mobile

Thoughts on WWDC 2019 Keynote (games-specific)

by

This was all around a very exciting WWDC Keynote, that I feel advances the entire Apple ecosystem in various ways. In this post though, I’ll primarily focus on the relevant parts with impact to the games vertical.

Xbox One and PlayStation 4 controller support

The facts: Apple announced Xbox One and PlayStation 4 controller support across the entire iOS, iPadOS and tvOS family. This was really but a footnote in the whole 2-hour plus proceedings, but there’s a lot of tea-leaves that can be read here.

For one, this seems clearly in service of the Apple Arcade service, and part of a wider games vertical push, coming later this year. For years I’ve lamented that Apple didn’t have a passion for the games business (unlike their passion for music), despite being uniquely positioned to really shake up the landscape. This may be a sign that this year is different, and Apple is making an earnest effort.

This may seem like an obvious thing, but getting over 100m1 hardcore-gamer-approved controllers “for free” is a big deal. It will go a long way in spinning up the virtuous cycle of more developers taking advantage of controller support and more gamers looking for controller-friendly core titles.

It’s also worth pondering whether Microsoft and Sony actively participated in realizing this, or just passively agreed. I’d bet more on the former, given the mobile gravity narrative I’ve discussed some years back.

Project Catalyst

I’m assuming Project Catalyst is what became of Marzipan. In any case, this was the initiative to make porting iOS apps to Mac easier. Today’s section on Project Catalyst was easily one of the keynote’s biggest highlights (only overshadowed by SwiftUI in my opinion).

I don’t think it’s a coincidence that Apple picked a game as one of its 3 testimonials onstage today. MacOS is at the end of the day a 100m install-base platform, and Project Catalyst had made it easier for game studios to consider adding Mac to the list of supported platforms.

Sign in with Apple

This announcement makes a lot of sense given Apple’s privacy-as-a-platform-differentiator value proposition. (The throwaway email generation is in particular a nice touch.) By making it mandatory to any app that offers a 3rd party login, Apple has also introduced a lot of uncertainty to developers who want to own their users’ account identity.

This is more aimed at ad-tech and Facebook / Google, to be clear, but it could be pretty messy for game publishers who leverage them for sign-on and virality. For cross-platform games it also introduces more potential account / social fragmentation, if players are offered several choices (e.g. Facebook / Xbox Live / Apple) to create an account with. Certainly the account creation flow just got more complicated for UX designers.

Ingredients for post-platform

The 3 bits above can all factor into the post-platform narrative. It just got much easier to envision a console-like experience with an Apple Arcade game, where a player is sitting in front of an iPad with a PlayStation controller.2 In this world there’s also no reason why the iPad can’t become the most popular platform to play Fortnite (given it’s predominantly a console game), and the iPad offers a degree of mobility not existent with PlayStations.

Of course, we still need to see what Apple actually delivers later this year, and it’s a cliché to say that it all depends on the content (it does).

General keynote thoughts, beyond games

Some general thoughts, for those curious:

  • SwiftUI was the star of the keynote. Even as someone who has not professionally coded at all, this was super cool, and makes me tempted to pick up a Swift programming book
  • The Mac Pro and Pro Display’s pricing is all about re-establishing the high-end of Apple’s line-up, especially as the iPad and MacBook’s use cases become blurry (did you see iPads now have mouse support?)
  • The Pro Stand with a $999 price tag was a bizarre marketing move, to say the least. Do you have to purchase either the $199 VESA mount adapter or the Pro Stand? (Looks like you do.)
  • The Minecraft AR demo – for now this remains “cool demo” and not very practical. It also seems like just laying the foundation for an eventual hardware (Apple glasses?) announce
  • I’m waiting for the next Apple Watch hardware refresh to get a replacement. I recently traded in my original series Apple Watch for a paltry $27
  1. Xbox One + PS4 sell-in is around 130M; actuals for total number of controllers might be quite a bit more, given that gamers may have multiple controllers per console for local multiplayer. ↩
  2. I’m being a bit deliberate in choosing the iPad in this example, given there are roughly 500m iPads in use which is far bigger than the install base of AppleTVs, and using console controllers with iPhones may be considered silly by portions of the western audience given the screen size. ↩

Arknights

by

Arknights is a new mobile F2P game from Hypergryph, a studio founded in 2017 in Shanghai. There’s not a lot of info about this young studio, aside from “the founding team are from Google, Massive Black and Cygames.” I’d guess it’s about 50 people in headcount.

The reason I’m writing about this game / studio is I think it’s a good example of the new gen of Chinese developers’ capabilities. This is not a Tencent / Netease studio with massive headcount and brute-force production. The scope and production budget seem carefully managed, while still achieving a sharp impression of high quality and polish.

Game overview

So what is the game? Arknights is an original IP / anime-themed / character-based / tower defense game, with f2p gacha monetization & progression. This video covers the gameplay:

The game art and UI leaves a strong first impression:

Main menu, with the menu UI tilting to gyroscope movement (creating a pleasing effect). The UI is surprisingly clean (especially compared to most Chinese games) given how many features are actually on display
Character info screen
Character skin select (more of a placeholder for now, as most characters only have the base skin)

Chinese studios have developed a specialization in anime-based IP for a while now. To name a few brands that have struck it big in China and/or overseas: Honkai Impact 3, Azur Lane, Girls Frontline, Onmyoji. Given the ecosystem of talent, I’m not surprised that a new studio is able to execute on new IP creation here well.

Core gameplay summary

What I am pleasantly surprised by is the gameplay: this is a fairly thoughtful iteration on the tower defense genre. (Scroll up to the video to see it in action.)

