Author: by Ethan Einhorn

[The first was an App Store sensation, but what of its sequel? Ethan Einhorn, Sega's associate creative director on Super Monkey Ball 2 spells out the processes that led to the creation of the second game in the series -- including all the major triumphs and mistakes.]

The original Super Monkey Ball on iPhone was the number one paid app at the launch of the App Store. Rolling Ai-Ai the monkey around in 3D tied perfectly with the tilt/accelerometer functionality of the iPhone; this new way to play allowed SEGA to re-invent a classic intellectual property. The first game continues to sell well, and is held up by enthusiast gamers as a worthy challenge to their skills.

An interesting moment as a producer on Super Monkey Ball came when three of our testers insisted that they preferred the iPhone game to earlier versions in the series, which used an analog stick for control.

I asked them how long it had taken them to get used to the new mechanics, and they all guessed about 45 minutes. It was great to hear that they loved the game, but the learning curve represented a big problem: when considering the needs of the broad audience the game was intended for, the ramp-up time was about 43 minutes too long.

Sure enough, we saw a significant split in audience response to the finished game, based on this issue. Very patient gamers tended to respond enthusiastically, while users looking for instant fun felt that the game bordered on the unplayable.

A large number of players found that they simply could not get past the seventh board in the game, because turning using tilt proved to be far too sensitive.

We updated the game to include an on-screen gauge that let players know when they were holding the device at the correct angle for play, and we added a tutorial, but we knew that more would have to be done to make this Super Monkey Ball truly accessible to the audience that was enjoying Peggle and Flight Control.

We knew we needed to rebuild from scratch, and we were given that opportunity with Super Monkey Ball 2.

In addition to streamlined controls, we wanted to make sure that Super Monkey Ball 2 delivered cute 3D characters (the first game used sprites), multi-player racing, and a selection of mini-games, all key components to the series' DNA on traditional consoles. As character-driven platforming went, we wanted this game to set a new high bar on iPhone.

What Went Right

1. Building time into the schedule to iterate on the core play mechanics.

We knew how important balanced controls were going to be for this sequel. If we weren't able to fix the first game's biggest flaw, the project would be considered a failure by all involved.

Making tilt control work properly in 3D is much harder than most people realize. Unlike a driving game, which uses only one axis (tilting left and right only), 3D tilt requires left, right, forward and backward movement simultaneously, and this can only be done well at a narrowly defined play angle.

Consider: playing Super Monkey Ball 2 at a 30 degree angle can be done comfortably, but if you hold the iPhone over your head (say, while lying in bed), 3D tilt will not register at all. Calibration options are often suggested to overcome this, but if you look at 3D tilt in games that offer such features, it's easy to "break" the games by maxing out the sensitivity sliders. When a player does this, the movable object on the screen (usually a ball) can't shift left and right anymore.

We felt it was best to settle on one playing angle, and then test the hell out of it. This, combined with looking very carefully at the motion control data from the Wii Super Monkey Ball game, helped us come up with a control scheme that's proven to be friendly to newcomers and experienced players alike. It took about five complete passes to lock this down, and it was important that we front-loaded our schedule with time to work on this.

2. During art passes, focusing on the Monkeys first.

Character is one of the most important elements of any Monkey Ball game; it's so much more interesting to roll around a cute simian than a shiny, blank-slate marble! In moving the characters from 2D to 3D, we needed to make sure that we had character poly count allowances that made the four stars (Ai-Ai, Mee Mee, Baby and Gongon) look as good on iPhone as they did on consoles.

Once we had the Monkeys modeled, rigged and lit, we went onto populating the environments, allowing the characters to take aesthetic priority. This made us much more comfortable determining how many background props to create, and which particle effects to employ in the final art pass.

3. Ensuring full compatibility.

Back when the iPhone App Store launched, developers only had to worry about three types of hardware (Edge iPhone, 3G iPhone, and iPod Touch Gen 1) and one firmware platform.

By the time SMB2 launched, there were iPhone handsets available with different capabilities (notably the 3GS vs. 3G), and firmware updates meant that some consumers would be up to date (with 3.0 firmware), while other users lagged behind (sticking to 2.21 firmware).

We didn't want to let any of our audience have a sub-par experience, so we ensured that Super Monkey Ball 2 ran perfectly on all platforms. The finished product looks cutting edge on any iPhone device, including the launch machines. This was also critical for multiplayer gaming (race mode); in the final version of the game, no advantage is provided to players with more powerful handsets.

4. Going big.

It would have been easy enough to use the same engine we had employed in Super Monkey Ball 1 (it still holds up pretty well), create some new character sprites, tweak the play controls, and stamp a "2" on the app. But that approach wouldn't have established a game that we'd be comfortable building upon moving forward.

