Kemono Battlers Design Thoughts

So right of the bat, our main goal was to create a mobile game, that could be released before the release of TFT on mobile, and capture that craving audience. But we did not want to just copy that game, we wanted to combine elements from multiple games within the genre and also put our own twist on it.

The other game that heavily inspired the design was AFK Arena, it is from there that we took the general idea of 3v3 combat with very distinct character abilities. We chose this as our inspiration, because one of the main things you notice about good TFT players when compared to mediocre TFT players, is that the good player actively looks at their opponents teams before a fight. To give even novices this tactical experience and to ease them into the strategic elements of our game, we really wanted the enemies team to be prominently visible and inspectable before the fight actually starts. We also thought about including a time limit to force more intense decisions, but ultimately decided against it, as we felt it was antithetical to our idea of a more strategic gameplay.

Another thing that falls within that same vein of strategy is that we wanted our units to be very predictable. We did not want as much randomness as some of our competitors have, because we felt that it would be an interesting factor if our game more closely resembled chess. This was actually a phrase that we use regularly when we tried to explain the concept to people who were not as familiar with the genre overall. "It is like chess, but both of you only play with a selection of the pieces. You will want to spend money because having access to the entire pool of pieces is vastly more entertaining than not having that." To further that goal, we kept the target selection of our units deliberately simple. Most units will focus on the closest enemy, while some assassin units will explicitly ignore that rule and attack the backline instead. This gave us a lot of design space to then spice it up with units that for example defend their allies and take over the enemies aggro, or units that can force enemies to change their attack target. However, we wanted to keep it to the point where you could conceivably play through the scenario in your head and be more or less accurate with your prediction.

This was because we wanted to have a "skill matters" approach, since we felt like that would keep the free to play players engaged for longer. afterall there are few things more satisfying to a free to play player than using strategy and game knowledge to defeat an opponent with objectively stronger units. The reason for this focus on our part was the idea that in the long term, the wales of our game would need free to play players in order to have someone they could beat up, while said free to play players would need to be able to fight back in order for them not to abandon our game. The solution to this connundrum is what brought us to instanced pvp, meaning here that we would let players set up their static defense team, and then have other players actively fight against that team, basically creating a situation where no real pvp would take place, but rather an infinite amount of user created pve content. A pleasant side effect of this would also hopefully be very easy to read data on what units might need balance changes, since player skill would only be a factor on one side of the equation.

While we are on the topic of game balance, this was actually a fun first for me on this project. This was the first time that I made a large excel spreadsheet to really crunch the numbers on unit stats. First I split our units into 3 archetypes aligned in a rock, paper scissors style triangle: tanks > assassins > attackers > tanks. I then used a handful of goal seeks and designer choices to come up with average stat lines for these roles based on things such as average time to kill, and effective power compared to skill (the latter basically being the disparity between best and worst matchup). Armed with those stats I then made a point buy system which we could use to create a wide variety of units very quickly, while still having them be similar yet different within their archetype. I also came up with 5 abilities of varying point costs for each of the archetypes, just to give us a baseline from which to judge future abilities. This is also the point where I want to mention that we had a fourth archetype, the supports. But my attempts to fit them into my formula was not very long lived, as I judged them to be too inividual with their vastly game changing abilities, that I basically decided to let supports be a case by case decision.

I will admit that this formula for creating units was not all that great, but I blame that mostly on the very limited testing data I had at the time, and the severe lack of time to make it. It was however a lot of fun to crunch these numbers, and I hope that I will be able to do similar things again in the future !