a wider range of stats with which to vary your deck playing abilities. The decision to take Goblin Barrel level 3 or Knight level 11 can affect your decks build.  

Finally your opponents are doing the same collection phase. This creates a more asymmetrical hidden information that means your deck cannot be built to easily counter.

Each of these 3 factors mean that wilder and less common card combinations are used and sometimes picking the most OP card in your collection (level 12 Barbarians) can truly beat opponents through sheer power. This plays out quite well as each clan war my clan has played a much wider range of decks.

Going Further

I think the solution doesn’t go far enough.  It should be tweaked to put even more pressure on the harvesting day and provide more depth in the War day for clans to strategise. Currently the number of cards dropped and number of level ups needed is the same as the main game. There is no reason to use these numbers apart from familiarity and in this particular game mode you never get enough cards to even see level 11. I would actually adjust the balance of the common and rare drops to provide even more “going for it” level ups, making the choice not just on strategy but stats as well.

Currently a player gets 2 attacks in order to earn cards and then their collecting abilities are capped. It would be more interesting if players had a number of lives in collection day, rather than a number of battles. This means that any player could earn exponentially more cards if their skills was high.  In order to limit players who are very skilled at the game the more wins you get the more levelled up an opponent’s King towers could becoming. In a way handicapping the victor.  This would then provide a range between 3-10 matches for each player depending on their skill.

The Clan War battle is also a single battle against a random opponent. In some cases you may have 2 battles, but this is only when the clan sizes are mismatched and you are on the lower tiered clan. You and your clan have no way to discuss or tweak your plays, based on learning from previous battles. It’s an all or nothing affair where winner takes all. What if the clan was given secret information or could earn the rights to see the opponents War Deck by winning challenges. By understanding the cards available for opponents you could again strategise more particularly on your clans deck crafting.

2. Yoyo Solution

Clan Wars does provide an alternative challenge to the ladder. It’s one more thing to do and I believe it would improve the total number of battles each user takes part in each week. However the ladder problem still exists in the single player game. For Clan Wars to be a truly successful update is should actually try to create a larger inter-clan ladder that all clans compete on worldwide ensuring that your personal ladder is of less important than your clans ladder position and you spend more time perfecting your skills for the Clan War.

At the moment this doesn’t feel the case as we’re still establishing clan win rates. However, I dont see a clear display of the Clan Wars leaderboard and no prestige is provided for those who are ontop. This means that overtime the wars fought will become less meaningful and apathy towards the wars might continue. Clan War leagues might be a better solution here, read the C.A.T.S review to understand the promotion and demotion for players and how it drives engagement.

3. Clan Engagement Solution

There is certainly more clan activity since the update. More communication, more replay sharing and an increased discussion about the game itself.  We’ve seen more clan leaders send messages to the group to get people involved and the social dynamics it creates are important. I also was personally congratulated for winning my particular fights in the war and that felt good. People within the group are responding and pushing each other to perform at their best and this is very healthy for the game in general.

Another suggestion here is that rather than creating multiple decks per clan for each war, the clans should craft a single war deck that all clan members would fight with. Clan members could each submit their own deck creations along with a name, and the top picks could be voted on for the war.  You could then imagine clan discussions recounting previous wars and how certain clan mates created crazy but powerful combinations that helped the clan succeed. You would also face the same deck with your opponents and therefore clan members could coach you through each war day.

War Day

For me the actual weakest aspect of Clan Wars is war day itself.  The collection mechanics and the strategic clan chats have geared you up for an epic battle against 5 other clans. However, what actually occurs is a single match against a random opponent.  Once you’ve played this round, the war is over.

The fact that the decks you spent so long collecting and crafting are only used in a single match. War Day feel more like a game of chance rather than skill. I felt this acutely on my first war when I won quickly because I was matched with a poor opponent.  It had built up, to a single battle that was a poor experience even though I had won and then I was left waiting for the next war.

The war could be expanded with a clan tournament.  Taking 8 clans and facing them in a 3 tier round robin tournament where if your clan wins you would obtain even more cards, dynamically shifting your war deck.  A clan would then spend 1 day collecting and 3 days battling to become the ultimate clan of the 8. Also rather than a single battle, each player should have 3 fights with opposing teams giving a bigger range for skill. The further a clan progressed the more involved in chat and strategy they would become. The Archetype issue would actually become a strategic point of discussion as you played more battles you could feedback to clan mates to let them know the opponent is like to drop a pekka. by giving the clan more and more to think about as results pour in.

Conclusion

Clan Wars is a well designed update aimed at solving some of the more complex design issues that occur in older free to play titles. It’s a great solution at solving the Archetype issue that is very apparent in the core game. We’ve also see that it provides some remedy to both the YoYo problem and clan engagement but these are likely to be short term fixes. My largest criticism is that the collecting mechanic and “Going for it” feeling are both underutilised and less important that simply turning up and having your battle with any deck. The clan as a whole feels united during collection day and disjointed in war day. I believe the team should go further to create memorable clan wars that are talked about for weeks or months after the event. Players who create the war deck that win wars will feel proud and acknowledged by the clan and could go down in clan history.  For me this is the weakness with the current Clan Wars update. It’s provided a new game mode that creates a dynamic and interesting deck building environment, but it doesn’t create memorable clan battle or stories. Each clan war fades quickly into the background and the next war takes its place. The more planning, strategy and interaction the clan takes in each war, the stronger the emotional connection will be towards the event and this is what will make players stick for the longer term.

