[In this extremely in-depth postmortem, Mediatonic's head of games Paul Croft and producer Jim Griffiths look back on the development of the PlayStation Mini/Xbox Indie (and more) title Monsters (Probably) Stole My Princess, which performed well despite some major snags.]

Monsters (Probably) Stole My Princess [YouTube trailer] was our first project for PSP and, thanks to the PlayStation Minis initiative, also our first for PlayStation 3 (in a roundabout way). Our background is in Flash game development for corporate clients and although we'd previously dabbled in some iPhone game projects under our own Mediatonic brand, Monsters was our first real stab at entering the console mainstream.

Once the project had been finished and launched via PSN in April 2010, we then created an Xbox 360 version using XNA, which launched in August 2010.

Concepts and Planning

Working within the constraints of the Minis platform, we knew we had to create a tight, small-footprint game that would be easy to learn, yet difficult to truly master.

We drew on our Flash sensibilities to devise an immediately accessible title that we could wrap in our own distinctive art style and place in a comedic universe, where our characterization and visual design would (hopefully) provide entertainment alongside the raw game mechanics -- and all in under 100 megabytes.

The concept design team was just two people -- myself (head of games, Paul Croft) and writer Jim Griffiths, who was also producer. We initially started out with the vertical platforming mechanic and went through a few themes before we hit on the core idea of Monsters (Probably) Stole My Princess.

Our original idea was for a platform puzzler where the player was a monkey with a bell, and the player could use the bell as a grappling hook to defeat enemies. "The Waterfall of the Dirty Gods" was our original concept pitch, where the player needed to climb to the top of several different waterfalls and ring their bell to stop the gods from washing their filthy bodies in the water and polluting the waterfall. However, we did not feel the theme was quite right for the game.

The concept then coalesced into a guy chasing after monsters that he thinks have stolen his princess, which adds a unique irony to what could have been a relatively dry and abstract game design, while keeping the platform race concept intact. Critical to this idea was the player character, the Duke.

For us, building a good game is all about creating a really compelling story and characters along with a great interactive design, so once we'd sketched out the Duke, the rest of the concept had a reference point to develop from. What's more, framing the game's central story as a demon hunting down the presumed kidnappers of his beloved princess was an excellent excuse for our artist to draw giant, awesome monsters.

The Duke

The Duke was well formed as a character before we'd even discussed his look with our art team. We were looking to develop a player avatar that embodied the 'let's go!' attitude of the game mechanics. As such, The Duke embodies the entire game concept of dynamic action and movement, as well as doing things just for the joy of doing them, regardless of the consequences. As the game was our debut title for Minis, we wanted people to be able to jump in and get going immediately, so it fits that our player character should somehow match that notion.

In deeper sense, that immediacy reflected our idea of what players would actually want to do -- jump around and bash monsters -- which is the exact attitude that the Duke has within the game's story. His search for the princess is as much about having an excuse to go on a gleeful (non-lethal) rampage as it is a noble quest to rescue a damsel in distress.

We could even go as far as saying this concept reflected the core values of the Minis platform as well as our expectation of what players actually want. However, that's the benefit of hindsight at work. At the time, it just seemed like the right way to go about making a good game!

The 'Bad' Guys

Naturally, any good player character needs "good" enemies to create the basic tension in a game. With our main concept developed, we needed to come up with a good hook for the player's opponents, so we designed the titular monsters with the Duke's colossal misconception at the forefront of our minds.

We thought that the most natural perception for players was to take the Duke's viewpoint, so we wanted to play with this as a mechanism for humor, which we worked very hard on getting right. We couldn't just spell out this misconception to the player, which, in any case, the Duke was only using as an excuse to have some fun, so we set about staging the 'reveal' in several ways to let the player "discover" what's really going on.

The most obvious method for this was that on first glance, all of the monsters are large and frightening and appear to be engaged in some kind of sinister activity (wrapping something in a web, standing poised with a knife about to plunge down) but when the player watches a little longer, the monsters are actually engaged in quite normal activities (wrapping a present for a birthday party, about to cut a cake for a teddy-bear's picnic) and peacefully going about their business.

This central idea came from the earliest sketches of our first prototype monster, Squishface Octo-thing, who is really an avid gardener rather than a merciless kidnapper of princesses. The rest of the monster cast followed from that first concept.

The second method we used for the reveal was to place visual giveaways at the top of each level that revealed the monsters' true nature and that they had nothing whatsoever to do with the disappearance of the Duke's princess. For example, Squishface had a beautiful little garden at the top of the tower, where he grew flowers. The spider creature wrapping a present in his web gets to the top and there is a little skull-spider party with all the young spiderlings wearing party hats (and hiding from the Duke -- awwww.)