To quickly summarize the gameplay ruleset:

  • Deck-building: players start each level with a deck of 12 characters, plus one optional additional character from a friend (or socially recommended)
  • Grid based, real-time combat, with goal of stopping enemy units from reaching assigned grids on the map. Energy charges up over time, and can be consumed to place units. There is a cap on number of units placed per map
  • Unit placement is based on class restrictions – generally, melee are placed in low ground (where ground enemies will pass through) while ranged are placed in high ground
  • Units can be recalled from the map. Recalled or killed units go through a revival timer, after which they can be placed again, at a higher energy cost
  • Level design elements:
    • map grids where enemies can be pushed off, or give special buffs and consumable abilities (an AOE stun)
    • Large amounts of special level rules, for example melee units only and no energy auto-generation
  • Unit classes that fulfill specialized roles, formed through the following building blocks:
    • Unit attributes: HP, attack damage (physical, magic, heal), attack range, , attack speed, attack effects (single target, AOE, slow), defense, magic resistance, number of units blocked, deploy cost, revive timer
    • Units have special abilities that charge up over time, with diverse effects ranging from basic attack steroids (attack speed buff, increased attack range, next attack hits 2 enemies etc.) to resource abilities (generate extra energy for players) and displacement abilities (pull / push enemies)
The 8 character classes, standard fare for RPGs but provides a good foundation for gameplay depth

This is not a particularly complex system but it has enough space for a lot of variety in characters, and the characters are certainly the heart and soul of the game. Players do need to invest broadly in many characters, as the classes each play specific roles and different levels will emphasize different classes.

The generic single level strategy is as follows: put down some Vanguards (melee, low cost, abilities related to energy generation) for early defense; with more energy, build up your core defense of Defenders (high HP / low attack, can block multiple units), Snipers (ranged physical) / Casters (ranged magic) and Medics (healers); recall your Vanguards to free up cap to put down situational units such as a Support that slows, or a high damage Guard to pick off a single threat. In line with the overall tower defense genre, this flow clearly has shadows of classic RTS games (build up economy, expand, adapt).

The specific level design will ask players to adapt their strategies, sometimes quite drastically: for example one of my favorite level types has no energy generation at all, which forces players to thoughtfully utilize Vanguards throughout these levels as a source of economy.

The common game loop is like this: try out a new level a couple of times to understand the level design and different unit placement strategies; if needed, make some tweaks to your deck to address the level’s specific challenges and/or invest in leveling up your characters, and repeat the level until you are able to perfectly complete it. This is where F2P business model tensions creep in (more on this later).

Macro systems

I don’t play gacha RPG games deeply, but the general systems here look similar enough.

Monetization – premium currency can be converted / funneled through various systems to advance the player’s progression. Most notably, it can be used on 1) gacha draw of characters; and 2) Recharging stamina, which is consumed when levels are played. There are also various real-money packages that can be purchased directly, which offer a bunch of progression resources (crafting mats) such as XP cards for leveling up characters.

Character progression – 4 major components:

  • “Potential” upgrade – duplicate gacha draws of the same character can be consumed to improve character attributes (e.g. reduced cost) up to 5 times
  • Character leveling – only through consuming XP cards (which can be farmed / purchased)
    • “Elite” conversion – at max level, consuming a bunch of mats to advance the character to the next stage (unlock new abilities and passives), resetting the level count. Characters have several Elite stages, which extends the leveling ceiling
  • Abilities leveling – consuming mats to level up the ability level
  • Relationship – accumulated through playing with the character (also using the character in the base building system – see below), grants bonus stats

Another significant macro-system is the base building system, which takes cues from the likes of XCOM and Fallout Shelter:

Base at a glance – certainly looks similar to XCOM
Zoomed-in view of one of the rooms in the base

The base building system primarily serve the following goals:

  • Another axis of progression, as the base is gradually expanded over time and requires mats / resources to do so
  • A reliable source of economy for players to produce mats, and also grow relationships with characters
  • A presentation of the characters
  • Light social interactions amongst players
  • A engagement habit-forming hook, as you need to check in regularly (at least once a day) to collect your outputs and manage your team

As characters are the central driver for player engagement, both in terms of IP / narrative (these are appealing characters that attract players) and progression (want to see these characters become more powerful). The daily engagement thus is fulfilling quests and replaying levels (once you’ve perfectly completed a level, it can be auto-played at 2x speed) for mats, and doing base management about once or twice a day.

A quick note on the game’s social systems. The game directly requests the player’s phone number to register, and doesn’t have any alternative logins (e.g. QQ / wechat openID login). This shows the dev / publisher’s intent to grow its own social graph, but at a notable cost. There are no group social features such as guilds; you can add friends, which gives you the benefit of borrowing your friends’ characters for levels, and also for some light co-op related to base management. As is standard with many gacha games, the game also recommends strangers (through recommending characters to borrow).

Gameplay issues

Briefly discussing the game’s issues:

  • The obvious tension between gameplay and F2P. Game levels have recommended character levels, which compels players to go through the progression grind. After a few days of play the game quickly becomes primarily about the progression grind as the content pace slows down dramatically
  • To make matters worse (and also common for many gacha games), game levels have low replay value. Once the puzzle of a level is figured out, there is very little outcome risk (since there’s low execution variation). The only question then is whether your character progression meets the level’s demands. And thanks to the internet optimal level strategies are only a few clicks away, further aggravating this issue
  • Weak social play. The game feels decidedly lonely and the acceptance rate on stranger invites feel lower than other gacha games
  • Base management quickly becomes a chore

Let’s talk 996 / crunch and GaaS

by

This is going to be a difficult topic to write about, and one outside my usual product strategy focus. However, it is an incredibly important topic, and one that I have some frontline exposure to, thus I do want to exercise my mental muscles to put together some coherent points.

