Arcade Spirits: The New Challengers
Twitter / Discord / Email

Developer Blog - Designing Choices in Visual Novels

Now that all the character reveals are out of the way, this developer blog for Arcade Spirits: The New Challengers will be focusing on sharing inside news and peeks into the game's development! We've got a long haul ahead of us as we finish the game (all while grappling with the pandemic and pacing ourselves to avoid burnout). During that long haul we want to keep our fans entertained and informed. This week, we'll be talking about how we design the many choices you make during the game!

The Goal: Self-Expression and Roleplaying

Our game revolves around the five identity traits (Kindly, Quirky, Steady, Gutsy, Basically) and relationship points. Our goal was to allow players freedom to express themselves without feeling railroaded. The choices are forgiving, with only the occasional moment gated off by being too weak in a particular trait. The relationship system is also forgiving with huge leeway for what qualifies you for romance and positive endings; you don't have to scoop every single point to achieve your goals. Getting most of them, or simply more of them in proportion to others, is plenty.

Nearly every choice in the game consists of three options:

Whenever a relationship point is available, the point is always behind two of the three options. That means the neutral Basically option is just as valid a selection and will earn you a relationship point, in fact! And by spreading the K/Q/S/G opportunities out, you can't rely on any one trait every time.

At first glance, this works great -- you can explore a variety of responses, use the Basically option to avoid a trait you're just not interested in, and even deliberately fail if you aren't trying to build a high relationship with someone. If you accidentally fail, it doesn't really hurt you in the long run. Hooray!

And... then we released the game. And then we started watching people play it. And then we saw where the system came up short. Since there's always room for improvement, let's talk about that!

Problem 1: Obvious Right Answers

Okay, let's look at this choice and see what's wrong:

"Gosh, I just love cats! Cats are my thing. Anybody who doesn't like cats is a terrible person. Do you like cats?"

The problem here is pretty obvious -- players looking to scoop as many relationship points as possible feel they have to agree with every single view and opinion that a character has, no matter how they personally feel about it. It's a reward-minded mentality and it's fed by bad design, because the Gutsy answer is "obviously wrong" and the Basically answer is boring and feels like it should be wrong.

Solution: Unclear Answers or Equally Valid Answers

There's a few ways to kick this problem in the butt. First is to make it less clear what is the "right" answer, giving all of them weight because you aren't going in knowing what will work:

"Hey, do you like cats?" (without saying how the character feels about cats)

This works, but using an unclear answer is like adding spice to your dish; too much feels punishing. Another approach is to design equally valid answers. The above actually can work as-is for an equally valid approach, if you change the question:

"I just love cats! Are you a cat person? It's cool if you're not, just wondering, y'know?"

By not front loading it with an extremely biased viewpoint and a promise of punishment for not agreeing, it works just as well. It doesn't feel as high pressure and makes the Gutsy and Basically options more appealing, knowing that they aren't going to think less of you.

Another spin on it (but I won't type out a full example because this is already taking too long and I wanna go eat nachos) is to have every response say the same thing, but with different spins and twists on the idea. Agreeing politely, agreeing aggressively, or shrugging it off with a quick agreement. This works better when it's less about checking your opinion and more about an obviously right course of action, like cheering up a friend. You wouldn't pick "No, you suck, go and die in a hole" when your friend asks for your help, right? Right.

Problem 2: Risk-Averse Thinking

Did you know that in the Mass Effect series, a whopping 92% of people picked Paragon options, despite Renegade options being available? Makes sense, though. Years of games have taught players that if you are impolite, risky, or aggressive that games will punish you for it, even when the designers do their best to make all options viable. That's compounded by a natural desire to be reasonable, agreeable, and pleasant when a game is all about relationship building.

Arcade Spirits hits the same problem, of course. When you're relationship-building and maintaining the health of your arcade, decisions tend to come down to "What is the safest play in this situation?" even when every single option presented is actually valid. The player doesn't know that they can't fail -- they just see a problem and want the safest solution.

Even less mission-critical situations can run afoul of this. In this case, these three choices all lead to long character dialogue scenes that help you build relationships... but when phrased this way, it looks like life or death:

If Alice is bored, maybe talking to her would help. That's appealing, but it's also boring, because... well, boredom. Bob is busy reading a book -- emphasis on the word busy -- so interrupting him could be seen as a risky or impolite thing. But Chuck...? Oh man, Chuck is anxious and that's a problem we have to solve! So, the player gravitates to Chuck, even if they really wanted to romance Bob or Alice.

Solution: Equal Pressure Choices

There's an easy fix for this -- don't make Chuck visibly freaking out and anxious. Chuck can still be anxious once you go and actually talk to him, but make the choice itself a lower pressure option. "Chuck is planning his next move." or "Chuck is considering what comes next." would be just as good. Same applies to problem-solving situations; make sure every solution looks like it'd work, and then let the player decide which one fits their mood.

Problem 3: The Kindly Meta

The first two problems naturally lead to the third... the Kindly option. Everybody loves Kindly! Who doesn't want to be Kindly? Bad people, that's who. Kindly gets results, Kindly solves problems, Kindly shows you are listening and helping and always there for the ones you care about. That's great!

But... Kindly is one of five options. And if we don't make all five options valid, players will only see 1/5th of the writing in the game. Wasted writing is wasted hours and wasted money, and as an indie dev, you really can't afford that. A complicated and extensive system of decision making that just boils down to "click the green heart every time it's available" is kind of wasteful, yeah?

In Arcade Spirits each character represented a unique combination of traits. Naomi was Kindly/Gutsy, Gavin was Steady/Kindly, Percy was Kindly/Quirky. So you'd expect Kindly responses work well with all three, Naomi should enjoy it when you get Gutsy, Gavin appreciates someone who is Steady, Percy likes Quirky...

...except when you first meet Percy, cracking a joke about his wealth upsets him. And that trains the player to stick to the Kindly path, because it's safer. Similarly Teo (Steady/Gutsy) and Ashley (Steady/Quirky) are not Kindly but never respond badly to Kindly and often aren't into Steady, so it trains the player to keep hittin' that green heart, again and again.

Solution: Re-balancing the Results

Sticking closer to the archetypes is good, with the occasional curveball to keep people on their toes. (Don't want to be too predictable.) By implementing all the solutions above -- making less times where Kindly is the obvious right answer or the safer path -- and generally keeping a better balance between the five results, we can beat the Kindly Meta.

That doesn't mean you can't be Kindly in the New Challengers, of course! We love our Kindly players. It just means Kindly is now equally valid, instead of being the dominant strategy. Everybody has fun, Kindly or not.

Problem 4: Nobody Wants To Be Basic

"Basic" was intended to be a neutral word, but has a slightly sexist cultural connotation of "being a basic bitch" that almost every single player knows by heart. This leads them to think the Basically options are failures, things you pick if you suck.

Solution: Just Rename It, Duh.

So... we're calling it "Flexibly" instead. Done. Honestly we'd go back and patch Arcade Spirits to use this word but, well, eventually you need to draw a line and call a work finished, y'know?

In Conclusion

We've learned a lot from streams and from talking with our fans. We've grown considerably since the first game, and are looking to make Arcade Spirits: The New Challengers a true roleplaying experience, perfectly balanced, as all things should be. We're looking forward to seeing what choices you make when our game launches in the distant future year 20XX!