When I revealed our cancelled project Snowball Earth from 2008 in my blogpost earlier this week, I said that one of the problems that publishers had with the game and that kept them from wanting to fund it, was that the main characters were robots. Apparently it is difficult to reach a large audience with robots, and the fantastic Wall-E is a rare exception to that rule.
Although we had a clear vision for the game ourselves, we did not ignore this critique, and our art team (Martijn Thieme, Ralph Rademakers, Olivier Thijssen and Gijs Hermans) started making new concept art for the main characters. This was made after the main prototype was already finished, and was intended purely to show to publishers how the game could change to fit their marketing department's requirements better.
Especially the dogs came out pretty hilarious. They were replaced with real dogs, but those still needed to be able to roll around. Our artists came up with a (for the dogs) quite dizzying solution...
We didn't think it was necessary to make both the dogs and the player non-robots, so a number of alternative robot-dog designs were also made, to see if our art team could come up with something else that might appeal to publishers better. I personally especially like the bulldogs, which add a lot of cool and I imagine could have made for a nice upgrade later in the game. :)
Speaking of robot dogs, this is the concept art of the original Robodogs, which are playable in the actual prototype:
Now to the main character. A lot of redesigns were made. The common element here is that they all needed to have the magnet for pulling the dogs back, and some kind of heating device for melting the frozen world. May I especially draw you attention to the one with the teapot on top of his huge hat? The Ronimo art team seems to have a certain tendency towards that kind of awesome craziness... :)
What I find particularly interesting here is the lesson that our lead artist Gijs mentioned a while after drawing these. They all come in groups and they share a lot of common elements, like the suit or the shape of the head. Many are variations to the same ideas. This might be a fast way of working, but Gijs concluded afterwards that in a sense it also cripples creativity: working from a previous design limits the imagination to things close to that previous design. The better way of doing concept art, is to make most designs really different, and just start from scratch every time. This results in a much more diverse set of designs and ideas. It also doesn't even necessarily take more time: not all designs need to be shaded as detailed as these.
The final image I would like to share today, I was particularly delighted to come across in our archive. Publishers were unsure about how 'cool' the project was, exactly. Was this a kids game, or was this for teens and older? We felt it was a pretty 'cool' game and not particularly for small kids, so one of our artists made a number of alternative skins for the main character, which could be unlocked. Zombie Robot! Hot Rod Robot! Bumlebee even! Yeah! I really like the attitude in the skins and poses in this one!
At the end of this concepting period, we chose the two below as the new characters. This was just before we cancelled the game, so they never made it into the prototype. The bulldogs because they were cooler ('more MTV'). The boy because we wanted to get rid of the giant head of our original robot hero. The big head was a problem with a relatively high camera, because the body was hardly visible beneath it.
Which makes me curious: which do you like better? Our original two robots, or these two?
So these shots were made after pitching to publishers, which leads me to next week's topic: how does pitching to publishers actually work? Before we did our first pitches, I was incredibly curious to know, so I hope that will make for an interesting blogpost. See you next week! ^_^
Today we release the complete playable prototype of our cancelled game Snowball Earth! This is the game that we worked on at Ronimo for a full year before we made Swords & Soldiers. Since Snowball Earth was never finished, we never really talked about it either and kept the prototype hidden. Such a waste! The prototype contains over an hour of gameplay and some really cool ideas, so it is awesome to finally show it to the world!
Snowball Earth is a really interesting game, both in how it failed and in the things it did well, so it provides tons of material for future blogposts. Today: what it is, and why it was cancelled!
Note that because this is an early prototype, this is all pretty unpolished! No tutorial, ridiculously long loading times, quite a few bugs and it only plays well with an Xbox 360 controller. However, the concepts of the game are really interesting and work well, so give it a try!
The trailer that we made early 2008 to pitch the game to publishers. Showing the player in the trailer was all the hype in Wii commercials at the time, so we tried a bit of that here.
Snowball Earth was our first official game as Ronimo Games. De Blob was not entirely made with the same team, and for Snowball Earth we had decided to really start our own company and make a commercial game. We were still students at the time, and this was our group-graduation-project. After we graduated, we kept working on it, but mainly on extra stuff, like the puzzles prototype that is also in the Torrent. The main game that we are releasing today was basically our graduation project at the Utrecht School of the Arts, and it was also our official start as a company.
Snowball Earth was to be a console game for the Wii or the Xbox 360. A 3D adventure in the style of Ratchet & Clank, although for our first game we wanted to keep things a bit simpler and more doable than the fantastic Ratchet & Clank Future: Tools Of Destruction, which launched around that time, and was a big inspiration for us.
After the colouring gameplay of De Blob, we wanted to find some new innovative mechanics, and went all out on that. So much so, that we ended up with not one, but two really interesting mechanics. These two mechanics are not just what makes Snowball Earth shine, but also its downfall...
The first and most striking mechanic is the melting. The entire world is frozen, and the player melts the snow by simply moving through it. This is a pretty spectecular effect: snow turns into puddles and then into grass, plants pop up everywhere and the ice disappears from the rocks. It is quite a sight to behold, and simply walking around while transforming the world around you is already a lot of fun.
When melting is enabled, everything around the player comes to life. I consider this completely dynamic system one of the most impressive things we have made so far. It was also an enormous amount of handwork to create, though...
