The Birth Of The Mobile MMOG

Major Fambrough: "You wish to see the frontier?"
John Dunbar: "Yes sir, before it's gone."
-Dances with Wolves

Mobile phones are becoming increasingly powerful computers, complete with operating systems and standard APIs. Although most phones on the market are still useless for anything more than phone calls, we are slowly moving into an era where many people will walk around with a (potential) networked game console in their hand.

As mobile phones increasingly resemble networked computers, they become more interesting as platforms for multiplayer games. Yet, as someone who helped develop three MMOGs (including World in War, which was recently released for mobile devices), I can vouch for the fact that application development for mobile phones is still in its infancy. Mobile game development lacks many of the resources that more mature gaming platforms offer. Based on my experience, I will describe what I believe are the essential requirements for making a successful MMOG for mobile devices - a genre that I'll refer to as "3MOG" in this article.

The MMOG Market Today

According to MMO Worlds (www.mmoworlds.com), there are almost 60 MMOGs on the market today (not counting the various expansion packs for games like EverQuest), and many times that number of MMOGs under development. The 20 most well known MMOGs on the market are listed in Table 1.

The MMOG genre has been a powerful buzzword in the game industry for years now. I see five interconnected reasons for this:

To date, not many 3MOGs have been released, but that will change over the next few years. A good example of a 3MOG currently on the market is TibiaME from German developer Cipsoft (see Figure 1). TibiaME is what you might expect a MMOG to look like on a mobile phone; it resembles popular PC-based MMOGs like Ultima Online and Lineage.

My company's first major 3MOG is called Football Manager, which we began developing in 2000. Now on the market, the game has thousands of monthly paying customers and is one of the largest Swedish online games. The premium television channel Canal+ markets it, but it's only available in Sweden and Norway at the moment.

The game is accessible via mobile phone too, allowing players to interact with their team at any given moment (Figure 3).

The Challenge

Creating a 3MOG offers many interesting development challenges, some of which I will describe. Note that my experience is based largely upon my experience as a developer in Sweden; some of the factors I describe may vary from region to region.

Latency

Latency is always an important issue when developing multiplayer games. Most mobile phones run on GPRS networks, which have very high latencies (bad for games). Whereas latency in network calls for PC games is measured in hundreds of milliseconds, for mobile phones latency is typically measured in seconds. Imagine trying to play a game of Battlefield 1942 and always seeing what happened four seconds ago.

While getting the correct packets on a mobile phone can take more time than on a PC, the speed of sending data is comparatively not so bad. It typically ranges from 9.6Kbs to (in theory) 172Kbs, compared to a PC's dial-up modem speed of 56Kbs. Therefore, as long as we can hide the latency in a mobile game's design, there is actually a pretty good network to work with. Table 2 shows some values that we have seen in our Swedish networks (these are anecdotal measurements).

Design to Hide

How can these long latencies on mobile networks be hidden from players? One simple answer is to use a turn-based game design. This kind of game works poorly over the Internet, since it's hard to hassle an opponent for taking too long during his turn. Thus, turn-based games have been much more successful in "hot seat" versions. Another drawback to turn-based games is that the more players there are, the less playtime each person gets.

Tick-based gaming is a solution to this problem. In tick-based gameplay, you take a turn-based game and allow all players to plan their moves ahead of time, and then the game executes all the moves simultaneously. This model is fairly common in massively multiplayer web games , but it hasn't been adopted by many high-budget projects (a notable exception is UltraCorps). The tick-based model works best for strategy games, in strategy sub-genres like "manager" games, war games, resource management games, and so on.

Of course, this is this is not the way we are used to playing MMOGs. However, the design of many real-time multiplayer games could be accommodated to handle much higher latencies if the game was redesigned with those technology limitations in mind. While a multiplayer first-person shooter probably couldn't be redesigned to handle high latencies (since actions are so direct and gameplay requires fast action), games that have natural delays in them might be able to hide latency. For instance, imagine a game in which you command a large Spanish galleon. When you give the order to lower the sails or make a full turn, it takes the ship some time. Latencies could be hidden between the time the order was issued and its execution completed.

Another good example of hiding latency is seen in The Sims, where the user influences indirect control over the game characters. Since the characters move around on their own, even without the player's direction, the player has something to watch while the command takes effect, and there are plausible reasons for the character to not react instantaneously.

Device Anarchy

In the beginning of the PC era there were many different PC manufacturers -- just like the mobile phone industry today. Over time, the PC market has become rather homogenous. Surely not to the extent of the console market, where you know precisely what hardware the game will be played on, but there are certain standards that you can rely on like screen size, amount of memory, and input devices like mouse, keyboard and perhaps joystick.

The mobile phone market doesn't offer as much hardware certainty as today's PCs. For instance, in Sweden there are about 80 phone models from about 10 manufactures. The screen sizes vary wildly, as do the number of buttons and their locations on the phones. It may go without saying that until there are standards for the most basic hardware capabilities, developing mobile games will be difficult.