A brief recap

  • 996 refers to the often unspoken, but in some cases explicitly spelt-out work hours at many tech companies in China (9am-9pm, 6 days a week). There has been an ongoing debate in China the past few years, and recently Jack Ma and Richard Liu, founders of local tech giants Alibaba and JD, both voiced strong support of the model (Liu claimed even though he can’t work as hard as he used to, he still works 8-11-6 hours). Also, I don’t think it’s a pure coincidence that the rise in discussion of this topic is happening while the Chinese tech sector is widely predicated to face its most challenging years (and thus widespread hiring freezes or downsizing across the big names).
  • As I’ve commented before, the gaming vertical is viewed as much more strategic / integral to Chinese tech giants (it’s a core part of many Chinese tech conglomerates’ business model – gaming is seen as an obvious way to monetize traffic you’ve aggregated on your consumer-facing web properties, similar to how advertising is often the de-facto model in Silicon Valley). So the 996 conversation in China almost wholesale applies directly to the Chinese gaming industry.
  • In the west, gaming is a much more insular / isolated industry (in relation to the broader tech sector), with its own history of excessive work-hours and its own label to the issue – crunch. Crunch has always existed, thanks to the combination of scope and polish arms-race, ever-high player expectations, and fixed deadlines tied to major seasonal launch windows (and quarterly earnings pressure for the publicly listed games publishers). However, in recent years, the dramatic growth of Games-as-a-Service (GaaS) threatens to further exacerbate crunch, as GaaS fundamentally means a never-ending live development & release cycle (until the project hits end of lifecycle and is no longer financially viable), which puts sharp focus on the inability of studios to meet players’ insatiable demands for content. Fortnite and Apex Legends, two of the biggest names in GaaS currently, thus both had articles discussing the content / crunch tension.

  • Underlying GaaS economics

    • It’s worth doing some dissecting of GaaS economics. Firstly, GaaS similar to many general internet services, exhibit strong winner takes all tendencies:

    • Games, in particular those in a GaaS model, have sharp network effects (the more players you have, generally the better the experience for everyone, and conversely beneath a player-base threshold the game is unplayable)
    • Games, somewhat unique to other forms of leisure (especially physical sports, which PVP games share many other attributes with), have a much higher threshold for extended time engagement. This leads to intense competition for players’ time and significant effects of crowding out alternatives. (Competitive PVP games are often intentionally designed with extremely high skill-ceilings, which reward skill acquired through a mix of natural talent and lots of practice, which creates the need for high time investment)

    This means that in the realm of GaaS, game studios more than ever are in an arms-race – the profits windfall, if you make it to the top 1%, are a step-function compared to if you are “just average”. (To illustrate: Honor of Kings was estimated to have made over $1B in Feb – the Chinese New Year month – alone.)

    To make matters worse – the means of production in gaming, the technology, hardware and tools are constantly evolving, and there is an ongoing variable cost in adopting and adapting your development to stay current; in mobile, there is also a Herculean effort required in device compatibility that is directly proportional to the addressable install-base your game can run on, which is a critical element to unlocking network effects.

    Furthermore – as the biggest games have gone mainstream culturally (current representatives: Fortnite being the prime example in the west; PUBG Mobile in a swath of emerging markets such as India and the Middle East; and Honor of Kings in China), the content requirements are ever more diverse. When you are servicing a player-base of tens of millions (or hundreds of millions in mobile), you can’t just develop for the niche hardcore audience that your initial game thesis was founded on. You have to cater to a broad array of needs, for instance social networking, expression of individuality, vanity / showing off, and the pursuit of collecting content. You also have to work constantly to keep your game fresh, with novel product + marketing ideas such as an in-game concert or tie-ins with big brands outside of gaming.

    Globalization and the China angle

    To further complicate things, whatever fragile consensus or common language (or action, such as unionizing) studios and employees in the west can reach with regards to crunch, is almost immediately thrown out the window if we add in the Chinese studios.

    Zooming out back to the tech sector at large briefly – it is obviously with a heavy dose of self-interest that in the past 18 months prominent voices in the Silicon Valley VC community have made statements such as this or the following:

    “996” is the demanding work schedule many Chinese founders have organically adopted: 9am to 9pm, 6 days a week. To us, 996 captures the intensity, drive, and speed of Chinese internet companies, many of which are moving faster than their American counterparts.

    The raw capitalist greed on display aside, there is quite a lot of factual basis to these statements. This is the by-product of a globalized economy and the comparative advantage of nations – many of the same factors that led manufacturing jobs (across many industries) to move to China (and then to Southeast Asia) are at play here. What’s relatively new to the gaming industry, but the impact of which will almost certainly be more measurably felt in the near future, is the head-on competition from Chinese studios for a slice of the global gaming market share.

    The outputs of Chinese studios have historically been limited to emerging markets (Southeast Asia, for example) or “down-market” segments such as browser-based games. But we’ve clearly witnessed a turning point the past 12 months, with the likes of PUGB Mobile, Ring of Elysium on Steam (another Battle Royale, that got a decent 8.5 review by IGN) and Arena of Valor on the Nintendo Switch. The results are mixed, and there’s no shortage of “noob mistakes”, but it’s surely a sign of things to come. And whatever gaps there exist today between these hungry new entrants and the blue blood western studios, we could see them evaporate surprisingly quickly thanks to the “intensity, drive and speed of Chinese internet companies.”

    When you are between a rock and a hard place

    It feels like so far I’ve been very long-winded at painting a grim picture for employees from the perspective of crunch. So are there any hope at all with this confluence of industry trends?

    First off, for the individual talents / employees – you have to follow your own life compass. Don’t let others or “the company” make the decision for you. In the big picture, and in the long run, your health and your family are generally more important than your passion and your work (and just to be clear, the games industry, perhaps more than most other industries, leeches off of your passion) – if you agree with this statement you should keep this in mind when you are making the short-term decisions at work. But aside from that point, I don’t think there are absolute rules. As a 25 year-old I worked the occasional 100-hour week as a management consultant, and while those days look comical in the rear-view mirror, I don’t regret doing them – I learnt some stuff and it was a memorable life experience. But I’d certainly have extreme reservations about doing anything similar these days.

