what is an Auto Fighter ? - 8 minute read
Let's start off by discussing the mechanics of Auto Fighter and the intentions behind it. Auto Fighter was originally called "Auto Battle Trainer", with a rough prototype developed by my boss and design director, Tim Fowers. The original design intention was to create something reminiscent of Super Auto Pets but with the presentation/design flavors of a fighting game. My job as a designer was to help elevate this prototype into something presentable and engaging to pitch to publishers and show off at development showcases.
The big objective at the time was the upcoming developer booth/showcase the team was going to have at Evolution Championship Series 2023, one of the largest and most prestigious fighting game tournaments in the world. The biggest challenges were striking a balance between the two different genres we were pulling from, addressing gameplay pain points, and showcasing the game in front of a huge global audience.
mixing it up - defining a hybrid genre
A big challenge the design team dug deep into was what it meant for Auto Fighter to be an Auto-Battler/Fighting Game hybrid. Given the two genres are fairly distant from each other, a key design objective was establishing what elements of both genres we wanted to embody in its gameplay design.
After brainstorming, I originally developed these initial core design pillars for Auto Fighter's gameplay identity that the team built upon:
1. The shop is RNG, but gameplay is skill-based.
2. Optimal play is predictable, sub-optimal play is risky but random.
3. Positioning defeats raw numbers.
From a design perspective, I wanted Auto Fighter to embody the skill/knowledge gameplay loop of a fighting game but have some of the chaos, luck, and builds an auto battler has through its economy.
oh the pain - addressing pain points
One of my core responsibilities was to modify, rework, and develop designs to help elevate the general combat and system mechanics in the game to achieve a more enjoyable game loop. To understand the current state of the design, I ended up playing the game a lot in its original prototype state.
There were some core pain points of the Auto Fighter gameplay loop that I identified in its original version:
1. Actions were overly individual in resolutions, the original set of actions seldom had long-term consequences or rewards - in turn, heavily rewarding short-term gameplans.
2. With no powerful incentive, all turns are ambiguous and random since nobody knows what order or actions the player picked for that turn.
With these two core design pain points identified I got to work to address these issues.
Thankfully, both of those issues would be addressed by introducing "stacking" or "positional" effects. Something that ended up becoming a core "feel good" design element. During testing, we learned the gameplay of "builds" off of these stacking and positional type effects was where the core of the fun was. So we leaned into it!
A big original issue with the deck of actions the prototype had was that a lot of the effects were one-turn and non-conditional. This would mean we would have a lot of cards that only really dealt damage or gave money for the next shop. Meaning damage was the only flat way to reduce the amount of interactions needed to win the game.
The original design intention behind the simplistic action effects/resolutions was for players to leverage the "trainings" players can buy - think of them like items in MOBAs, they buff your a selected action card with a unique effect/stats.
Though something that did exist in the original design was an effect called "Shifting" and "Fade". Shifting was an effect that shifted your opponent or your own order-of-action (OOA) if you won. Fade made a card disappear after use for one turn. These two effects were very powerful ways to manipulate your OOA but required good foresight and a good setup to pull off.
The presence of Shifting and Fade was a fundamental "seed effect" that pointed us in the right direction. The ability to string sequences of actions together was a really fun element of the game that needed to be highlighted and was something we tracked during user playtesting. Shifting and Fade were one method of doing it but was only really utilized in playtesting by more experienced players or if you had a specific OOA build.
In a round-about method, I actually took some inspiration from the game Darkest Dungeon. A core part of a lot of the adventurer's kits in Darkest Dungeon was the position they're in. Maintaining an optimal position/order was a key part of our design goals - we thought the idea of some actions having additional key effects in specific positions or near other actions fit perfectly into our design goals.
This birthed the "Positional" and "Next To" effects - these two effects that I designed were made to now make specific actions heavily incentivized to be in specific positions. This would solve two problems at once.
A huge problem in the prototype was positioning being completely unpredictable due to no incentives. Attempting to guess your opponent's order was more or less useless, instead maximizing economy and then just randomly throwing out buffed cards was the optimal strategy. If your card lost, it was fine - you had 4 other extremely powerful cards queued that would make back the health different for any loss. If both players were engaging in this style of play, the game would devolve into a coin toss with random effects. This broke our design pillar intentions of making positioning king over raw numbers and focusing on skill-based gameplay. At the same time, the ability to perform economic plays like this is valuable. We see OOA builds as healthy for Auto Fighter so we couldn't just nerf economy-focused OOA build plays into the ground.
What Positional and Next To effects solved is that we can now keep cards strong (or even stronger) but mandate they be played in specific positions or near specific cards. It makes some cards incredibly difficult to shuffle around and limits the spaces where changes in position can happen. This further actually incentivized OOA builds and further solidified healthy counters to key OOA builds. It was an elegant design the team could now use to moderate power/develop counters but still give players the ability to create unique OOA builds. Now positions were no longer random - they usually had a calculated purpose to them assuming the player was engaging with the systems. If they weren't they were always using sub-optimal tactics and would increase their chances of losing.
The second thing Positional and Next To solved was the issue of short-term vs long-term play. A lot of cards in the prototype had no additional effects on a win outside of the aforementioned Shifting and Fade. The reason is if we just had an "On win, attack for free" effect, you could do some really powerful stuff. Opportunity for damage in Auto Fighter is highly valued, being stunned for one turn is almost a death sentence. Positional and Next To allowed us to include on-win/on-loss types of effects since they were in their "power position". The player trades predictability for raw power. It would lead to implicit positional play where cards that didn't need a position would now have their range of positions influenced by another keystone card.
Taking these new effects, I was assigned to rework almost all the base prototype cards and develop new content cards for the game using our XML-based card creation system.