By aggressively upping the quality bar in every part of the game, and building all-new play experiences into the package, we are able to increase consumer confidence in our ability to manage our core brands on iPhone. When a Sega sequel hits the App Store, you know it's not just going to be more of the same.

So what, then, are the key features that support our claim of going big?

Competitive Multiplayer: The original Super Monkey Ball on Gamecube, a launch title for that platform, gave players a chance to race against each other in a split screen mode. iPhone's local wi-fi compatibility allowed us to bring this feature back, with the added benefit of allowing all four players to enjoy a full screen experience. This adds a lot to the game's replay, as all 115 boards support multiplayer racing.

Mini-games: Even the DS version of Super Monkey Ball features mini-games, and we couldn't let the iPhone sequel launch without addressing this notable omission from SMB1. Solid pass-along multiplayer fun was crucial to cultivating a good multiplayer experience, so we chose experiences that worked best taking turns: bowling, mini-golf, and target.

It would have been great to have all three mini-games available at launch (more on that later), but we're happy we got at least one in (bowling). Some players have noted in user reviews that they actually come back to bowling more frequently than they do the main game!

Full graphical revamp: When the App Store debuted, Apple used SMB1 to show off what 3D gaming looked like on the iPhone. The game still looks good, but we've come a long way since then. In the era of Real Racing and N.O.V.A, we needed to make sure Ai-Ai and his pals still set the bar for graphical excellence on the device. Editors have been favorably comparing the visuals of SMB2 to the Gamecube version of the game, so I'm confident in saying we hit this mark.

5. Listening to consumers.

One of the best things about the App Store is that publishers get immediate feedback from their user base, in the form of user reviews. Sega takes that information very seriously when working on updates and sequels.

Much of Super Monkey Ball 2's feature set was formed in response to the feedback we got on the first game, and as we move forward, we're closely eyeing the comments people are dropping about their experience playing Super Monkey Ball 2.

When users take the time to write down their feedback, they should know that people at Sega are reading the feedback and taking notes.

Customers who reviewed the first game gave us insight into the existence of a real split between the needs and patience levels of core and casual players.

Core users will take the time to master challenging mechanics, and will even come back to rewrite review posts once they have a better handle on the experience. Casual players are less interested in an intense challenge; they want consistent rewards, and they need them to pop up frequently.

If you're working on a legacy product, be prepared for lots of requests from fans for you to adhere more closely to the qualities of the original game -- there was a clear demand for mini-games, given the SMB1 user feedback.

We also learned that people tend to love or hate a game -- most users will give you either 5 stars (highest score) or 1 star (lowest score). And users will make it very clear when they don't understand a specific mechanic. All this information helped us better tailor the SMB2 experience to the broadest possible audience.

When it comes to SMB2's Customer Feedback, we learned even more.

In the first few weeks, we faced some of the most critical feedback, because users had to either go out of their way to post their notes (since they aren't asked for their feedback during play sessions) or respond to a prompt that pops up when they removed the app from their device. Most of the consumers that are happy with your game won't say anything about it until much later, when they're finally removing the game months later to clear up handset space.

Overall, the early SMB2 feedback indicated that people thought the game was much better looking and easier to play; that perhaps it was a little too easy and short (though I like to think 115 boards isn't too short for most players); that rapid price drops are now an expectation; and that people are generally okay with having core content promised in updates, provided they won't be asked to pay for it when it appears.

We ran into an interesting situation on user reviews with SMB2 at launch: we learned from reviews that the game would not work on jailbroken handsets. We only test our games on Apple-approved firmware, so we were taken off guard when we received a high volume of complaints about this. Players with jailbroken handsets are in the minority, but in the first week, nearly 50 percent of our reviews were from consumers frustrated that the game would not run properly on their jailbroken devices.

It should be noted that these complaints were coming from jailbreakers that had purchased the game -- you can't post feedback if you haven't paid for a game. This is an example of a vocal minority that can appear much larger when your App first hits the scene, and highlights a real challenge -- what do you do when users become upset over an issue that is out of a publisher's range to address?

Of course, we listen to everyone, and do what we can to keep everybody happy, but it can take some time to sort out the big issues that most players want addressed. If the number of patches you can post are limited by the constraints of the project (a common issue when you're working with external development resources), you want to make sure that you laser-focus on the issues that are most important to the largest audience.

Another great place to get user feedback is in the Touch Arcade player forum. This has become the most popular place for dedicated iPhone gamers to share their thoughts about new releases, and you get a great sense of how your game compares to everything else on the market.

One advantage of watching forum posts is that you get consumer feedback from people who are considering a purchase, but are on the fence. What needs to be there to convince them to jump in? And how do those playing the game respond when asked the most common Touch Arcade forum question: is the game worth buying at its current price?