Due to the Star Wars Battlefront II controversy, the industry is taking a far closer look at what monetization practices are ethical, and whether the industry can police itself or needs further regulation to avoid misuse.

In the meantime, it’s likely loot boxes will still be featured heavily in the top charts as the revenue potential of gacha and loot boxes is hard to ignore. Using a random drop system has allowed many new genres and core loops to flourish.

However, designing for gachas isn’t a simple design process. Not all genres and not all types of gameplay can be ported to support a loot box design. We’ve already talked about some of the necessary ingredients:

• Part 1: Ensuring your gacha system has enough depth to sustain drops over time
• Part 2: Ensuring your gacha system has enough width to ensure that each drop is useful to a player

Now, it’s time for the third element: how to handle duplicates. It’s what we call an edge case, but it’s a process that will define how your game will feel over the long haul: Do players feel like duplicates are useful or useless?

Duplicates vs Bad Drops in a Gacha System

The first thing to master when it comes to a gacha system is how to think differently about two situations that can arise; duplicates and bad drops.

For example, let’s assume that we have a Gacha system similar to Overwatch – our boxes only drops cosmetic items. As a result, each item that we drop is permanent (the player keeps it forever and it can’t be “consumed”) and players are chase after the cosmetic items they want for the characters they play as.

In this system, a ‘bad drop’ could be a cosmetic item for a character that I don’t play as – maybe in the future I will, but for the time I’ve been playing I haven’t taken to the character in question. As such, this is most definitely a bad drop.

Ideally, I should be able to convert this item into something of value so that I can eventually get the items that I want. In games such as Overwatch and Hearthstone, this means converting any bad drop into a dust-currency, which allows you to purchase the items you want directly.

However, also inherent in this system are duplicates. In this situation, I receive the same cosmetic item for a character that I already have, which feels like a big loss. It’s doubly frustrating if the game drops a high rarity duplicate (i.e. a legendary skin) as this feels like a massive waste – I was lucky enough to receive a legendary item, but unlucky that it was an item that had dropped before.

As previously suggested, games like Overwatch and Hearthstone handle this by allowing players to convert these items to dust, essentially treating a duplicate the same way as a bad drop. However, the amount of dust dropped is a fraction of the cost of purchasing the skin you want directly, so players still feel terrible when they pick up a duplicate.

As a result, Overwatch eventually went public about adjusting the drop logic to avoid duplicates as much as possible, while Brawl Stars even removed duplicates outright. However, in my view it doesn’t need to be this way. Removing duplicates from your system reduces depth, and puts more pressure on your team to develop more content. Ideally duplicates would be celebrated by players, making this rare occurrence into something of value, rather than serving as a regretful outcome.

In light of this, let’s look at how to build out better gacha duplicate mechanics:

Six Mechanics for Handling Duplicates

#1 Duplicates Aren’t Duplicates

A different way to avoid the pain of duplicates is to make sure duplicates rarely happen.

One way to do it is to make each piece of content generate in many subtly different ways. For example, a weapon or character can drop, but certain sub-elements are randomized and generated.

Using this method, if a duplicate item drops, there are smaller comparisons that players can make between the drops. This is done when gear or drops are both procedurally and randomly generated and there are enough smaller detailed stats that players actively want to optimize.

For example, in the first Destiny you could get the same piece of gear dropping many times.

However, each drop had randomized stats and perks associated with it, causing players to head into a chase in the end game to find unusual builds of gear. The game included perks that offset the problem of some guns being overpowered in competitive modes like The Crucible. While this obviously went overboard causing severe balancing issues, this shows the power of procedurally generated gear – it deepens the chase and makes duplicates something players actively go after.

However, this system can result in players ending up with mountains of weapons and gear that they don’t want to use. As a result, designers need to find ways of converting all bad drops into something of use to players, such as:

• Gold to purchase more weapons
• Dust to re-roll the weapon perks of your choice
• Resources to upgrade the weapons that the player actually wants

While such solutions put the duplicate issue to bed, it also puts more pressure on the bad drop system.

#2 Repair

One system that hasn’t been used often is the repair system..

Fallout 3 used this effectively by asking players to collect duplicates to maintain their gear. Have an amazing piece of gear? It will eventually deteriorate and be less effective over time. To repair it, you can pay a large amount of currency or find duplicates of your gear to repair for free. If the deterioration is felt as fair to players, this can create a repeatable grind to find duplicates of your gear to maintain its highest possible gameplay effectiveness.

This system is likely avoided because of the consumable feel that drops from the gacha become: The feeling that an amazing item will drop, but one that’s only useful briefly. It’s a feeling that anyone who played Zelda: Breath of the Wild will definitely find familiar

#3 Fusion (Unlocking Potential)

[Source: Both Guns Blazing]

Fusion is the typical way that Japanese and Chinese games have made duplicates relevant. These games typically focus on selling stat improvements over cosmetics, and because of this they focus their duplicate mechanics more on unlocking higher stat growth.