Of course, the Duke actively disregards this evidence even when it's staring him right in the face. That could have made him seem TOO aggressive or vicious, so we tried to handle that carefully and make it so that everything he did was without actual malice. He was only ever interested in the chase and the drama, and once he'd knocked the monster around a bit, he'd happily wander off in search of the next adventure.

All of this may seem quite convoluted in hindsight, but it all tied in to what we wanted to achieve: a fast-paced game with a fast-paced story that let the player think about it, enjoy it and try to work it out if they wanted to, but that players could just ignore it if they didn't want to pay attention, which would lead to a little surprise as a payoff at the end of each level.

Humor and Narrative

Other aspects we needed to get right early on were the characterization and narrative details we couldn't convey through character graphics and traditional narrative exposition mechanics. With the tiny file footprint, we knew we couldn't rely on voice or movies, so did a lot of work in getting these two perfected with the high-concept approach we had in mind for the game's overall story, as well as trying to make sure the grand joke of Monsters (Probably) Stole My Princess was maintained (and was consistently funny).

Humor is very, very difficult to get right. Sometimes you can have an idea for something, and you'll say it in one line and people will get it and the image it conjures up is funny and they'll laugh. But will that one little laugh translate into something that will work for an entire level? Other times it's the opposite, and you'll explain something and everyone will look skeptical but you have to stick with it because you can see how it will fit in.

Ultimately, we had to be viciously hyper-critical of anything that we were writing or designing. We actually put ourselves in the mind of a player coming to the game for the first time, while being forced to play it at gunpoint, who's just looking for the first opportunity to loathe everything that you show them. If you can create something that you feel will stand up to that kind of attitude, then you're (maybe) going to be ok!

Humor is always a hard thing to get right, but as some of the team had already worked on games like Amateur Surgeon for Adult Swim, we did have some experience to build on and a working knowledge of how to structure humour in gaming context. This helped give us some confidence that we would be able to deploy some effective and entertaining humor in Monsters.

We believe that a funny moment in a game is everything coming together in a perfect storm; it's the situation and the text from the writer, it's the perfect sound effect from the audio team, it's the expression on the character's face from the artist, and it's the flair on the movement from the animator. If you nail it, we believe people will remember that moment and remember your game. If you mess it up, then you really have nothing.

As a studio and a team of game creators, we'd say that making games that are a little bit edgy, strange and that we find funny is what we really love to do, which is why we put a lot of effort into making our humour work!

Prototyping and Development

With the creative work well underway, we prototyped the raw game mechanics in Flash with one of our in-house coders, which gave us a really rapid turnaround time for iterating the design in those critical early stages, as well as a prototype that we could test on our personal office PCs and laptops.

We took iteration to the point where we had a build that consisted of a little gif image chasing after a large blue block, that allowed us to make really quick and simple tweaks. Can you double jump? Triple jump? Wall jump? How high roughly should you jump? How many platforms should there be? Does this show any hints of being fun? "Let's find out" is obviously the easy answer!

We developed the combo mechanic shortly after the prototyping stage. We could tell that the core concept worked and was pretty fun to play, but there was still a need for something extra in the interactive design, and introducing combo scoring for consistently 'good' play was a logical step to take. Once we'd added the platform combo system, the game immediately felt like it had taken on an extra dimension.

Without a doubt, we'd say Flash is a fantastic hot-house/laboratory prototyping tool for any small-footprint game project, irrespective of the final target platform. We could easily trial visual assets and animations alongside the game mechanics and develop the final design and look with the minimum of fuss. This took place alongside the PSP development process, where we rebuilt the prototype from scratch and then fleshed it out into the final game.

One additional (and really helpful) aspect of starting in Flash was that our PSP coder could immediately see what he needed to translate. We were also really lucky that he'd arrived at Mediatonic just after finishing another Minis project, which gave us invaluable insight into the peculiarities of the Minis platform as well as the PSP itself.

Content Creation and Playtesting

The core of Monsters (Probably) Stole My Princess was coming together on the PSP when we started the meat of the game -- the level design. Again, we used Flash for our level editor tool, for much the same reasons as the game prototype. We were able to pull together a very flexible level designer really quickly, using a lot of the ready-made components in the Flash IDE, and it saved us a huge amount of development time.

Iteration was as critical to this process as it was to the basic interactive design and we made sure to employ heavy playtesting throughout the process, particular. The medal targets for each level made for some really challenging difficulty levels, so we had to playtest these with lots of different players to determine the best boundaries for medal awards.

A key aspect near the end of the development cycle was bringing in friends, girlfriends and relatives to play through the entire game. We collected a load of metrics to assure ourselves that the balance was 'right' - average completion times, average number of attempts per level and so on. From the statistics we harvested, Monsters' Story mode has an average completion time of about one hour for a new player, which we think sits perfectly in the Minis 'sweet spot'.

