Cavemania Postmortem: Or, How We Went from Age of Empires to a Match-3 Game

Dave Pottinger is President and CEO of BonusXP

Cavemania, which we launched this month, is a tough mobile game to categorize. On the one hand, it has elements from Age of Empires. You train villagers, gather resources, and end up killing a lot of stuff. On the other hand, you are also playing a Match-3 game. So, it’s simultaneously similar to PuzzleQuest or Candy Crush. As it turns out, this “inability” to cleanly describe Cavemania was a great indication that we had found some unique gameplay.

Of course, we never actually meant to make a Match-3 game with Cavemania. It started life as something completely different. I mean, who would intentionally try to make a Match-3 game these days? So, as we expand the worldwide release of Cavemania, we thought it would be useful to look back on the game's evolution and share a few Cavemania interviews from a recent game conference.
 

BonusXP Who?

Like many independent developers, we have quite a bit of old school “AAA” gaming history. In fact, the Age of Empires inspiration for Cavemania is no accident. Many of us started our gaming careers at Ensemble Studios, the birthplace for Age of Empires. It’s safe to say that strategy is in our blood. Cavemania is BonusXP’s second project.

Choosing a New Game

After we released Monster Crew last year , we had a team meeting to figure out what we were going to do next. We only had four full-time people in the studio, so we were interested in a smaller game. Ideas we discussed:

Everyone was really into the RPG idea, so we started there. We generously put one whole, entire person on the project. We continued to use Unity. We pulled in Matt Rix’s excellent Futile plugin to help with 2D rendering. We set about building our own UI system to address one of the bigger pain points with Unity development. We think this is our 89^th UI system. Give or take.

The RPG got prototyped in a week with some cutting-edge programmer art. It quickly morphed into a crafting-based RPG within a few days. The game’s codename, Combo, even comes from that crafting focus. The week after, some Zelda-elements crept in. We were solving puzzles and doing mini-dungeon raids.

But, after those three weeks, we begrudgingly admitted that we just didn’t have the staff to make the RPG the way we wanted. Even with a minimalist art style, we just wanted more content than we could reasonably do. So, we went back to the drawing board. We had the start of a nifty tile-based engine. What could we do with it?

Match-3? Are you kidding?

To this day, no one remembers (or is willing to take credit for?) the suggestion of making a Match-3 game with Combo. Who in their right mind would suggest that anyway? Can you think of a more crowded, bloodthirsty genre to enter?

Nevertheless, it was an easy game to prototype. We batted around some ideas for making it unique. We had all this code lying around that dealt with combat and moving guys around. Why not use that? The next day, we had characters on the board fighting in and around the Match-3 items.

We made seven different prototypes of this “Match-3 plus Units” idea over the next five weeks. We ended up with a prototype that we really liked. We had bumbled into something that was genuinely fun and felt like nothing we’d ever played.

By this point, we accepted that the content load was going to be more than we wanted. We had units, which meant animations. If we were going to do all that, we might as well have customizable characters and vanity items, too. We explored different art styles with two goals: Find a fun look for the game that our small team could execute. We quickly hit upon a Seussian/Papercraft-inspired style and went one step further by outlawing arms and legs. Animations suddenly got a lot easier!

GDC 2013 was now five or six weeks away. We set about polishing up the content and gameplay so that we could show it off there. While we’ve spent a lot of time continuing to iterate and improve the game, the core gameplay has remained remarkably true to the original prototype.

Cavemania Developer Scorecard: Wins and Losses

As Cavemania goes to market, here's a tally of the things that went well during development, and some takeaways for next time:

Win #1: Unified Senior Team

Dallas, where BonusXP is based, has a surprising amount of game talent. We were fortunate enough to build an incredibly strong team of people we could instantly trust. We’re up to 15 people now. Except for our last hire, we’ve known and worked with everyone else for many years. Those senior people tend to be capable of wearing multiple hats, which is fantastic. We don’t have room for someone who just does one thing. Most folks regularly work in multiple disciplines on a daily basis.

Those existing relationships also allowed us to bring together a team that wants to work in an extremely collaborative manner. We discuss everything as a team. Everyone is involved in decisions from hiring to the games we make. We don’t mandate crunch. We sync as a team twice a week and otherwise coordinate over email, chat, or our online task board. Everyone gets design credit. We don’t have leads. Everyone works from home on Fridays.

As a small, self-funded startup, we’re not flush with perks. We try to enable folks to work without distraction. We pay a good wage and provide top of the line equipment. We’ve recently secured corporate healthcare and are building out a new office. But, for the most part, people work at BonusXP because they love the environment. In our experience, that passion indelibly shines through in the final product.

Win #2: Iterative Development

A fundamental piece of our process is the daily build. The game must always run. We get nervous and twitchy when we can’t make builds. We expect forward movement each day, so the build becomes this unifying bellwether for progress. We thought it would be hard to maintain the daily build pace when everyone was stuck in some grungy weeklong task. That honestly never came up. Someone always had something new to show off each day.