Fusion mechanics are designed in a way that requires the player to receive a duplicate in order to increase the stat potential of a card. As such, while you can upgrade a card up to high level, unlocking the ability to upgrade it even further requires you to “evolve” or “awaken” the card with a duplicate of itself.

When looking at the stacking probability needed to get the highest star rating, it’s easy to see why they do this. You can drive a lot of depth in a gacha system by asking players to chase after duplicates without adding more content.

The problem with this comes in the randomness of the system. Getting a single duplicate becomes so important in this system that players can become very frustrated. Players have no grindable path to unlock the potential for their favourite characters. Hence, designers came up with a new system: Sharding.

#4 Sharding

As duplicate systems changed over time, there was a need to make them more flexible and granular.

To solve the issues of fusion, gacha games started to experiment with shards instead of duplicate fusion, best seen in Western Gacha games like Galaxy of Heroes. With shards, each character can’t be unlocked until you have collected a certain amount of shards. In the above example, Grand Moff Tarkin requires 80 shards to be unlocked.

However, that’s just to unlock the character. To upgrade the character to its maximum potential, the player would need to collect additional shards, so “duplicates” are simply just additional shards needed to progress to the maximum potential.

With characters now needing hundreds of shards instead of single drops to reach the maximum characters, games added mechanics which allowed players to grind for specific shards, so players that are looking to upgrade or unlock their favourite character could grind specifically for it. This wasn’t possible with the fusion system before, since giving a single card could mean massive progress for players. In short, sharding allows clear progress.

However, there remains one big problem: opening up a gacha pack you’ve paid for and receiving mere pieces of a character – nothing that you can use there and then. It’s a transaction the player almost always regrets and, as a result, Supercell came up with a workaround.

#5 Unlock & Upgrade

Clash Royale provided a completely new framework for how to handle duplicates. It took the best of the Shard framework, made the handling of duplicates restrictive, yet still has a gacha system that feels fair.

With Royale’s system, each card is unlocked after getting the first card. This feels far better than shards because getting a new card feels amazing – there’s no more paying for “parts of a character”.

After you unlock the card, the card becomes a duplicate sink. In order to upgrade the card, you need to collect a number of duplicates of that card. It removes any needed management of duplicates, while giving a clear path for players to upgrade their cards.

Due to the design, players will unlock cards fairly quickly (you only need one card), but the majority of the chase is after the (thousands) of duplicates necessary to upgrade your cards to a competitive level. This system has significant depth, allowing Supercell to be generous with the cards it gives out, and keep players collecting for years.

However, despite its perks, this design still has disadvantages. For one, Clash Royale has to work really hard to try to ensure that as many cards as useful to players as possible. Otherwise, getting a duplicate for a card you aren’t using is completely useless (the only way to get value from it is to trade it away to clan mates). This works very well for CCG style games, but many games can’t support this level of gacha width – where every item from the gacha is theoretically useful.

#6 Unlocking Better Cosmetics

All these mechanics thus far are primarily focused on handling situations where duplicates give out better stats – they “unlock the potential” of an item so they can be upgraded further. This works great for games that are RPG-based and are comfortable with players speeding up progression (ex. Clash Royale), but most competitive PvP games can’t do this, such as Playerunknown’s Battlegrounds, and League of Legends. Competitive PvP games can really only sell cosmetics. So, how do you add value to duplicates for cosmetic economies?

This is considerably harder, which is why most cosmetic driven games end up allowing players to convert duplicates into dust (ex. Overwatch) or allow players to sell them on a secondary market (ex. Counter Strike: Global Offensive). League of Legends has even dabbled in at first not fully “unlocking” the cosmetic, but only allowing the player to “rent” the cosmetic. Getting duplicates eventually allows the player to convert their duplicates into a permanent item.

However, beyond this, the only thing you can do is make duplicates of cosmetic gear unlock cooler/better visuals of a cosmetic item. In Counter-Strike: Go (CS:GO), they use a “decay” system to do this.

In CS:GO, each item is dropped with a randomly assigned “decayed” attribute. This could mean that the item looks brand new, or is heavily worn down. Getting a duplicate allows players to find items which have far less wear, meaning that players aren’t just chasing that “item”, but also chasing the best looking version of it. The visual differences between “Factory New” and “Battle Scarred” are striking – making the value of having the highest valuable version of the item very important to players that are chasing after cosmetics.

As such, similar to stats, cosmetics can have a “unlocking potential” of their own – you just need to make sure your cosmetic items can have varying degrees of visual quality.

Summary: Duplicates aren’t Bad Drops

In any Gacha system, regardless if you’re just dropping cosmetic items or gameplay impacting items you, as a designer you are responsible for ensuring that there is as little remorse or regret from players – for making sure that each purchase of a loot box feels rewarding to players.

• Gacha depth helps ensure that you can sustain drops from a gacha.
• Gacha width ensures that each item is as useful as possible.
• However, Duplicates are inevitable, and how you handle them is important to achieve the balance between a system that feels fair to players and doesn’t cripple your studio by producing lots of content.

