Friday, 29 June 2012

A bad kind of competition

A short blogpost this time, with a small but valuable lesson that we learned at Ronimo a couple of years ago.

At some point we figured it would be wise to make two artists make designs for an important character, and then from all the designs they make choose the best. At first glance this looks like a really good idea: two people will come up with a much wider range of designs and the chances of reaching something really good are a lot higher than with one artist.

However, there is a gigantic downside as well: in a sense making this design becomes a competition, and one of the two artists will lose. This can be incredibly demotivating, especially if an artists 'loses' a couple of times in a row.

So we decided to not do this again, and just give a single character to a single artist to design. That artist makes a lot of different designs, everyone can give feedback and in the end we choose from his concepts. We can still have variation and something to choose, but whatever is chosen, it is always a design by the artist working on it, so he never 'loses'.

For this same reason, we also rarely move a task to someone else when a specific artist appears stuck on it. It is often better to struggle a bit, or to let it rest for a week or so, or to get help, then to give the entire topic to someone else and see him find something cool within a day. This feels too much like someone else did a better job, and we don't want that kind of competition in our office. Of course, if someone asks for the task to be moved, then that is totally okay.

I guess some people might excel under competition and stress, but in general, I feel it is a good idea to try to keep any form of 'winning' and 'losing' outside our development process.

(Except when playing games, in which case everything else must give way for the glorious purpose of pwning n00bs.)

Sunday, 24 June 2012

How to really involve a larger team in the creative process

At Ronimo we strongly believe in involving the entire team as much as possible in the creative process of game development. No matter whether someone is a lead, a developer or an intern, everyone's opinion is valuable. This is, however, an incredibly difficult thing to do. In the past months we had 16 people working in our office. That means sixteen different opinions, and even just understanding each of them is already incredibly difficult, especially when it comes to difficult-to-explain graphical concepts. Let alone if you want those 16 people to agree, or to least all feel like their voice has been heard and seriously weighed in.

There are two big problems involved in this, and this blogpost outlines a method that tries to work around both:
-A decision might be made without really taking everyone's opinion into account.
-The process might take endlessly, as everyone has a different opinion and we keep trying to find a middle ground.

So today I would like to talk about the way we do that at Ronimo for our art styles. Since we started on De Blob 6.5 years ago (when we were still in school) we have been in a series of larger game projects and in each project we have learned more about how this works best for us. To keep this blogpost focussed, I am only going to talk about art style development today.

Before I start out, let me first note that this is a very flexible process for us. We don't follow a step-by-step plan and we constantly adapt to what works in the current situation. This blogpost is just my analysis of how this practically pans out for us, and since I think it works quite well, I figured other creative teams might have a use for this information.

Step 0: Communicate visually
This is not really a step in itself, but more a general concept. Describing art styles in words is incredibly difficult and in practice always ends in everyone having a different vision in their minds. So we try to communicate visually as much as possible. We don't do art style meetings or brainstorms without a computer with internet at hand, so that references can always be shown and pointed at. Clear communication is key and visual concepts should be communicated visually.

Step 1: Collect inspiration
We usually already have a rough gameplay concept before we start on the art style, and that means that most people also already have some ideas about how the game should look. So we start by making a topic on our internal forum where everyone just posts inspirational images. Images can come from any source (movies, comics, art, concept art, other games), and may contain just a couple of interesting elements, or be relevant in their entirety. Since posting other people's images does not require any drawing skills, even programmers can easily propose styles here.

Step 2: Create concept art
Our forum topic will quickly contain several dozen images and ideas. From those our artists create a series of completely different designs, trying to take in all the ideas. This should result in wildly different images. The art team tries to also express the ideas that they personally don't like. Everyone's views ought to be represented here somewhere.

(Check this previous blogpost for bigger versions of these images.)

Step 2b: Forum feedback
While the art team is making concept art, they regularly post their results on our internal forum, and everyone can add suggestions for variations and changes.

Step 3: The big feedback meeting
Now that we have around 10 vastly different designs, we do a big meeting with the entire team, around a big screen that shows the images. Since this is a meeting with the entire team, it can easily take several hours. First the art team shows each image and gives some background info on what their ideas were with it. Then we do a round where everyone in turn gets to express their opinions and elect a couple of favourites. Our producer writes down the votes and at the end we discuss the results.

Letting everyone speak in turn here is very important for this meeting: an open discussion with 16 people always has some people constantly talking, and some others who never say anything. Giving everyone a turn gives more equal weight to different personalities.

Step 4: Iterate on the favourites
Usually the meeting has not resulted in a single image that everyone is happy with. So the art team takes the two or three favourites and makes variations on these, taking into account any feedback given during the meeting or on the forum (e.g. "How about style B, but with more colour and with that tough guy from style D"). Again, results are posted and discussed on our internal forum.