As you can see in Table 3, WAP has the largest market share in Sweden with over 60 phone models available, but WAP supports fewer technologies suitable for interactive games. In Sweden, two million people out of a total population of 8.5 million have access to a WAP-based phone, but only 100,000 have ever used these capabilities . Among the programming environments for interactive applications, J2ME covers the largest mobile phone model segment, but it has several drawbacks when compared to C++. For instance, with J2ME you must rely on the Java Virtual Machine and you have less control over function calls -- and that can mean less control over the network connection. Network disconnections can add latencies of up to 10 seconds per network call, which can be a big drawback for a 3MOG. In addition, with J2ME you often end up trying to fix bugs caused by the Java Virtual Machine, not your code.

Screen Size, Input and Other Limitations

Screen Size. There are two major hassles when it comes to screen size on a mobile phone: it is small and there is no standard size (not even a standard proportion!).

Building an interface that supports different screen sizes is incredibly hard. A heads-up display that has been optimized for 800x600 pixels will not look as good on 1600x1200 (and excludes game play in 640x480) and will take valuable development time if your ambition is to make it work at all resolutions.

The only real workaround to this problem is to use tile-based graphics and accept that text will break for a new line whenever it feels like it. Full-screen 2D bitmaps are almost out of the question.

Note that small screen size is not always a bad thing when it comes to game development. Players do not (yet) expect the quality of graphics on mobile phone to match that of Half-Life 2, and these lower player expectations can work in a developer's favor. When it comes right down to it, the display size has little impact on the success of the title.

Application size limit. Application size is limited on many models. A common size limit is 64KB, which is tiny for MMOGs. Some more capable phones have an extra memory card, allowing applications to be several megabytes in size. Needless to say, it is easier to develop games for the latter.

Color depth. Color depth also varies substantially between phones, but luckily there are not an infinite amount of color depths from which to choose. With a few exceptions, most phones with application development possibilities support 12-bit (4,096 colors) or 16-bit (65,535 colors) color depth. Both are quite sufficient, so game developers do not have to focus too much energy on the differences between the two.

Operator Issues

Internet connectivity for mobile phones isn't as easy as it is for PCs. There can be no MMOG without networking, and when it comes to Internet connectivity for phones, everything relies on the capabilities of the operator. On mobile phones, network packages are transmitted via the operator's software, and in many cases those messages are like frogs crossing a highway: sometimes they make it, sometimes they don't. This problem is bound to work itself out with time. Currently, however, all operators that you plan to support need to be thoroughly tested (which isn't cheap if you want to launch your game in several countries simultaneously). My suggestion is to launch your game in you own country where it's easiest to contact mobile operators to troubleshoot any connection problems. A word to the wise: operators have their own schedules and product plans, and many aren't concerned with the deadlines and constraints of a small game developer. Make sure you build some flexibility into your schedule and save yourself some stress.

User Behavior

Phone usage behavior varies greatly from PC and console playing habits. This must be taken in consideration when designing a 3MOG. Studies show that the average gaming session on a phone lasts just a few minutes. In some respects, this fact bolsters the case for 3MOGs since a persistent world can make better use of short playing cycles than a game that requires a player to start a new session each time the game is played.

Mobile games must behave politely and accept that the player's situation must always come first. A game in which the player's character dies and can't be resurrected -- just because the player got off the bus or answered a phone call -- will aggravate users and result in fewer players and lower revenues. Devising multiplayer functionality to accommodate frequently distracted players is one of the great challenges facing 3MOG designers. Indirect control of your character might be one way to overcome this. If the game character can make its own decisions, the sudden absence of the player is not as obvious.

The Publisher's Role

As a 3MOG developer, you're working in one of the game industry's frontiers. As such, you'll be on your own in many respects - perhaps even without a publisher. Publishers are very risk averse these days, and they all seem to be waiting for someone else to make a move in this genre. Perhaps that means that there's a window opportunity for a brave publishing company.

In our case, we financed our titles without publishers and brought our games to market by making deals with media companies instead. Media companies often have large audiences already and are always looking for new ways to make money from them. MMOGs and media companies seem to be a good match: premium games looking for a large audience.

If you have an idea for a 3MOG, the most likely way it will get financed is from your own wallet. Don't despair, though -- one of our biggest lessons has been that it doesn't take a bankroll of $50 million to get a game started. We've found that players like feeling that they are part of the game creation process, and will accept fairly rudimentary gameplay if they have a hand in helping the developers mold it into something grander over time.

The Opportunity

Admittedly, if you are thinking about creating a 3MOG, this article might not offer you all the encouragement you were looking for. Don't give up just yet, however. Knowing the obstacles is at least as important as knowing the opportunities. Now that we've touched upon many of the hurdles you could face, let's look at the opportunities that await.

A commonly asked (and very important) question is, "Will mobile gaming become really big?" In response, I've heard people say things like, "I doubt it -- I'll never sit on my sofa and play a game on a tiny phone display". The truth is, neither would I, and I don't believe mobile gaming can compete with console gaming in those same terms. However, I do believe that it's a complement to console games, for several reasons:

Fortunately, the similarities between web and phone applications are not very large. If you're your game was written in Java, it might even be possible to reuse classes between different platforms, too.

In spite of all the exciting abilities the phone, as a gaming platform, it has not the ultimate sound and visual experience and will not have for a foreseeable future. By adding a web (or even a PC) interface, your game can become a richer experience. Cross-platform games also increase the feeling of a game that expands out of its medium, and strengthens the players' idea of what the game world really is.

Resources