There are seven examples of mechanics you can use to handle duplicates and give them value:

• Dust: Allowing players a path to purchase items they want
• Duplicates aren’t Duplicates: Using procedural generation to have subtle differences between drops
• Repair: Duplicates can power up a previously owned item
• Fusion: Unlocking further potential
• Shards: Breaking fusion up into a more granular path
• Unlock & Upgrade: Unlocking higher stat levels with duplicates, no option for duplicates
• Unlocking Better Cosmetics: unlocking better looking versions of the same cosmetic

Each have their pros and cons, but hopefully can help you decide what is the best path for your game.

When Supercell launches a new game, it sends shock waves around our industry and players alike. On June 14th, Supercell released Brawl Stars — and in typical fashion, we all jumped on to give it a try.

But there was something special about when Supercell launched Brawl Stars. The game was Supercell’s first outside of the strategy genre. Brawl Stars is the first action-based multi-player game for Supercell, and notably, the most casual MOBA style game launched for mobile to date. Supercell also publicised the launch, beginning with an e-sport style tournament. This isn’t a typical soft launch; they are already building up a massive community and driving a strong streamer culture around this game. This was a bold move for Supercell. Supercell has been known to stop games such as Battle Buddies, Smash Land and Spooky Pop when they don’t look like they will become a top 10 game. Going into this soft launch with so much confidence is bold.

But weeks after the game has been launched, industry veterans began to weigh in and started noticing the cracks in the design. Many have already dismissed the game as an unlikely game to launch, despite having a massive following already from streamers and e-sport fans. Currently, the game is sustaining in the top 10 grossing in Canada and driving a massive community around it. Despite the concerns, this game could end up being a surprise hit due to the strong multiplayer gameplay.

But ultimately as a game designer, what I see from Brawl Stars is an amazing game that is weakened by a poorly designed gacha system. It fails to deliver on what a gacha system needs to do, and it will ultimately not last in its current incarnation. Comparing the system to Clash Royale, Brawl Stars system is considerably weaker and will result in lower revenue on a per player basis. Even if Supercell can drive downloads organically, this will hold it back from where it could be.

While I believe the game is incredibly fun to play and may just succeed based on its multiplayer component alone, ultimately the game will be weak on a revenue-per-player basis.

From this analysis, it begs the question:

What is it about the Brawl Star mechanics which weakens the Gacha? That comes down to Depth.

“Depth” of a Gacha System

Something to clarify is about how designers look at depth of a gacha system, and why this matters.

The depth of a gacha system ultimately defines how long it will last, roughly what the maximum spend a player could spend to reach the end of content, or how long a player would need to play before reaching the end of content. This is usually defined as the number of drops it takes to complete the gacha.

A “drop” in a gacha is defined as giving away a single item. For example, in Clash Royale a drop would be synonymous with a single card dropped from a chest. Some designers also call this a “pull” — but for this article, I will call them drops.

Keep in mind that a drop does not mean a chest. A Chest has multiple drops in the case of Clash Royale, but a chest in Brawl Stars only contains a single drop. Also, not all drops are alike — a drop from a legendary chest in Clash Royale is not the same as a drop in a wooden chest — since the legendary chest has different probabilities for selecting higher value items. But when roughly measuring the depth of a gacha — you can ignore (average out) the “quality” of a drop.

Drops are important because the ultimate goal in free to play games is to maximise long-term retention and maximise the cap of the economy. To drive strong long-term retention, players need to have a long lasting sustained desire to pull from the Gacha. The more drops this takes, the longer the system will last.

The more drops a gacha can sustain, the more generous a game can be, the higher revenue per player, and the higher the long term retention would likely be.

On this metric, Clash Royale’s system dominates Brawl Stars, comparing their soft launch states. Designers usually have 3 key variables to maximise Depth: Content, Duplicate mechanics and Pacing. In all 3 of these cases, Clash Royale’s systems outperform Brawl Stars.

Problem #1: Content

Content is usually the easiest problem to point to with a shallow gacha system. Brawl Stars has 15 characters (for now) whereas Clash Royale had 42 at their soft launch.

What this gave Clash Royale was a longer period of time in which players were likely to get new content, as well as the ability to control the pacing of the introduction of this content. With 42 cards at launch, Clash Royale was able to pace the pool over time using Arena tiers. So players knew they needed to play for awhile before they could even gain access to some of the upper tier cards.

On top of this, because they were able to launch with this much content, each interaction with the gacha system felt novel and interesting, especially between arena tiers. So playing through arena 1, each time you opened up a gacha chest you typically got new cards. Each time you levelled up to a new tier, you were introduced to a whole new set of cards, all of a sudden the gacha got way more exciting to open (even inciting purchases like the limited offer for each tier!).

For Brawl Stars, with 15 characters, all available in the gacha from the beginning, with only a few as legendary, this leaves Supercell in an inflexible position. They need to keep all 15 in the pool from the beginning, otherwise, players will get duplicates too fast from the gacha. By only having a few legendaries, the path to complete the gacha feels fast. As a paying player of Brawl Stars, I’ve dropped a small amount of money, but already feel like I’ve unlocked a majority of the content that the game has to offer.

With more content, Brawl Stars would have considerably better control over the player experience and make it last far longer.