The sound effects for Snowball Earth were made by Aline Bruijns, and the music by Erik van Woudenberg. They graduated in audio in the same year as we graduated in game development, and they did some great sound work on Snowball Earth! Aline later started the sound studio Audio Rally Sound Design.
By now several other games have done kind of similar mechanics, like Prince of Persia 2008's restoration of life after a boss has been killed. The effect there is a lot less dynamic, but it was still kind off odd to see a game release with a similar idea after we had already cancelled our own take on it. (By the way, I absolutely love that Prince of Persia game, definitely my favourite in the series! I think it's a very rare open-world platforming gem!)
The other unique mechanic of Snowball Earth are the Robodogs. These are little robots that run along with the player. When you shoot them with the right stick, they become physics balls and can be used as projectiles in combat, or to get them to spots where the player cannot go. The player can press the right trigger to pull them back. This allowed for some great combat and puzzles. For example, hitting enemies in the back does extra damage, so deliberately shooting your Robodogs past an enemy and then pulling them to hit him in the back is a great tactic, and a lot of fun to do.
The physics gameplay with the Robodogs turning into balls.
In terms of story, the premise was as weird as one would expect of a Ronimo game. In prehistoric times, a little alien robot on an orbiting spaceship accidentally pressed the wrong button and thus froze the entire earth! As punishment, he is sent down to fix his mistake. He goes looking for sources of heat, like volcanoes, to melt the snow. However, during his quest he finds strange fridges everywhere, and discovers that he has been deceived: a fridge manufacturing company has illegally sold all these fridges to the yetis, and this is what really froze the earth! On his epic quest, our little hero must battle the evil polar animals, close all the fridges, and prove his innocence!
With the two unique mechanics and all the hard work that went into Snowball Earth, why was the game cancelled? The answer is quite simple: no publisher was willing to fund the game, and it was too big to make ourselves. We pitched to all the big publishers: dozens of meetings at Gamescom 2007 and GDC 2008. Quite a few liked the game and kept communications with us going, and some even invited us to do internal pitches to the entire team in their offices, but in the end no one wanted to really go with this.
I think there were two main reasons why publishers didn't want to fund us. The most important was properly that we had no track record. Making a big 3D game with a team that has no console experience whatsoever is an extremely risky endeavour, and no publisher would want to fund that until we had proven that we could actually finish a game on console.
At the time I was totally convinced that we could make this game, but right now I am very happy that no one funded us. We wanted to make the game in only 1.5 years, and wanted to grow the team from 7 to 15 developers for that. Just finding 8 experienced developers is already a difficult challenge... Let alone finishing such a big game with such an inexperienced team!
Swords & Soldiers and Awesomenauts were both much more realistic plans than Snowball Earth, and they both took way longer than we expected. If we had actually started making Snowball Earth, it would probably have taken us at least 3 years to make it. In fact, with our current team (13 full-time developers plus a number of interns) and my current experience (two multi-platform console titles), we still wouldn't be able to make this in 1.5 years. Snowball Earth was just way too big for our team size and total lack of experience at the time...
The other reason publishers didn't want to fund the game, is that it was a difficult sell to their marketing departments. What similar games were out there, and had been a success? At that point 3D platforming had not been a big market for years (except for Mario Galaxy, of course), and our game also had all these original mechanics that could not be easily compared to other marketing campaigns.
Of course, I know now that making weird stuff is a good way to gain attention and thus great marketing for indie studios like us, so I don't necessarily agree with these marketing arguments. But there is another, deeper problem within Snowball Earth: the mechanics around the Robodogs were all pretty complex, and thus for a slightly more mature audience. The melting mechanics, however, were very kids-friendly, but didn't bring us any gameplay depth besides the simple fun of walking around melting everything. So our two innovative gameplay concepts just didn't mingle all that well. Now add that our main characters were robots, and that robots just don't work for mass audiences. Except in Wall-e, of course, but that's not enough to go by.
We tried addressing these issues by focussing more on the physics gameplay. So after we had graduated and had finished the main prototype, our game designers made a whole bunch of puzzle levels. They bundled those in a prototype level with hardly any graphics, but some really interesting gameplay. This level is much less polished, and the physics are not precise enough for such puzzles, but it does show the great potential of these mechanics. With more time, I think we could have turned that into a great 3rd person physics puzzler. These puzzles are pretty difficult with all their bugs and the lack of a tutorial, but I still think the ideas are great, so they have been included as a second level in the version of the game that we are releasing today.
Some gameplay from the puzzles level of Snowball Earth.
To react to the publisher critique on the robots, we also made tons of designs for non-robot characters. We have some absolutely ridiculous concept art of dogs in round spacesuits, so I will post those soon as well!
After a year of work on the game and no money and no publisher, we finally got access to Wii devkits. This allowed us to start developing Snowball Earth on console, so that we could finally prove to publishers that we could actually make this game. However, what if publishers still didn't want the game at that point? Then we would have wasted even more time on what already seemed like a failed project! So we decided to cancel Snowball Earth altogether, and start working on something smaller, something that we could actually finish ourselves, on those Wii devkits that we got. The became a little game called Swords & Soldiers...
(Of course, the history from then on was paved with a lot more success: Swords & Soldiers came out on Wii, PS3, PC, Mac, Linux, iOS and Android, and got great reviews and won several awards. Awesomenauts was recently released on PS3, 360 and PC, again to great reviews. Seeing Snowball Earth fail because it was too big, was exactly the lesson we needed to learn to keep it smaller and more doable in our next games!)