However, competent players were able to complete the game considerably faster, which skewed our results a bit. Given that we were working to the notion that Minis would be appealing to skilled, educated gamers as well as total newcomers, we felt the overall balance wasn't unduly affected by that skewing.

From Minis to XNA to Wherever!

The success of Monsters (Probably) Stole My Princess on Minis meant we were in a great position to port the title to other platforms. Wanting to establish ourselves on the main home consoles, and doing so on a tight budget, led to us embarking on an Xbox 360 version by a slightly unconventional route for a professional game development studio -- the XNA/ Indie Games channel.

This also had the benefit of laying some groundwork for the future, especially with Windows Phone 7 about to launch, along with our plans to possibly expand Monsters onto the PC/Windows format and broaden our potential customer reach even further. A big benefit of doing a 360 version is that we could use HD assets and remain within the file footprint restriction, making it the "prettiest" version.

Ultimately, though, the prime reason for the 360 version was simple curiosity in adding another platform to our portfolio. A huge bonus in this regard was that our PSP coder was also well-versed in XNA, making the whole translation a great deal easier!

XNA was a great way to break into the 360 platform with minimal risk, though it comes with some unique issues, such as the community-based approvals procedure.

When you're used to platform holder approval times, waiting for enough community members to go through the approval chain can be frustrating, but overall (and as a way of testing the water), it's a great route to getting your foot in the door, so to speak.

What Went Right

1. We got the game finished on time

Despite working on a completely new format for us, we were able to go from scratch to (very close to) our desired launch date in around five months, when Monsters (Probably) Stole My Princess was one amongst several projects the studio was working on.

We were pretty proud to have got the game out "on time" as we also ran into a couple of problems during submission, which put some significant delays on launching. The years of working on short term flash titles meant that we had some good working processes in place that directly applied to this project, even though it was more ambitious than most of our previous titles.

2. We created a popular character and fleshed him out via story exposition within the file footprint constraints

We got awesome feedback from players about the game and the Duke -- people loved him and the game's overall humourous tone, as well as the gameplay itself. We're pretty proud to have achieved this in 100 megabytes and with words we'd written ourselves!

3. We got enough encouraging feedback and sales from the Minis version to port the game to other platforms

We got plenty of positive feedback very quickly after the initial launch, which spurred us on to do the 360 version via XNA. This was helped enormously by the making sure our website, Twitter accounts and Facebook pages were highly interconnected, giving us a lot of avenues for customers to give rapid feedback.

With the Monsters (Probably) Stole My Princess IP having enough impact to win praise, we're confident about porting it to a wider range of formats too -- with the Flash base, this will make it an ideal fit for smartphones and browser-based gaming platforms and we're now looking at doing an iOS version.

4. We got enough momentum to commit to another Minis title

The critical and platform holder reception was hugely encouraging, which gave us real momentum to start a second Minis title (Who's That Flying!?), where we'd apply the same development techniques and be lucky enough to enjoy similar acclaim. We were also really fortunate to have a lot of valuable support from Sony. The teams there helped us promote the title on the PSN store and have been really great to work with.

5. We got a lot of exposure and attention from the press

Monsters (Probably) Stole My Princess reviewed really well in the media (including a glowing review and editor's pick from IGN), which did wonders for our profile and status. Mediatonic was established as a 'go to' brand on the Minis platform and we got a lot of interest, both from gaming media and customers looking to hire us for their own projects.

It was a bit of a shock for us, having spent years doing Flash titles, which were quite "under the radar" for the specialist gaming media and then doing one PSP Minis title and getting a quite a lot more visibility and attention all at once. However, I guess you could say we needed all that Flash experience to really deliver a quality debut on a platform that's much more visible to that particular media sector.

What Went Wrong

1. We couldn't include more content

The main criticism we got for Monsters was concerning the length of the game -- this was mostly due to the file size limitations, but also the fact that there were not many other Minis titles in existence, we did not have much to base our game structure on in terms of what works best for that specific format.

We had a plan and we stuck to it -- make the tightest, best story mode we can and don't pad it out with filler. That meant that the story mode was (very) short and sweet, with more content in other modes.

We could have dropped the cutscenes and made the monsters smaller, or made other asset cuts to increase gameplay content -- but then we'd start to lose the things that really made the game shine.

The deeper reasons behind a perceived lack of content are actually expanded upon in other points to follow, which include the way we structured the game modes and the way we did animations in-game, but overall this issue can be put down to our inexperience with PSP development and the Minis format constraints and, perhaps, not catering well enough for the expectations of the PSP/PS3 userbase and relying too much on our experience from Flash and iOS gamers.