For Supercell to correct this problem it may not come in the form of new characters. Brawl Stars gameplay is not the same as Clash Royale. Clash Royale’s core gameplay supports and pushes players to have a collection of cards, especially since each battle requires 8 cards chosen. Brawl Stars only asks the player to choose 1 character. If they add too many characters, this may lead to players losing the desire to collect them all. Having too many characters can lead to players just choosing one they like and ignoring the rest. Brawl Stars will need to find new ways of dropping desirable content, and it may not be in the form of characters. Content can come in the form of special abilities, perks, equipable weapons, customizations, which each could add considerable depth to the progression system, and drive players to upgrade more than just their favourite character.

Problem #2: Duplicates

Content typically isn’t a terminal problem on its own. Content is simply the base in which the gacha total drops has to work with. If content were the only thing that was important, Hearthstone’s 1,000+ card collection would dominate over Clash Royale, but this isn’t the case. The fact is that Clash Royale got away with significantly less content than Hearthstone at its launch because of its duplicate system.

Even with a smaller set of content, a strong mechanic for handling duplicates can make a gacha mechanic last.

The most terminal problem that was introduced with Brawl Stars was the mechanic for handling duplicates.

In Brawl Stars, getting a duplicate character in the gacha meant that you were instead rewarded with a single blue chip. This mechanic is similar to Hearthstone, where you can exchange duplicate cards for a small amount of dust. Players can exchange the blue chips in for unlocking characters, although the number of blue chips necessary to unlock many of the rare characters is insane.

As a result, each time I have purchased gacha packs from Brawl Stars I’ve felt completely regretful. After I unlocked a majority of the characters, each chest has a high probability of dropping a single blue chip over unlocking a new character or gaining some elixir (the currency necessary to upgrade your characters). Having a string of gacha packs that just give out blue chips, especially if you’ve unlocked all the content, would surely cause many players to churn.

Clash Royale doesn’t have this problem because it drives significant value from its duplicates. Duplicate cards are necessary to upgrade the card. Getting a single card unlocks the card for use, but to have the fully upgraded version of the card, you need duplicates of it.

This is what makes Clash Royale’s Gacha system last. Thinking in terms of the number of drops, even with a base amount of content of 42 cards, requiring each card duplicate to be found hundreds of times (depending on rarity) exponentially increases the number of drops necessary to reach the end of the economy.

Even thinking about maximising a single legendary card can show you that it takes a lot of drops. It’s reported that Supercell drops 1 legendary card 0.43% of the time in their gold level chests. If we use this as a base, and a pool of 6 legendary cards, that leaves the % of dropping your chosen legendary to be 0.0716%. In order to upgrade this card fully, you need 37 drops of this card. So, on average, a player will need over 50,000 drops before their single legendary card is fully upgraded. That’s a system that LASTS.

So for Brawl Stars to utilise its minimal content better, it needs to think about duplicate mechanics similar to Clash Royale. Potentially duplicates increase the max upgrade level of a character. Potentially duplicates unlock new special abilities. Without it, players will simply lose interest in the gacha, or feel as though the high price tag to purchase chests are just not worth it.

Problem #3: Pacing

With gacha systems, designers have one final variable to control how long their gacha lasts: pacing.

Not all gacha systems support a huge amount of drops, but to counteract this, increase the time it takes for a player to get another drop from the gacha. For pacing, game designers typically have a couple methods to use:

• Pace how often the players can open the gacha
• Pace how many drops the gacha gives

Clash Royale gives a lot of drops daily. With free chests, crown chests, clan chests, and regular chests, each day players can get plenty of free drops to feel progress. This is mostly because Clash Royale’s duplicate system multiplied by their high amount of content supports such a high amount of drops.

With Brawl Stars, because of the low level of content and the fact that duplicates aren’t necessary, this left Supercell designers in a bind. They had to pace their gacha significantly slower. They did this by tying chests to coins, and by making chests only give 1 drop each. Comparing this to the experience of opening a chest in Clash Royale, Brawl Star’s gacha boxes are far less rewarding. The reward pops up, you get a single currency of something, and then you’re left feeling — “That’s it?”. This problem is magnified when each drop can be amazing or terrible feeling. If I get a new character or some elixir — this feels good. If I get a blue chip… I feel like all the time I put into collecting coins for that box was worthless.

Clash Royale’s chests on the other end can guarantee rare or legendary cards, and even if I get a duplicate, it still feels beneficial. So even as I reach the mid-game and end-game where I have a majority of the content, every time I open a gacha I feel like I’m making progress, and I have a chance for big gains.

Supercell had pace Brawl Stars chests this harshly because their economy only supports a certain amount of drops. If they increase the number of drops a chest will give, this will mean they either need to increase the pacing (increase the cost in coins to purchase a chest) or they will be allowing players to speed through content significantly faster — something they can’t afford with the low amount of content they have so far.

The Path Forward

Supercell’s Brawl Stars is an amazingly fun game to play. As the community has shown, there is a huge desire to play an action-based MOBA on mobile, and clearly, Supercell has capitalised on this with Brawl Stars. This game has a strong chance of succeeding simply based on its rabid community building around its multiplayer core gameplay.

But as we know in free to play, a strong core gameplay is only the first step towards success. For Brawl Stars to become a Supercell-sized success, it’s about how long their systems last.

Improvements could come with more content, it could come from better pacing of the gacha, but driving more sustainable drops likely will need to come from a better mechanic for duplicates to avoid a content treadmill. Taking a page from Clash Royale’s system and finding a way to make duplicates a key part of reaching the end of content for its gacha mechanics. Doing so will exponentially increase the lifetime of their gacha systems, plus drive stronger retention and monetization from their user base.