    Secondly, for studios, I don’t think there are any silver bullets, but the following might be able to move the needle:

    • This is incredibly challenging to do, but having an honest conversation with players about scope, polish and the grueling realities of game development could help partially reset unrealistic player expectations. Make players demand for a better work/life balance on behalf of employees. For poor analogies, see “responsibly grown coffee beans” or “carbon neutral products” – basically anything where the cost of rising to a higher standard (offseting a negative externality) is passed on to the consumer in a feel-good way
    • Don’t compete in red oceans; go for blue ocean opportunities. Stop participating in the scope and polish arms-race. For example, work on something like Minecraft when the rest of the industry is working on something like Call of Duty; or invest in a distinctive lower-fidelity art style that is cheaper to make given your pipeline and tools. Obviously, easier said than done
    • As a more specific instance of the above – work on a low scope project. Supercell is the best example in my opinion of doing this repeatedly with great success, despite my critique of their challenges. Hearthstone is another famous example
    • Collaborate with Chinese studios. For one thing, from an economics perspective, think of it as similar to industry consolidation, which helps with reducing competition (and thus reducing the arms-race). But more fundamentally, there are deeply complementary assets that western and Chinese studios could cross-leverage. This is why I still think Diablo Immortal was/is absolutely a right initiative for Blizzard – Netease to pursue

    (…And with that, I’d like to wrap up this post, which has in its own way grown way out of scope. This has been one of the more difficult posts to write, and the longest time/energy I spent on one in years. I hope it sparks some thoughts.)

    “Two Billion Gamers”

    by

    I’ve been meaning to write a post about cloud gaming for a while now – there’s a rather long draft sitting in my editor from several months ago (hope I can dust it off and ship it), back when Google generated some headlines with the Assassin’s Creed Odyssey streaming beta.

    Anyhow – the topic picked up again the past couple of weeks, this time with Microsoft Xbox making some bold proclamations.

    I for one, wholeheartedly welcome the “two billion gamers” concept, that apparently Microsoft has been talking about for at least a year now. This is an audacious vision and easy to rally around. (Tooting my own horns a bit – 3 years ago I wrote about the path to a billion MAU game… coincidentally also in a post about Microsoft.)

    The problem for me, with Microsoft’s stated strategy, is there are certainly competing paths to this vision, and Microsoft’s version is (of course) based on what’s feasible for Microsoft. However as an independent observer I do feel other paths are more likely to realize this goal.

    What do I mean by this – consider Microsoft’s industry position in the various gaming platforms:

    • PC – clearly Microsoft still is a dominant platform holder in the PC segment with the Windows platform. However the entire Wintel ecosystem is structurally challenged, and have been eclipsed/leapfrogged by mobile in most areas of personal computing. That puts a serious cloud over the future prospects of PC gaming, which remains one of the core niche cases for PC hardware
    • Console – Xbox unfortunately is an also-ran this gen
    • Mobile – Microsoft has tried in earnest, but failed, to establish its own mobile OS. Android and iOS have won, in various ways. Mobile has also turned out to be the majority growth driver of the entire gaming industry and its rapid ascension has toppled the old world order (see the rise of Chinese developers that I’ve frequently written about on this blog)

    Given this set of circumstances, it is logical for Microsoft to attempt to change the basis of competition – namely, to bet all-in on cloud gaming as a disrupter to established platform advantages, and go to a Post Platform world. To be clear, it faces obvious challenges:

    • At least so far, the stringent network requirements (especially latency for good feel) means cloud gaming remains a mature markets play, and thus dramatically limits the addressable audience. (For the two billion gamers vision to be achievable – markets like China, India, Indonesia and the Philippines must be unlocked, and at a rough glance cloud gaming faces severe hurdles in those markets)
    • It is not yet clear how other platform holders (i.e. Sony, Nintendo, Apple, Google) will react. Will they allow 3rd party devs to utilize Microsoft’s solutions on their platforms, and to what extent? For example, if a game is streamed onto iOS, can it thus bypass the App Store 30% revenue cut? Thus it’s unclear what exactly the value proposition is being offered to game developers (some will no doubt try)

    And this now leads to the competing paths part I alluded to earlier. In the emerging markets I mentioned above – China, India, Indonesia, Philippines, Vietnam, Thailand etc. – there is already a single platform with the clear runway to two billion gamers: mobile. Thus it seems more probable that Supercell or Tencent can ride the mobile gaming evolution to serving two billion gamers faster than Microsoft or Google can with a cloud gaming solution.

    This is a much belabored point – but North America truly is a mobile laggard and that biases the industry on its mobile perspective. So much so that when one of the biggest western publishers, Activision Blizzard, talk about “mobile is a top priority”, it inevitably invites players to complain and mock the company for being out of touch with its player-base. (Some of that criticism is deserved, but not all.) Still, I give Blizzard plenty of kudos for sticking to the mobile priority vision, and I suspect they will bear some surprisingly good fruit in the not too distant future.

    One last random tidbit that’s worth mentioning in this free-flow post: Vivo announced a fairly affordable ($450) high-end gaming phone, iQOO, last week. It sports a top of the line Qualcomm Snapdragon 855 chipset, but what’s more interesting to me are the gaming specific hardware features – the two bumper buttons (when you hold the phone in landscape mode) is a significant gameplay upgrade for games such as PUBG Mobile. I’ve long clamored for a mobile controller peripherals push by the platform holders such as Apple – building it directly into the device is also a great path. I expect more Chinese phones to add such features – and these Chinese brands just so happens are generally doing well in all the aforementioned emerging markets. (Sadly, I expect Apple to be quite a laggard here as well.)