What Went Wrong

1. Overestimating the demand for local multiplayer gaming on iPhone.

One of our primary focuses on Super Monkey Ball 2 was to offer a local wi-fi gaming experience that we felt would be comparable to the competitive mode on Mario Kart DS. As stated earlier, on SMB2, up to four people can simultaneously race through any of the 115 boards in the game, provided they are near a wi-fi hotspot.

It's a very fun, balanced play experience that very few consumers seem to be bothering to use. Virtually all of our feedback from consumers has focused on the single-player experience, and if we had known that interest in the multiplayer would be so limited, we may have dropped it in favor of adding more content to the solo experience (or adding more mini-games).

2. No 3.0 firmware functionality.

In an effort to ensure that our game worked seamlessly on all hardware/firmware configurations, we opted not to present any 3.0 firmware-exclusive features. It seemed safest to build to the 2.2.1 firmware instead, which guaranteed full compatibility across all handsets.

Drawbacks to this approach included substantially reduced user interest in the multiplayer option, as Bluetooth support (only available on 3.0 firmware) was not made available as an option for connecting devices.

It also reduced our ability to let consumers know when new features were available, and made it impossible to integrate a micro-transaction driven storefront.

3. No leaderboard or achievement support.

It'd be great if the iPhone ecosystem featured a universally employed achievement and leaderboard system, as Xbox 360 and PlayStation 3 do. But since it doesn't, the issue of integrating these features becomes a bit complicated, particularly for major publishers.

Fortunately, these are the kinds of features that are possible to activate through updates, but it would have been ideal to have launched with them day one.

Super Monkey Ball 2 is about racing to the finish line as quickly as possible while trying to collect 10 bananas without falling. It would have been great to let your friends know that you're able to beat a stage with incredible speed by posting to a leaderboard, setting a bar for them to try and reach.

Some of the challenges late in the game are pretty tough; even getting to the Crown World (which requires you to collect 10 bananas in all 110 standard boards without falling) is an accomplishment. When I perfected every stage, I wanted to see an achievement unlock! I wanted to have my skill on public record! Instead, I found myself posting a message to my Facebook contacts after I jumped out of the game. Achievements and leaderboards can significantly improve the lifespan of a game -- they keep you going well past finishing the final board.

Currently, I believe that achievements and leaderboards are more interesting to core players than casual players, but that doesn't diminish their importance. If you've been playing Xbox 360 or PS3, and you just got a shiny new iPhone, you're going to want to see achievements pop up -- it's a key component to the enthusiast gaming experience. I believe that most major games will support leaderboards and achievements in one fashion or other by the end of 2010.

4. Splitting the content offering into two pieces.

When Super Monkey Ball 2 launched, the "minigames" button lead players to a landing page that allowed them to play Monkey Bowling. The page also promised that Monkey Target and Monkey Mini-golf would be coming soon. It was not made clear that the missing content would be coming free of charge, and no expected release dates were provided for the expanded content.

We answered these questions in interviews (yes, the content would be free, and the new content would hit in the first quarter of 2010), but most players didn't get the message. As a result, there is an increased risk that many people have finished the main game, and have dumped the App from their device... not knowing that substantial free content is on its way!

The game's release date had been pulled up to ensure that we hit the holiday season. The original strategy was to launch in February, alongside the new Wii Super Monkey Ball game. The shift was good, in that it was a demonstration of Sega's confidence in the quality of the product, but it was also difficult, because it put us in a position of not having all of the intended game content ready for launch.

The initial $9.99 asking price last November would have seemed more justified if all of the content had been there from day one. We were very happy with the value proposition the launch package offered, and we believed that it was better for consumers to enjoy a great game right away, rather than making them wait for all three mini-games to be finished.

But with all three mini-games in place, SMB2 feels more distinct from SMB1, and that differentiator would have been very valuable to have at launch. It'll be interesting to see if the delivery of that free content helps to boost the overall sales of the game.

5. Not effectively using SMB1 to cross-promote SMB2.

When SMB2 launched, there was a surprisingly robust uptick in sales of the original game. Looking back, I wish that we had done more to use SMB1 as a platform for informing consumers about the value of the new game. One example would be to offer an update to SMB1 featuring three sample boards from SMB2 (effectively, a "lite" version of SMB2 made available only to SMB1 owners).

Conclusion

We're proud of the work we've done on Super Monkey Ball 2, and we feel that it's the definitive version of the brand on the iPhone platform. A lot was learned from the development experience, and we continue to work towards improving and deepening the game's player experience through updates.

The App Store has evolved significantly since the original Super Monkey Ball was released, but even in light of the sheer volume of content available to consumers, we've seen first-hand that there's still plenty of room for premium 3D games to thrive.

Data Box

Developer: Other Ocean
Publisher: Sega
Release Date: November 2009
Platforms: iPhone, iPod Touch
Number of Developers: 9
Length of Development: 9 Months
Development Tools: Proprietary