Brawl Stars has the DNA of the next Supercell hit. They may just need to make some last minute adjustments to make it the next billion dollar game. I’ll be cheering for them.

Supercell has dropped a bomb on the mobile gaming market. Their new game, Clash Royale, soft launched just as 2016 got started. They have soft launched in only 8 countries, but this game is already a sure success. Supercell has already committed the game to a global launch in March.

Supercell has made a lot of smart choices with this game. They have a fun, competitive, forward-thinking game that exemplifies what modern free to play design should feel like. Previously I’ve talked about just how difficult Multiplayer on Mobile is to get right, yet here Supercell threw out the rulebook. They’ve now proven that Synchronous multiplayer can work on mobile. Many have even gone as far to say this is the first successful MOBA on Mobile.

Whatever you want to call this game, it will be a success, and it did so while breaking many of the rules.

But enough praise for the game, today I’d like to talk about my favourite subject when it comes to mobile game design: sessions. Specifically, where I think Clash Royale succeeded in creating session design that pulls players back each day.

They did so with 2 clever systems:

• Free Chest Systems
• Chest Slot System

Overview of the Game

Clash Royale is a card-based real-time strategy game. The best way to explain it is to watch:

Player use cards to spawn various units to attack opposing player’s towers. The goal is to destroy their towers before they destroy yours. The strategy is in choosing when and where to place your cards: to counter your opponent’s units, and to ultimately press the opponent enough to destroy their central tower.

Overall it is a hectic strategic game that lasts only a few minutes. It feels like a real-time hearthstone match mixed up with the clash of clans gameplay.

The Core:

• Winning a battle will reward you with chests (in various ways)
• These chests give you random rewards: gems, coins, and random cards
• Cards can be upgraded with enough duplicates of the same card, and enough coins
• To win, you need a variety of Levelled up cards

The Goal:

• Players want a collection of competitive cards
• To win as many matches as possible
• To get as many crowns and trophies as possible
• To reach highest Arenas
• To reach the top of the leaderboard (With your clan or by yourself)

The loop is focused on collecting and gathering cards. Not unlike Hearthstone. The big modification though is the ability to upgrade these cards.

To upgrade a card, you need to collect duplicates as well as coins. The real key comes in the rarity of the cards. Some cards are inherently better than others (ex. the Giant), and since they are RARE or EPIC, they drop a lot less than others. So not only do you want to collect these rare cards, you also need to collect a lot of them to fully upgrade the card.

This strong desire to collect and upgrade your cards is what drives all systems in the game. Each session is about attempting to get as many chests (and thus cards) as possible. To collect cards the fastest, the player has to play by the rules that Supercell desires to drive retention and monetization.

#1: The Free Chests System

To analyze Clash Royale’s sessions, let’s start with the most obvious system: how Clash Royale starts and ends its sessions.

For any game, good session design is marked by two things:

• You’re rewarded each time you come back to the game
• The game quickly gives you a short-term goal, that can be accomplished within that session, or at least within a few sessions

This is usually accomplished in most games by a few things:

A Rewarding Start:

Good sessions always start off with a instantly rewarding mechanic. Most games aim to have a collection of resources each time you return or a Daily Reward System. This gives the player a good feeling instantly after starting up the game.

Short Term Goal:

But having an instantly gratifying mechanic isn’t enough. The player must quickly form a goal which will drive the player further into the game. They need a goal which asks them to engage in the core gameplay.

This dynamic is usually created by a Daily Mission system or wanting to use up all Energy.

Clash Royale creates these 2 dynamics with 2 systems: A free chest every 4 hours, and a crown chest after collecting 10 crowns.

The free chest system marks the beginning of your session: you come in, open up your free chests. It feels rewarding just to come back.

Secondly, the crown chest. To open you must collect 10 crowns from opponents. This gives me a nice short term goal. Even if I am far away from ranking up, I want to collect 10 crowns so I get the crown chest. Realistically this goal can be accomplished in 1 session, or at least within a day.

This is perfect for driving a strong session length. A clear goal as soon as they’ve opened up the app. Something that the player feels good for accomplishing.

This chest can be opened once per 24 hours, which gives a strong daily goal for players. Players wanting to get the maximum number of chests come back each day and play enough matches to collect 10 crowns.

These 2 chests, which take up a small portion of the UI, incentivize strong sessions per day and strong session length.

#2: The Chest Slots System

Secondly lets look at the Chest Slot system.

Each time you play a round, if you win (score more crowns than the opponent), you will receive a chest. This chest is randomly chosen from Silver, Gold or Magical. Each chest takes time to open: 3h, 8h or 12h. You can only open 1 chest at a time, and to restrict things further, you only have 4 slots to store chests.

No other game on mobile has used this pattern for pacing players. This is the first I have ever seen someone attempt something like this. Instead of pacing the players through energy or construction timers, they went with a system that limits the rewards players get. Players can play as often as they like, but in order to progress and upgrade their deck, they need to pace themselves.