    This example is to say – the mobile gaming ecosystem continues to rapidly evolve, and there are signs of further movement upmarket. (Speaking of which – the foldable phones by Huawei and Samsung, while unlikely to be more than proof-of-concept demos this year, speaks to interesting mobile hardware leaps in the next few years.) So if “two billion gamers” is the only metric to pursue – my money would be on mobile to hit it first.

    Supercell’s challenge

    by

    Supercell’s CEO Ilkka Paananen published an open letter / annual outlook that had some fantastic and honest discussions of his company and its products. I don’t know if gamers would care too much, but as a fellow game developer, I loved it, and had some points to comment.

    Before I dive in, a meta recap of Supercell – thanks to Finnish financial reporting rules, it has had to file annual reports and thus its performance has been in the public’s view. I cobbled together the below graph using a bunch of news reports over the past years:

    The narrative is pretty straightforward: after truly explosive business growth during 2012-2015, Supercell has gone into a period of decline. At the same time, the company has steadily added headcount, showing it is confident in its long-term future. (By all accounts it is still a very light-asset company, having less than 300 full-time employees. Just for comparison – Blizzard pulled in $2B in revenue in 2017 as well, and its headcount is easily 10x Supercell’s.)

    Back to the letter: the thing that struck me was in discussing the business operations, how much focus was related to China:

    One of the really big steps we took in 2018 was that we decided to start building a game studio in our Shanghai, China office. In the early days of Supercell, I thought that we would always be a single studio company – just to keep things simple. But I changed my mind when I got to know the Chinese games industry better. I admire how the best developers in China think about social game play and also how much new quality content they bring their players every month. There is definitely a lot us Western developers can learn. Our goal with this new studio is exactly the same as with our studio in Helsinki: create games for the global market, games that are played for years and remembered forever. We feel like this is a unique opportunity to bring together the best of two different worlds.

    And then later:

    One of our goals this year is to get better at creating more content for our players. This is a more interesting challenge to us than you might think. On one hand, we like our small team sizes because we believe that is one of the reasons we’ve been able to produce innovative games with fun core gameplay. On the other hand, there is only so much content a small group of people can do, no matter how talented they are or how hard they work. Anyway, we’ve concluded that this is something that we need to get better at. How do we keep the small team sizes that are so important for innovation, while getting much better at serving you, our players, with more content?

    We have now made the first few steps to improve this. One, we’ve partnered with some talented external studios who will be helping us to build more and better content. Two, we have invested more into tools & technology that will help us create content more efficiently. And three, we have slightly grown the size of the live teams (but only the live game teams) to be able to serve all of you better. What all of this means for a game like Brawl Stars, for example, is more brawlers, more skins and more environments being added more frequently.

    China wasn’t mentioned in the two paragraphs above, but IMO was an essential driver of the underlying business challenge facing Supercell. When Mr. Paananen wrote “I admire how the best developers in China think about social game play and also how much new quality content they bring their players every month,” I do believe this was not just paying lip service to his Chinese competitors, but an actual admission.

    Chinese developers work crazy hard (“996”), are not afraid to throw bodies at problems (some of the biggest mobile games in China often have dev + live teams of 300-500 people), and the industry has spawned a sophisticated production eco-system. What’s most noticeable is the clear production quality upgrade that I’ve talked about in the past, which is funded by a maturing value chain of outsourcing shops and specialized vendors. This enables Chinese games to afford to engage players with a seemingly endless stream of content at free or low price-points – gameplay modes, cosmetic items, quality upgrades to legacy content etc.

    In the face of this, Supercell’s games, despite often having superior & innovative gameplay, just look like demos; and the aggressive monetization of power ironically becomes a particular point of frustration for Chinese players. I say ironic – as Chinese players have no qualms monetizing for power, it is the transparent design, limited by content shallowness, that players complain about.

    Thus, Mr. Paananen’s comments about trying to strategically tackle content production, while maintaining the company’s small size (and the benefits that comes with being small). In my view this is not a nice-to-have initiative – if Supercell wants its games to succeed in China, which it seems to, this is a strategic imperative. Otherwise, Supercell games will be confined to a niche audience in China (and the business outlook limited), as it cannot retain the mass audience due to losing the content war to its Chinese market peers. If Supercell can achieve a breakthrough here – it will be the kind of boost it needs to reclaim growth.

    2018 year in review

    by

    This post is unfashionably late by a week, but anyhow, here are a few major themes for 2018 in my view.

    Chinese apps conquer the world

    Hard to resist the hyperbole, but 2018 saw some incredible progress for Chinese apps’ global ambitions. To be sure, the biggest successes were predictably from emerging markets (very notably, India), but apps like Tik Tok, PUBG Mobile, and Knives Out have taken formidable positions (and sizable revenue) in developed markets such as US and Japan.

    To pontificate on this, I believe the following factors play a role in this outcome and ongoing trend:

    • Relative lack of investment from American competitors in these emerging markets, due to perceived lack of infrastructure and low consumer spend. There’s also the question of business model fit, where Silicon Valley’s dependence on ads performs poorly in emerging markets
    • Chinese devs’ openness to hustle as needed and over-invest (willing to be inefficient but highly effective). The Chinese playbook is to try everything and see what sticks – but do it in lightning speed, which requires a high upfront investment1. This is also against a backdrop of a tightening home market (and for games, a regulatory freeze that signaled the arrival of winter), where “going overseas” is more than ever a strategic imperative for more and more businesses
    • A potential mindset advantage – unlike American companies which (stereo-typically) prefer a “one-size fits all” approach to global opportunities (and which often really means, built for the North American market, and hoping it is compatible with other markets), Chinese devs have seen how this did not work (for US products) in China. 2 They also likely more deeply understand the value of empowering the local team to make big decisions3

    The beginning of the end, of the PC (x86) platform

    Another hyperbole, it may seem (or a massive understatement, depending on which sectors/markets you look at), but consider the following that happened in 2018:

    • Intel’s significant ongoing woes with its 10nm manufacturing processes, in contrast with TSMC’s 7nm process that has already seen mass commercialization (e.g. Apple’s A12 Bionic chip)
    • Windows on ARM is now a thing, and Mac on ARM is on the horizon too with project Marzipan

    Both of these are symptoms of the gravitational pull of the mobile ecosystem. Specifically for the PC gaming sector, I’d say these are alarming long-term signals –

    • There is a non-trivial likelihood that x86 stops being a consumer computing platform, wholly replaced by ARM in a convergence form factor like the Surface
    • x86 gaming may survive as a standalone high-end market, but will face hostile underlying hardware economics (not unlike, say, DSLRs)
    • PC gaming could possibly survive the demise of the x86 platform, but it likely will be a massive blow to legacy content: many games from the past few decades may become unplayable (cloud gaming could be a solution there, but cloud gaming has its own critical dependencies)

    At this point we have to pause and say, what are we even talking about when we say “PC gaming”? It’s not up for much debate that the underlying Wintel platform has been disrupted, but if keyboard+mouse lives on as an input paradigm, is that all that’s needed for “PC games” to live on?

    It of course is not as simple as that – the migration away from x86 will be painful at the execution level for developers and end-users. Add to this mix tremors in the distribution (Discord and Epic becoming publishing platforms, and initiating a race to the bottom in rev-share), and the next few years look quite turbulent and interesting.

    Predictable mobile clones, and unpredictable market adoption

    For the last theme, I want to go back to the world of Chinese mobile games. 2018 was a year that continued themes I wrote about previously, where every genre conceivable had an earnest mobile clone effort. The thesis is simple: Chinese players prefer the mobile platform, so every game globally that has a fresh idea on PC/console was ripe to be taken to mobile.

    What was dramatically unpredictable, was just how strong the appetite was. Going back a year, even with the benefit of having played Netease’s first stabs at mobile battle royale, I would have said this genre has severe adoption constraints on mobile, mostly centered around the input. Oh how wrong I was. Also the sheer audacity of Tencent’s playbook – to launch 2 competing licensed PUBG mobile games simultaneously – surely invited many a raised eye-brow. Now all of that feels like ancient history – more players globally play the battle royale genre on mobile than on all other platforms combined, and it’s probably not even close.

    But if this were just about PUBG Mobile, which I feel I have talked about ad nauseam, it wouldn’t be a theme. There is more – Identity V, and LifeAfter (both by Netease) took concepts from relatively niche games Dead by Daylight and Rust, and successfully launched them to a wide Chinese mobile audience. The market performance of these games, despite clear technical drawbacks (in particular for LifeAfter), shows a huge appetite for “fresh gameplay”.

    The other side of this coin also bears a mention: Chinese game devs, in particular the large in-house studios of Tencent and Netease, now have well-rehearsed processes to quickly assemble and deploy large-sized teams (100s of devs) against opportunities deemed strategic. This means that any game that does not have a mobile strategy, regardless of how irrelevant the original devs believe mobile to be for them, will quickly (3-6 months) have a mobile clone if they stumble across success. (In this regard, single-player AAA games, especially those strongly narrative driven, remain relatively safe from clones.

    This is why, despite western gamers’ loud protests, Blizzard et al must march towards mobile – it’s not just about profit-seeking; in many ways it’s about long-term business viability.

    1. One gaming example is King of Glory‘s Battle Royale mode, which if I were to guess took 100 devs a few months of work; it did not gain market traction, but I don’t doubt that the studio would make the same bet again. ↩
    2. Example: PUBG Mobile has not only different store-fronts for North America vs Japan, but also different content. ↩
    3. For more on this point, Kai-Fu Lee expands on it at length in his book AI Superpowers. ↩

    2017 year in review

    by

    Going to try to talk about the 3 most interesting themes I felt, reflecting on the past year. As always, these themes take a heavy China perspective.

    The rise of PUBG and the even bigger rise of mobile PUBG-likes

    Globally in gaming, PUBG was the biggest breakout success of 2017, coming out of left-field (what are chances of a modder partnering with a Korean studio to successfully launch a global title?), and achieving impossible heights. If you said at the beginning of a year that a Korean PC game would break 3M PCU on Steam (and easily dethrone Dota2 as king of that platform), everyone (and especially all the industry insiders / experts) would have laughed. And yet this is exactly the appeal of the industry – what seems like ludicrous ideas can completely transform the landscape in a brief amount of time.

    I wrote about what I thought made PUBG such an appealing experience; the follow-up to this success story is even more noteworthy though. In the brief span of a couple of months, a horde of Chinese mobile PUBG-likes were launched, and a couple of them, Rules of Survival and Knives Out (both from Netease) also hastily released global versions. These games have performed surprisingly strongly in key markets globally – Knives Out has over 100M registered users and 20M DAU (majority of which is from China, of course), while Rules of Survival has some 50+M registered users (and performed better than Knives Out in the US).

    If you had a chance to try any of these games, you will see that they defy all common (western) notions of what a mobile game is. They are janky and hardcore, and are almost literal ports of the PC PUBG experience. It’s easy and (to a large extent, fair) to deride/dismiss these games as copy-cats; but that doesn’t answer the question of why they are so successful (e.g. who’s playing these games? Lots of people across a far broader demographic than PC gamers, it turns out), and what PUBG should do.1

    Chinese mobile games’ huge production quality upgrade

    The production quality upgrade of top Chinese mobile games is something that has been a steady evolution, but it has hit me in the head recently just how big the improvement has been, even compared to a couple of years ago.