2. We put more content in Score Attack than the Story Mode

It's hard to see this as a thing that actually went wrong, so to speak, but it's something we paid a lot of attention to with our next Minis title. The criticism about the game's length kind of missed the fact that Score Attack had 24 stages, which is a decently-sized slice of gameplay for a 100mb game, and where we thought the longevity would lie.

Generally, we saw Story Mode as a way to tell a great story, set up the character and his world, and introduce the gameplay concepts to the player. Score attack mode was where the 'real' game began, so players would hopefully train their skills throughout Story mode and then go on to rinse the combo-based Score Attack.

Ultimately though, Monsters had two types of players. People who loved the combo mechanic and wanted more would go deep into Score Attack, but players who did not quite "get" the combo aspect would finish Story Mode very quickly and with that, the game was "done" for them.

We could have done more with the combo aspect to encourage that second type of player to get into the spirit of it a little more -- maybe by requiring a minimum combo amount to damage the level bosses. This would have made things more interesting and "enlightened" the more casual players to the innate satisfaction gained from mastering the combo system.

3. Platform holder procedures and updates affected our release windows

It's unavoidable really, but we ran into delays for launching on both PSN and Indie Games due to issues at the platform holder end. In PSN's case, Sony updated the PSP emulator, which actually contained a bug that led to us failing submission. We had to re-submit and this incurred a significant delay. With the 360's Indie Games, we had to approve via the XNA community, which was reliant on enough members going through the process. This took a lot longer than standard, internal approvals.

4. People didn't realize there was a secret sixth level

We'd hidden an entire stage in Monsters, which was unlocked if the player attained gold medals for the five main story levels. It had all-new artwork and a unique boss (Kong in a Thong) and, in terms to the story, took place in the Duke's dreams as a prequel to the rest of the plot.

Unfortunately, hardly anyone seemed to notice this. Considering that we took a lot of flak for our story mode being too short and that by missing the secret stage, players were missing out on an entire sixth of the story, we'd consider this a thing that definitely went wrong.

Avoiding overt hints that the level existed was a deliberate decision, mostly because we wanted people who unlocked it to be completely surprised and bowled over by their discovery. However, we failed to properly indicate that the level even existed, so any unlocks that did happen were by serious players and ardent completists discovering the sixth stage accidentally.

5. We were inefficient with some assets

To save time in the development process, we used pre-rendered PMF movies for the Duke's finishing moves instead of writing an engine to render them dynamically in-game. These animations mounted up as the Duke has a number of different finishing moves that can be applied to any monster, so they ended up using most of our space!

In fact, 80 megabytes of the game's footprint is taken up by these animations. We wanted to show off the talents of our skilled animation team at Mediatonic, and these movies were one of the main ways in which we wanted to communicate the story and the character of the Duke, so we felt it was a good spend of resources at the time to get the visual qualities we were after. In retrospect, we could have put a LOT more game content into that 80 megabytes.

The fact that we used PMF files also caused us quite a few unexpected issues later on, as they kicked up a few problems on the PS3 emulator. When we were porting the game to XNA, we ran into more problems as XNA video playback is not great and we couldn't play the movies in full screen HD, which made things look a little stilted.

Conclusion

Monsters (Probably) Stole My Princess was a hit for Mediatonic. Aside the financial aspects, we got nearly universal praise for the game, either based on its gameplay merits or our comedic approach and strong sense of visual design. This really helped us establish ourselves on the growing UK boutique studio scene and promotes our creative freedom as being a core asset of the company.

We were really flattered by some of the plaudits, particularly the IGN editor's award and we were able to pull out some amazing quotes regarding the game and ourselves. IGN also gave us the 2010 award for "funniest game on PSP", which meant all the hard work we put into effortlessly humourous had paid off. VGChartz gave us an award for the best PSN game of 2010, which was a really big mark of recognition for us.

As a result, we now get a lot more attention from the gaming media and as mentioned above, Monsters (Probably) Stole My Princess has proven itself enough to warrant more ports to a wider range of formats. Having our own IP get this level of acclaim and visibility is fantastic, giving us a whole new world to explore.

We also really got to grips with developing for the home console digital distribution platforms, which is invaluable experience. With Monsters (Probably) Stole My Princess and Who's That Flying!? completed, we're starting preliminary work on our third Minis title and really hope to carve a significant brand presence on the format.

Data Box

Developer: Mediatonic
Publisher: Mediatonic / SCEE
Release Date: April 22nd 2010 (PS Minis), August 24th 2010 (X360 XNA/IG)
Platforms: PlayStation Minis PSP/PS3, Xbox 360 XNA/Indie Games
Number of Developers: 1 programmer, total team of 6
Length of Development: 5 months
Lines of Code: 50,000
Development Tools: Flash, Illustrator, After Effects, Native C+ coding
Babies born during the project: 1
Babies born during the project named "Duke" after the central character: 0