This system can only work if they know that :
#1: players won’t grow tired of playing their game… no matter how much they play
#2: their matchmaking and card upgrade system can prevent players from progressing into the higher leagues too fast

#1 is no easy feat, but I believe they accomplished it. Clash Royale is a game, like Hearthstone, that has a shifting meta, no clear answers. Every battle feels different, especially because its synchronous multiplayer.

#2 is based on the big change they made over a pure Trading Card Game system. Because you can upgrade each card, eventually the player will be confronted with decks that are stacked against them. No amount of skill will be able to defeat a deck with higher level units. Because of this, players will eventually need to play the chest opening game. There’s no avoiding it.

Matchmaking aside, what about the overall feeling of the sessions?

This system fulfills the goals of Flexible Sessions. Rather than blocking the player from playing the game, they ask the players to be smart about how they spend their time.

But what about having to come back every 3 hours to clear out a single chest? Why not allow for chests to be opened up automatically? Opened up in queue?

My guess is that Supercell knows the pain that the chest slots creates, and this is intentional for retention and monetization purposes. Players have to organize themselves to hit all their timers. This uncertainty of hitting their Chest Timers drives players to come back, and pay to speed up the timers when they know they won’t be able to return optimally. I know for myself this chest slot system has converted me into paying to skip timers.

But regardless if you’re chest slots are full, the player can continue to play, which really is what drives the flexible sessions. Even if you’ve filled up your chest slots there is a lot of productive things you can do in the game:

• You can continue to play and push as far up the leaderboard as you can go with your current cards
• You can continue to collect crowns for the Crown Chest
• You can donate cards and request cards from clan mates
• You can chat and read messages from other clan mates
• You can watch other battles from around Clash Royale (and be teased of late game content or tempted to speed up progression…)

So although the Chest system is restrictive, its not nearly as restrictive as a straight up energy system. And having this “soft” restriction allows highly engaged players to opt-in to leaving the game when they feel smart about it.

Conclusions

Supercell have a big success on their hands with Clash Royale.

They crafted strong sessions with 2 systems:

• A Free Chest system that gives rewards just for arriving and setting a strong session goal
• A Chest Slot system that effectively paces players without energy

This base of strong session design is driving strong retention and monetization. I don’t expect Supercell to change much as this game moves towards global launch. I expect that they are mostly focusing on making their end game deeper and more competitive. This will drive the game even further up the Top Grossing charts, and drive even stronger long term retention. This game will be on the charts for a long time to come.

Overall Supercell clearly have opened up new doors with their designs. It shows that synchronous multiplayer can work on mobile, and energy is not needed to pace players properly. Lets see whether this ushers in a new “Clash of Royale Clones” or developers can apply these design lessons to new games on mobile.

Supercell’s most recent soft launch is called Smash Land. It’s been in soft launch phase in Canada and Australia since March 31st 2015 (About 2 months from this post). There is no doubt that Supercell’s soft launches are huge news for the mobile free to play industry. Supercell is notoriously picky about what games that make it to soft launch. Each new game goes through rigorous internal feedback, and only the best games survive. The games that hit soft launch are games that Supercell genuinely believes have a shot at the Top Grossing charts.

Smash Land is based on “Monster Strike”, a massive mobile free to play game in Japan. In December 2014 it took over Puzzles and Dragons’ top spot in the Japanese charts. Similar to how Supercell started Clash of Clans with looking at Backyard Monsters, Supercell now looks to simplify the design of Monster Strike so that it could work in the Western markets. But in Supercell’s simplification of a game that performs so well in Japan, has the game stayed intact?

Has what remains kept what is required to be a successful game?

Smash Land’s Core Battle

Supercell decided to keep core battle game the same as Monster Strike. The core battle mechanic is a Physics-based RPG battle. Almost like a game of pool, the game is mostly about predicting how balls on a flat surface will bump and move to create a preferred outcome. In Smash Land, the game is about lining up one of your characters so that it bounces between walls, enemy characters and your own characters as many times as possible. The player then collects up to 10 different heroes, each with their own special ability. For an overview of the mechanic, watch this video:

Overall the core battle feels smoother, cleaner and is much easier to understand than Monster Strike. Each character feels unique because of their special abilities which feels great.

The gameplay is very strong for a mobile F2P game. Its easy to pick up and understand for any player. The feeling of skill is strong — I can predict a few bumps and feel smart about setting up strong combos. On top of this, because of the nature of physics, Luck comes into play. Like Peggle, physics is usually pretty easy to predict after the first shot, but after the first few collisions it becomes almost impossible to predict the outcome. As a result each move can result in some “Post-Action Luck” which is critical for casual games.  Players feel smart and each shot is unpredictable.

Overall they’ve taken the best bits of Monster Strike and applied it to a more focused experience. It’s a great battle system that is easy to get addicted to.

Heroes

Outside of the battle, players can also engage in upgrading their heroes stats. This is really where Smash Land departs from Monster Strike.

Monster Strike contains far more variety of stats for each character:

Just comparing these two screens you can see the dramatic comparison between the games. Its much easier to understand Smash Land compared to Monster Strike.

However, at what cost is this simplicity? In Smash Land the major differences between the characters are special abilities and their health to damage ratio. In Monster Strike, the team you bring into battle requires far more strategy as you progress in the game.