    Some notable games that illustrate this:

    • Netease’s Onmyoji, launched at the end of 2016 and which is currently in closed beta for North America, is a game that screams “look at my shiny anime art style and impressive VO featuring top Japanese voice talents!”
    • Honor of Kings’ (Arena of Valor in the west) 60fps update, which launched in January 2017, completely changed the game’s experience. It’s as impressive a visual update as, say, Apple’s rollout of Retina displays IMO
    • Honkai Impact 3: anime-style ARPG, from a Shanghai-based studio

    What these games represent is the coming of age of Chinese studios: these teams have mature dev and art pipelines, and have established processes that support large scale development on mobile. By large scale, I mean 200+ headcount projects (such as the Tencent mobile PUBG games in development), which I dare say I can’t think of one western dev at this scale on mobile.

    And furthermore – these large teams all have launched numerous projects, so they have gone through trial by fire. So I think Chinese studios have a firm lead here just by virtue of accumulated experience.

    The success of the Switch and what it means for mobile

    In more general gaming, the Nintendo Switch also heavily outperformed versus expectations, largely driven by strong first-party titles such as The Legend of Zelda: Breath of the Wild.

    What I personally found amusing, was the player raves around the mobile benefits of the Switch – e.g. Mario Kart parties during lunch-break at the office. Not to take anything away from Nintendo building a great experience – but this is exactly the type of thing that Chinese gamers have enjoyed for years on mobile, with core multiplayer games such as Honor of Kings.

    If you were to compare the Switch with an iPhone (or top of the line Android), the only real advantages (from developers’ perspective) the Switch enjoys are 1) physical controller 2) Nintendo first-party games 3) premium pricing for games which players have accepted. For 2 & 3, it’s hard to see that change given app store dynamics and Nintendo’s interests, but 1) is an area where I could see some material changes.

    I know in some ways this sounds backwards, but I believe firmly that controller peripherals offer a major opportunity to elevate core gaming on mobile. This is a classic multi-sided chicken & egg problem: without strong adoption of controllers, games have little interest to support/demand them; without great games, controllers can’t sell. I wouldn’t be surprised to see some shift arising from China – there are already some interesting controller plays at small scale, such as this brand of controllers with dedicated software optimizations for playing Honor of Kings / Rules of Survival.2

    1. PUBG’s China publisher, Tencent, is not sitting around to find out – it has not 1 but 2 PUBG-licensed mobile games in production. ↩
    2. So far this is still janky, as these hardware players don’t have official support and have to operate in a grey-market fashion in terms of how they integrate with the games – a bit like the jailbreak app days on iOS. ↩

    Revisiting mobile platform advantage

    by

    This post, which is mostly a rehash of ideas I’ve written about previously, is inspired by the following two gaming industry stories:

    These two stories are very interesting when looked at side by side. On the one-hand, Overwatch is certainly one of the best new-game launches ever (for console/PC platforms), and has great potential to engage millions of players for years to come (in the clichéd games-as-a-service model). On the other hand, its impact (whether monetary or number of players or player-hours served) is completely overshadowed by Honor of Kings, and we are not talking about a small gap – it’s probably a 2-5x difference today and could grow to an order of magnitude difference (10x).

    In one sense, a hardcore game having fewer players than a more accessible/casual game is nothing new (top Facebook games easily had tens of millions of players). However, compared to other popular mobile games (e.g. Pokemon Go), Honor of Kings is a much more hardcore game and it certainly serves plenty of hardcore gamers.

    It is from this lens – viewing Honor of Kings as a game that’s closer in spirit and purpose to League of Legends/Overwatch as opposed to Candy Crush/Farmville – that the expert opinion in the 2nd link above is even more interesting. I have a lot of respect for the opinions voiced from the 4 industry peers interviewed – they made many reasonable points, such as concern over the average session length as a blocker for attracting players. However, I also think these opinions are founded on some assumptions about what mobile games are / aren’t which may not actually hold.

    The biggest shift in perspective required is not viewing mobile as an inferior platform versus console/PC for gaming, but rather a superior platform. Mobile does have some severe constraints (such as the lack of physical feedback for input, and input often taking up valuable screen real-estate), but many commonly-cited constraints are artificial. Take average session length – Honor of Kings has easily proved that almost 200 million players in China 1 have no problem regularly spending chunks of 20 minutes for one match, which is certainly mind-boggling for anyone used to thinking of single game length of under 5 minutes as a golden rule for mobile. If you are able to suspend belief and imagine players spending hours a day gaming on their phone (which they do in China), your perspective of what games are possible on mobile changes. Another common constraint I see is somehow phone-screens “are not large enough for complex gaming”, and devs end up optimizing for tablets2.

    Put another way, I see a self-reinforcing cycle – if devs don’t believe in the potential of mobile and blindly accept conventional-wisdom constraints, then they can only make games that operate under these constraints3.

    I’m often reminded of phone industry experts reactions to the iPhone when it was first announced 10 years ago. A lot of very smart people made some terrible predictions, when in hindsight the conclusion was so obvious. I feel more and more that mobile gaming will continue to grow and grow, and eventually force devs that prioritize console/PC to make some very painful transitions.

    1. Based on analyst estimates in this Bloomberg article. ↩
    2. e.g. Vainglory was clearly seen as a game intended to be played on tablets, as evidenced by its marketing videos ↩
    3. i.e. the casual arcade / casino and async strategy games that dominate the US app store rankings ↩

    Innovating beyond Clash Royale

    by

    Some short thoughts here, after sampling Netmarble’s Star Wars: Force Arena, which can be described as a hybrid between Clash Royale and Vainglory.

