Design Methodology in Avatar Warrior’s Journey: Designing for System-Wide Balance
Avatar Warrior’s Journey is a 2D endless runner with a card collection mechanic. As you complete goals, more map locations open up and you get a chance of running through these new environments. Your character would equip Gear through cards that can be levelled up and these cards are acquired through a blind-pick Card Pack mechanism.
In this series of articles, I’ll run through some of the design processes behind balancing these mechanisms.
At LucasArts, one of my favorite projects was this unreleased social game called Outpost that you can read somewhat about here. One of my key interests as a designer is to formalise a process to work with numbers and I carried over some of that learning from Outpost into AWJ. You can read about the basis of that thought process in my previous articles about player progression here — Part I and Part II.
In AWJ, you gain currency directly related to the number of orbs you collect during your run. Like in any endless runner, as you cover more distance, you’ll encounter a speedup and at intervals, a bonus mode allowing you to collect more orbs.
Game Currency is used to purchase Gear cards and you would equip Gear to reward you with a bonus, be it in increasing your Orbs gain or a chance of more specific obstacle encounters. Collecting duplicates of a Gear card will increase the soft cap and running with a Gear card equipped will go towards levelling up that card to that cap.
There are 3 major points to focus on for the purpose of currency balance with regard to progression
- New players are likely to run a shorter distance both by design as well as by player skill progression.
- Each tile features a different number of orbs and as you go further, tiles will contain different routes that a player can take as well as different orb scores.
- There’s really nothing stopping a really good player from running really really really far, hence gaining currency at a much faster pace than expected.
- There’s also nothing stopping a player from collecting all orbs or just some orbs during his run.
The latter two points are consistent with our design philosophy of wanting AWJ to not be a pay-to-win type of game.
The first step was to lay down a baseline for how far we want player progression to be and through that, a rough estimate of orbs gain can be calculated. We looked into a few possibilities including how an average player might play and how a good player might play. The variation in this is important because it controls how much incentive a player has to improve his skills and also it’s useful as a metric to measure for pain points in progression at a later stage.
Next, we looked into the cards to create a conversion for Gear level gain. Each level gain is measured by distance and this corresponds to a number of games played.
With this, we next looked at how to create a fair system for Gear bonuses. Gear bonuses are a strategic means for the player to receive more rewards depending on his/her play style and should be viewed as an investment. With any investment, the key goal is to receive a pay-off that exceeds its price. One of the key considerations is how many games would a player have to play before he/she starts profiting and since this balance is expected to be tweaked after the game goes live, we want to easily see the quantitative effects of any tweaks we make. For all of these reasons, we’ve decided on using distance as a basis for comparison across systems.
From here, we’re able to project a conversion rate between Gear value, the monetary worth of it, the number of games played as well as the fluctuation of value between a newcomer and a veteran, the latter being useful for measuring rate of progression.
Personally, I feel that a good variable choice as a ‘currency exchange’ is important especially for F2P games because once launched, you wouldn’t want to mess with the pricing for ingame currency (ie. if you priced $1 for 100 Gold, you don’t want to change this). First-adopter players who have bought into your system are likely to be your most loyal fanbase and you don’t want them to feel cheated. Unless it is so far off the mark that you have to change it, you’ll want to instead look into rebalancing other parts of your system to make it coherent.
In the case of AWJ, we realised very early on with playtests that players aren’t getting as far as projected. We wanted a game session to be of a certain length and we ended up altering the tile difficulty and bringing forward the bonus mode so players were gaining more distance and orbs earlier on.
The last topic I want to touch on here is the type of metrics that are useful and what we looked out for during these playtests. As mentioned above, with skill-based games, there will be outlier players who are exceptionally good or bad. At an early phase, our classification of good and average players was based off the projected percentage success of collecting orbs. While testing with a small group, it is important that the quantile accomplishments are used rather than an average because these outliers are more likely to skew results.
With a small test group, it gets increasingly difficult to gauge for balance since their skill level would have increased. One of the indicators we have built in is a 1–10 scale for rating tiles according to factors like the obstacles that are on it. This allows for a more measured means of gauging a player experience and for projecting the difference in difficulty a gameplay change will make the game.