You need a balance of elemental types on top of ensuring you’ve got strong special abilities that are complementary. My guess is that while Supercell’s game clearly scores points for understandability, it will seriously limit the long term replayability of the game compared to Monster Strike. Players just won’t have nearly as much to strategize about in the long run.

Smash Land also departs from Monster Strike in how upgrades are handled.

Heroes are upgraded with gold and time. So the player collects gold from playing matches or collecting them from treasure hunts, and turns this gold into upgrades to their heroes. The cost of each upgrade escalates very quickly. As a result, the game really starts to require many, many battles before you can afford a single upgrade.

Smash Land’s system is far simpler than Monster Strike. Monster Strike takes cues from Japanese Gatcha games like Puzzles and Dragons. To upgrade your heroes you must collect hundreds of characters and consume them to give experience points to your heroes. For a great overview of Gacha, read here.

Leaving the Gatcha system out for Smash Land is a big risk, what remains is a far too simple economy that quickly becomes a grind.

Monster Strike’s system with consuming & collecting monsters has a massive advantage in the long run compared to Smash Land. Instead of just 10 heroes, Monster Strike has almost 1000 collectable monsters in the game (source). With this massive set of monsters, they have created a system where players have much more excitement for the long run.

As I’ve spoken about before, to alleviate the feeling of grinding it’s all about creating random spikes of progression. Similar to games like Diablo you need to find ways to add luck to your progression. Ensure that each battle can result in a lucky outcome which could dramatically increase their pace of progression. In Diablo this could be finding a legendary weapon on the ground which makes it a breeze to beat the enemies following.

In Monster Strike, instead of powerful rare swords, players can randomly get awarded rare monsters from the gatcha system. The player now feels lucky, like the game gave them something for free that should have cost them real money or a lot of time. Because the player got this rare monster, they can rush through previously hard levels and feel great.

Overall the rate of progression may be slow, but because there are these moments where progression randomly spikes, players are far more likely to engage for a long time.

This variable progression is missing in Smash Land. To progress, you must upgrade your  heroes in a linear path. Each victory gives you a calculated amount of rewards. The cost of upgrading a hero grows each time.

Overall, with only 10 heroes and very limiting upgrades, the metagame is just too simple. I have the same heroes as everyone else, the same upgrades as everyone else, so there is no moment where I feel like I’ve got a really unique set of heroes that are more amazing than my opponents. Without this unique feeling, it is hard to get attached to my characters or get attached in the long run.

Desirable Stats

Smash Land removed plenty from Monster Strike when they simplified the heroes/monster collection structure. But regardless of how many collectable characters you have in your game, if you want players to engage in an upgrade system you need to ensure that those upgrades are desirable.

In Smash Land my drive to upgrade is very weak. The battle overall feels very Skill & Luck driven (as I described above). The outcome of battles has more to do with getting repeated bounces over how much each player’s heroes had levelled up. In many cases I won with far fewer levels than my opponent, or I lost at the hands of an opponent that had far fewer hero levels than me. This translates in less player demand for upgrading their heroes. Instead of having a strong desire to have the strongest possible team, players will blame victories on their skill or luck and will more likely be content with their team as is. This is a difficult balance to get right in any game. For more on Stats vs Luck vs Skill, read on here.

But for this game, where its whole monetization plan is dependant on players upgrading their characters, Stats must take more precedence in the outcome of a battle.

If the player has a decreased desire to upgrade their heroes, then this will break how the game monetizes. Hero upgrades are at the core of how this game makes money. Players grind for coins (or spend money), on top of have long timers (8 hours or more) to upgrade their characters. After spending money in the game, speeding up the upgrades all of my characters substantially, I really didn’t feel any more powerful in the game. I lost subsequent multiplayer battles, and was now facing an even higher upgrade cost for my heroes. In the end spending money in the game really just didn’t feel worthwhile.

Overall Thoughts

If this wasn’t launched by Supercell, this game would never have gone under so much scrutiny. The game on its own is polished, fun to play, and ticks all the boxes for being a successful free to play game:

• Strong Pacing of content
• Multiplayer gameplay to provide long tail retention
• Guilds to bring players together without requiring Facebook
• A simple game mechanic that’s easy to pick up and play, hard to master

But when you put them all together in this game, the metagame is too simple:

• There is not enough variety or strategy in choosing heroes
• Upgrading quickly becomes tedious and a long grind
• There is not enough desire to upgrade your team to compete at the highest level

So how will this do on the market?

So far it seems Supercell is keeping this game in a quiet soft launch. Comparing this soft launch to Boom Beach, by 2 months Boom Beach was higher in both the download charts and grossing charts within Canada (source:AppAnnie). That points to Supercell keeping the marketing costs & number of new users down for the time being while they improve the game. Supercell is very rigorous with their soft launch games. Just last year they released “Spooky Pop” which failed to hit their targets. As a result they decided to cancel the game.

Can Supercell turn this game around during the soft launch? I think it will be difficult. They cut so much away from what made Monster Strike work, its hard to see if small feature additions will be able to rebuild what’s necessary. It will only happen if they completely rebuild their Hero progression systems.

I think Smash Land should be an example for all future mobile game designers. Simplicity can open up to wider markets, but the focus on Free to Play must be on long term retention, not the widest audience. Game designers must strive to create enough longterm depth in their metagames, or else they will fail.