When I designed geoDefense I spent considerable time fussing over about five different tower types. Really four, with the vortex tower being sort of an enhancer thing. The spreadsheet work alone was pretty intense. Getting the values just right so that they made a perfectly balanced mix, while also trying to account for the things that were not numerically quantifiable, consumed weeks of effort.
So when I laid out the vision for Tiny Heroes and I said that ultimately I want as many as fifty defenses, I knew the particular pickle I was proposing. Tiny Heroes will hit its mark and include over 50 towers by the time the first expansion (The Workshop) ships. The initial release alone has something like 40 (it's an imprecise count because some towers have secret alternatives, which count as new defenses you have to suss out). Moreover, I was extremely keen on the notion that each of these defenses had to be valuable and interesting in their own right.
In the defense game genre, there are definitely two sorts: The kind where you have a relatively limited set of towers, but you can upgrade them in one or more dimensions. And then the type where there are tons of different types but perhaps little or no upgrade path (more like the very well done Plants vs. Zombies). With geoDefense I went with the former and with Tiny Heroes I opted for the latter. Although nothing's precisely as clean as that, since, as I have said... there are ways some defenses in Tiny Heroes can be upgraded.
The design team, then, was tasked with making sure each of these made sense and was balanced relative to every other defense. Sure there is a progression of power in the defenses, but even the very first one you are introduced to (the Ballista) will still be a go-to defense on your last level. However, since the number of defenses you can take with you each level is limited, eventually you will have to choose. We wanted to make it so that choice was difficult because so many are so useful... and cool.
Add to that the concept of combinations. Defense games tend to have combinations of towers that work effectively together just by their nature, and some games make their own game mechanic about combos (such as the wonderful Onslaught). Tiny Heroes is sort of in the middle. A lot of defenses work together naturally, and different people find different ways to arrange things. We gave careful thought to how this would play out and tested extensively. But more than that, we created stuff that was purposely designed to be extra powerful if you could figure out how they enhanced each other - or in military lingo: force multipliers.
We collect a detailed list of defenses the testers really like as well as those they found to have no real use. The fact that anything would show up on the second list may sound like a bad thing until you realize that one tester's list has little in common with the next. At least that's the point at which we now have it. In other words, there are a lot of combinations that work really well. And when someone thinks a defense is "useless," what it really means is they never took the time to figure it out, because it turns up on other tester's favorites list. And that's okay because it means there's a lot of ways to play effectively and that's what we were aiming for.
I have to say that making this all happen is not something I could have ever done alone. Our team over here allowed me to set out the big ideas, the flavor of things, and they ran with it. Taking the ideas and molding them into something that worked (or didn't as the case may be), and doing the hard grunt work of balancing this wide disparity of effects into a cohesive whole. Believe me, spreadsheets were of only marginal use in this game design because it's impossible to numerically quantify the relative merits of a Jelly Cube blobbing down a dungeon corridor vs. a cauldron of Sticky Tar vs. a treasure-returning Pack Rat.
In the end we fell back to the technique I used to "get it right" in geoDefense and geoDefense Swarm: Endless iterations of tweaking and testing - but this time with a lot of testers, careful data gathering, and analysis.
The net result is a game with some really sharp game balance and first-class defense design.