    First of all, it’s no surprise devs are looking at Clash Royale and trying to build upon its success formula. From the initial wave of cheap Chinese knockoffs, we are now beginning to see the more serious attempts to innovate. Aside from Star Wars: Force Arena, there’s also the recently announced Smite Rivals, which we can get a sense of from this teaser video:

    https://www.youtube.com/watch?v=73W0QbYK4eU

    Both Smite Rivals and Star Wars: Force Arena implemented the idea of 3 lanes. In Force Arena’s case, the 3-lanes setup is tied to 2v2.1

    Force Arena also went the extra step of the MOBA-like camera angle, which likely went hand-in-hand with the addition of the hero gameplay. I’ve mixed views on this concept. On one hand, having unique hero characters per deck is intuitive and awesome (and ties in closely to the fantasy); on the other hand, actually controlling the heroes feel cumbersome. There might have been another direction with making the heroes just unique cards that anchor the decks.

    This camera angle adjustment also required a battlefield map that needs panning to navigate around. In my opinion this was a high cost to pay – it costs the player time and energy to move around the battlefield, and a lot of the action is happening off-screen.

    Another issue I have with Force Arena is the units design. Being a IP-licensed game, there are basic rules around what units could be. Granted, this is a lot of the fantasy that drive fans to play this game, but at the same time, it is likely limiting game design. At least in my limited play-time, the action does not look as intuitive as Clash Royale, and many units feel “meh” (a lot of humanoids, and the shapes look similar).

    The net result of all this is a game that’s more exhausting to play, and less satisfying to watch. To me it’s still a worthwhile experiment, and doing the comparison to Clash Royale helps highlight some key design insights (e.g. such as keeping the action on one fixed screen). I also still think as a general space this area is ripe for further innovation.

     

    1. Incidentally, this was also a direction some coworkers brought up in our water-cooler talks, so it’s exciting in that sense to see the idea brought to life and tested. ↩

    Thoughts on Top Eleven

    by

    As a casual player of football sims, I’ve been very impressed with the thoughtfulness of Top Eleven, and here are some thoughts to sum up what I found interesting.

    (For an overview of the game – check out this review. This is a game that’s been around for 6 years and surpassed 100M registered players in early 2015.)

    At a high level, Top Eleven‘s core design thesis appears to be “how do we take the incredibly addictive gameplay of Football Manager, and make it a massively multiplayer social/mobile game?”. This spills out into the major gameplay systems:

    • Leagues made of human player-managed teams, with a long and tightly regimented progression system using the real-world concept of seasons (every 28 days in real-life form a season, and cup and league matches are distributed throughout the season so that on average you play 1-2 matches per 24 hours)
    • A 2d match engine familiar to any Football Manager veteran, and the associated tactics and training systems
    • An real-time auction-house (much like Diablo 3′s auction house) that serves as the transfers market
    • Being a free-to-play game, a set of virtual currencies that restricts player actions and provide some amount of pay for power

    My first impression after playing the past 3 months, is that this is a well-tuned set of systems, and the player experience is pretty satisfying even for someone who hasn’t monetized (my basic principle for playing mobile games is to not monetize and test the design for a non-paying player). In every season I’ve competed in I’ve won the League with the limited resources available to a free player, and I’ve won the Champions League (a more competitive tournament) once. The holy grail of course is the treble (winning the League, the Cup and the Champions League in one season), and that is challenging but doesn’t seem completely out of reach.

    In particular, I’d like to call out the League progression design as simple, effective and clever. It’s effectively a cohorts based design –

    • when you join you are placed in a league with players who started around the same time as you, and therefore have similar amounts of resources;
    • Every season the top 50% of the league are promoted to the next level, while the bottom 50% stay in place;
    • For each level, there are tight restrictions on the quality of player you could acquire, regardless of how much money you are willing to spend.

    These measures ensure that on average the players progress through the game at a similar pace and are always in an environment where there are worthy opponents.

    Similarly, the auction house design is also simple but extremely effective. There are a few additional options for player transactions, but the basic auction house is a real-time feed of player listings with deadlines, using an English auction format:

    • Players can only see and bid on listings appropriate for their level – again, carefully segregating the player population and controling the experience, and also creating a healthy economy of auctions (a higher level player’s 3-star NPC is an all-star for a lower level player);
    • The seller sets the initial floor price, and each bid increases the price by a set amount;
    • If there is only 1 bid for a listing, the bid wins when the listing expires;
    • If there are more than 1 suitor for a listing, the suitors face off in an unlimited number of short-session follow-up rounds (starting at 1 minute, and quickly reducing to 20-second rounds);
    • Each round a suitor must place at least 1 bid to be eligible for the next round, and the auction ends when there is only one bidder or none (the highest bid from previous round wins) in a round.

    The catch for this system is that each bid consumes a super-rare virtual currency called a token. (An engaged, highly active player can expect to earn 30-50 tokens for free per season; in other words, a little more than 1 token per day.) This gives each action a lot of weight, and creates interesting psychological influences on players. From players’ perspective, it’s advisable to avoid a pro-longed bidding war for a single listing, but in the spur of the moment (20 seconds to make a decision), it’s easy to be trapped in a deadlock.

    This design also creates room for lots of auction strategies, which creates uncertainty and fun for players. For example a basic technique is to track an empty listing and put in a bid in the last few seconds, to ensure the token is not wasted. Sometimes though, this backfires and you will see several last-second bids, which sets up a bidding war. Similarly there’s lots of mind-games in the follow-up rounds: do you wait to put in a bid in the last few seconds of a round (which puts you as the price leader for the next round, and also can surprise a rival who didn’t put in a bid); or do you bid early each round to signal that “I’ve got plenty of tokens, I’m going to win this no matter what”?

    Having said all the above, Top Eleven is not without its issues. In particular, churned players’ teams pose an interesting problem. In my 3rd & 4th seasons, a vast majority of the teams in my League were clearly occupied by churned players. This meant that their neglected teams were weak and didn’t pose an interesting challenge, and in effect the lengthy League season came down to a few matches between the active players. This may be due to the inherent high churn at the beginning of the funnel (my current season seems to have the right mix of teams), but I wonder if there are better ways to solve this.