Those daily builds go out to a core group of testers (employees, sig ots, and some lucky friends we’ve duped). We were also graced with plenty of additional friends and industry colleagues that participated in our extended alpha test. With such a small studio, we don’t have a dedicated internal playtest team. This assistance was invaluable.

While we actively provoke discussions about the game, we also try to limit discussion as soon as we have something to “try next”. Once we have a path forward, we let the owner execute the next step and go from there. This avoids over-discussion. We’ve ended up with a happier, more engaged team. As long as we’re moving forward, we’re fine. Like a shark, but less toothy.

We encourage breadth-wise development. We’d rather see a shallow version of a complete feature before we invest a ton of time. Yes, this can sometimes take longer. But, that’s okay because we’re more unified along the way. We’ll always make that tradeoff. We pair that indulgence with a ruthless focus on re-scoping. If we can avoid doing something we don’t really need, we can redirect that time onto polishing something we already have.

Lastly, iteration doesn’t stop with launch. Everyone can rattle off the evils of analytics, but any tool can be abused. If employed for the Forces of Good™, analytics can be wicked useful. We put in a lot of detailed stat tracking before our Canadian beta. Afterwards, we went through a week of rapid iteration on our initial levels. We looked at where players were quitting and which level objectives weren’t completed. We fixed how we taught special powers and polished a ton of other pieces. As this simple graph shows, the difference in retention is remarkable.

Win #3: Great Partners

It’s impossible to go it alone in this industry. Well, not impossible per se, just really hard. We decided to partner with a publisher to create something bigger than we could do ourselves.

Cavemania is our baby, so we spent a lot of time searching for the right publisher. We wanted to make sure we partnered with someone who loved the game and understood our philosophy. Yodo1 has been a fantastic fit for us. Secretly, I think they like the game even more than we do. They’ve been aligned with us every step of the way on prioritizing what matters and what doesn’t. Like us, they’ve always put quality before monetization and even helped us with some of the code.

We also have to give a huge shout-out to the vendors and contractors that helped with Cavemania. Gl33k provided all the sound effects and music for the game with their usual aplomb. ElementX Creative rocked our trailers. The incredible team at LocLabs saved our butts with ridiculously fast localization turnarounds. Additionally, at any given time, we had five or six different independent contractors filling gaps in animation, icons, and even programming.

We also made the decision to grow to two teams during Cavemania’s development. That has always been our plan; we just accelerated it. With all of our eggs in the Cavemania basket, we would have wanted to ship it sooner, perhaps before it was ready. By having a second team that’s in production on a second game (with a different, equally awesome partner), we’ve been able to let Cavemania bake until it is ready.

Alas, not everything was perfect…

Takeaway #1: Mixing Casual and Hardcore isn’t Easy

This one is tough to admit, but if we’re being honest, it’s probably true. Once we decided to go with the “Units plus Match-3” angle, we hoped to make a simpler, more casual game. Unfortunately, the very thing that made the game fun and different from other Match-3 games is somewhat at odds with a more casual audience.

Cavemania has a lot of combat. We enjoy that, but it’s possibly a turn-off for some folks. We tried several variants without combat, but they weren’t nearly as fun. Similarly, it’s hard to have a deep, upgrade-focused strategy game without showing numbers somewhere, so we did that.

In the end, we made the choice to stick with what was fun for us. I think we’ve done a good job blending that deeper gameplay with a straightforward freemium business model. I’m sure the players will tell us if we’ve made the right choice.

Takeaway #2: Start Localization Sooner

It’s a rare game that starts localization perfectly on time. We are certainly not one of those! We knew we were waiting too long to dig into localization. But, with a small team, tradeoffs have to be made. We had to do a ridiculous scramble to get all languages in place for the worldwide launch. We definitely would not have succeeded without the seemingly infinite help from LocLabs and Yodo1.

Takeaway #3: Biz Dev can be a Drain

We are a small studio. We don’t have dedicated staff to handle business development, legal contracts, HR, or admin duties. Hell, we even have to bring in our own snacks.

Jokes aside, we make no bones about putting things in the right order. We are building a long-term studio where the employees come first. When there is a conflict, the business concerns win out vs. adding another feature to the game. That’s the right cosmic order, but it does make the games take longer to get done. We easily “lost” an entire month on Cavemania due to those studio overhead responsibilities.

On the plus side, we did hire a producer halfway through Cavemania. He also handles all of our monetization and hardcore balance work. The importance of having someone who loves numbers (but can balance them with good design skills) cannot be ignored in today’s industry.

What’s Next for BonusXP?

We are planning a lot of additional content and features for Cavemania. New levels, new units, and even some fun multiplayer surprises. We will have to steer clear of the dreaded “content treadmill” pitfalls, but we think we have some good answers to those problems.

Beyond Cavemania, we have our 3^rd game in production now. It’s a larger mobile title that will launch next year. It might just have a little strategy in it, too. We’re also having some meetings next week to talk about our 4^th project. Maybe we’re finally ready for that Space Diablo game!