Step 5: Make a final choice
Now we do another meeting like in Step 3, only this time with the goal of in the end choosing the final style. First we discuss a bit, then we do a vote where everyone can elect his 2 favourites. The style with the most votes ends up in the game.

(Check this blogpost for more on why this style was cancelled in the end.)

Step 6: Make game!
The process is done and we have a style and theme! Woohoo! The art team can now start working on designing all the different characters, elements and levels in the game, and on trying to get the style working in our engine.

A key thing to understand here, is that this process involves a limited number of meetings with the entire team, and each meeting has a clear goal. Since we usually cannot reach a unanimous decision, voting is used to make sure we never end up without a conclusion. These limitations guard us from iterating endlessly based on all the different (and shifting!) opinions.

With sixteen people in the team, the end result will rarely be a unanimous decision. However, by involving everyone in the team, and by making the art team make designs that incorporate everyone's opinions (even the ones the art team does not actually like themselves) we have a process that really gives everyone a voice and gives even the weirdest ideas a change to flourish.

Sunday, 17 June 2012

The Epic Shoes Of Awesome Winning

I often talk to people about that good indie marketing requires doing all kinds of crazy stuff to get press to talk about your games. The good thing is that such "crazy stuff" usually is the same as just "awesome stuff". So even if a 'stunt' would fail and isn't written about that much, we still end up with awesome swag. I intend to some day write a couple of blogposts about marketing, but in the meanwhile I just can't wait to show the awesomeness I have for you today. I would like to present: The Shoes.

Yes, we have Awesomenauts shoes. Only one pair, but they are just insanely awesome, and custom-made by Kyozokicks. And here's the best part: we are giving them away to someone who likes Awesomenauts on Facebook or who follows RonimoGames on Twitter. So if you are not doing one of these things yet, then please do so now and get a chance at winning The Shoes Of Cool!

These shoes are even cooler than the custom Swords & Soldiers Move controllers we made two years ago. I realize I am saying a lot of "cool" and "awesome" and such here, but I just really love this kind of marketing! ^_^

Yellow shoelaces! YELLOW SHOELACES!

Sunday, 3 June 2012

What Awesomenauts almost looked like

Last week I showed the various styles that our art team concepted for Awesomenauts. But I kept the most important one for this week: the one we actually chose to make! The Purple Style. We chose this one and worked on it for several months! This is what it looked like:

click for high resolution

I still consider this image to be one of the pretties things to come out of Ronimo so far. So just imagine how bad I felt about letting this one go. It was the right choice, though: this purple style is extremely difficult to read. It looks great, but in which teams are those characters? How could we visualise what skills a characters is using, when the colour palette is so limited and the backgrounds so dense?

This style finally got cancelled when, while playing around with the gameplay, it turned out that Awesomenauts was going to be very chaotic. This chaos was part of the fun, but a chaotic game requires a lot of clarity in the art, which is just not possible in the Purple Style. So we switched to something that was inspired by the likes of Earthworm Jim and Ratchet & Clank.

Since we worked on the Purple Style for several months, lots of cool concept art was made for it. Our ideas for the setting were not very clear, but some keywords were "Shangri La" and "eastern".

When we started working on Awesomenauts, we were quite unhappy with Flash and wanted something else. Flash was too difficult to handle for anything but flat shading, and we wanted more details and gradients in our artwork this time. So the idea at first was to create 3D models and render those to textures. This turned out to be too much work, though, so when our former intern Marlies Barends showed us After Effects and how excellent it is for animating detailed 2D art, we switched to that for all our character animation.

However, in the meanwhile I had already been playing around a bit with how to get those 3D models to look good for a 2D game. A couple of years before that I had made a 3D model of Captain August (a webcomic by Roderick Leeuwenhart), so I added some basic animation to that and created a couple of cartoon materials in 3D Studio MAX to see whether that would work. It doesn't look as 2D as I wanted, but it does look quite interesting, I think:

I also tried dropping that guy in the 2D environment art that we had:

Of course, we didn't just work on the characters, but also on the environments. Since this would play in a forest, our art team wanted to add variation by using the entire height of the forest. So the top of the level would play on top of the trees, showing mountains in the distance, while the bottom of the level would play on the ground, between roots and small streams.

Finally, while looking through our concept art archive, I came across this one, which I had never even seen before! I really like the flowing lines here, so I'll end this post with this great piece of linework:

Next week, I'll explain what process we use at Ronimo to involve all 16 developers in choosing a visual style for our games. It is difficult to involve so many without grinding to a halt in endless discussions and brainstorms with way too many people, but I think by now we have found a pretty good process for this!