Getting That Balance

Recently I’ve been working on The Day After, my board-game that became a board-game themed PC game. Primarily I’ve been writing the scripting language for it so I can start sewing in game mechanics and messing around with things. I’m using a scripting language because you can iterate (and debug) quicker that way. Anyway, working on the general gameplay model lead me to working on the scripting language, which lead me to nailing down some of the concepts I had floating around. One of them is Roles and balancing them for gameplay. I thought I might chat about that.

The Day After

In The Day After, you are a group of people who have survived some spectacular catastrophe. I’m still working on the catastrophe – for arguments sake, let’s say it’s like zombies. Something’s happened and a large portion of the population have gone mad, killing each other. The Survivors find themselves hemmed into a city by the military quarantine. If they want rescue, they need to complete a bunch of objectives and try to survive until the rescue attempt.

Like most zombie movies, your Survivors are a rag-tag (perhaps stereotypical) bunch of people from all walks of life. Their different skills help the group survive in different ways. The way I want it to work is that everyone has a Role, which gives you certain strengths and weaknesses. It also gives you personal goals that you need to accomplish during the survival/rescue stage.

My current list has just under 20 roles, including: doctor, soldier, spy, socialite, scientist and courier. Each have been devised to cover some aspect of survival: killing things, keeping your folks alive, scavenging, and figuring out what caused the catastrophe. The variety of roles offers a lot of different ways to approach the game, and hopefully offers interesting replay options. For example, a game might be easier with a doctor and a soldier in the team, requiring fairly straightforward strategies. If you have a socialite and a courier, you don’t have that safety net, so you have to adapt to survive.

At this point I’m getting the feeling that I have too many roles, or they’re not distinct enough. Having heaps of roles isn’t a bad thing given the modular way the game works. It’s just that players might not care about the more oddball options. The Roles also reflect the game mechanics. Do I have too many options? Too few? Is there enough interconnectedness and emergence between the game mechanics to offer a variety of strategies and approaches. And how do you balance it? I’m hoping looking at balancing will put a spotlight on some of these other questions.

The Opposition Model

One model for balancing roles that I can’t use, but might salvage something from is TF2’s. I know it well. You have 9 classes, each designed with a particular weakness (and secondary to that, a main strength). Each class has hard or soft counters.

For example, the Heavy is a big guy who deals a lot of damage from short- to mid-range. However he is big and slow, making him a perfect target for the Sniper. The Sniper can stay outside the Heavy’s range and get a well-aimed kill shot off without being in much danger. The Sniper has a counter in the form of the Spy. The Spy can sneak up behind him and stab him in the back while he’s busy staring down the scope. The Spy, in turn, is vulnerable to the flame-spewing Pyro, who can stop the Spy from sneaking in… And so on. Because teams have identical options available, you can push a strategy a certain way by having a different assortment of classes. A BLU Sniper is the same as a RED Sniper (except for weapons, but let’s ignore that). But if BLU has too many Snipers killing your precious Heavy, you can just go Spy and kill the lot of them.

I think this system of opposing forces is excellent, but not relevant to The Day After. It’s geared towards two equally matched teams being able to configure themselves for their current strategy. In The Day After, we have one team versus the world. Opponents are meant to be mown down, so you don’t want to balance the Roles versus the bad guys too much. The idea of each team member contributing in their own special way via their Role is a neat idea though.

Build Points Model

In RPGs, one way of rolling up a character is that you’re given a certain amount of “build points”. Your six main attributes (Strength, Dexterity, Constitution, Intelligence, Wisdom and Charisma) have a certain base value and you can increase each by spending build points. You can also increase your build points by taking penalties in some attribute. Buying a lot of points in one attribute is more expensive than buying the same number of points, but spread across a bunch of attributes.

To be sure, I don’t want players in The Day After to roll up their own freeform characters, mostly for theme and simplicity reasons. But I may still use some sort of similar system to create the Roles. I think a 0-5 system might work well. 1 is basic ability, 3 is average, 5 is excellent. 0 is specially allocated to say a Role cannot do this thing.

There are a lot of subtleties in how you do this. For example, are all attributes worth the same? Is increasing your Shooting ability by one the same value as increasing your Scavenging ability by one? Maybe, maybe not. Shooting things is universally applicable. Scavenging is only useful when you can.

Should increasing an attribute by 2 be twice as expensive as increasing two attributes by 1? Or should it be a non-linear relation? If non-linear, are the bonuses worth it? Suppose you have an extra point in Shooting versus your friend. If you’re only doing an extra point of damage, but it’s inherently cost you (say) 4 times as much, is that a fair tradeoff?

The way D&D works is that everything is more-or-less about combat. In the vast majority of games, most adventures are decided by combats. Clearly Strength, Dexterity and Constitution have their place here. The last three non-physical attributes  play into things like spell-casting. Typically Charisma has not much of a role to play for most characters, so some players scavenge points from Charisma to augment their Strength (for example). Nevertheless Charisma might play a role outside of combat (perhaps even in avoiding it!)

The Day After is about surviving. There’s a combat role, but I didn’t want that to be the only way to win. One way I’ve thought about it is that combat ability helps you not die, but the other abilities allow you to actually achieve objectives or find equipment that helps you in the long-term. Following from this, if you have a wide variety of non-combat skills, then it might be harder to achieve objectives (because your abilities might be spread thin across objectives).

In any case, if you have a different emphasis on skills, you can weight points on their relative usefulness and scarcity. Maybe Hacking is rare, so you can increase Hacking by 2 for the same cost as increasing Combat by 1. There’s still a tremendous amount of balancing to be done.

The Simple Model

One disadvantage with the points model is that you have to assign every Role a grading in every ability. This is easy with something like the Soldier – just give him a bunch of combat-related bonuses. For something like the Courier, you give him speed… but then what? Can you justify it, in-theme?

One solution is to make every Role a one-trick pony. Or perhaps have a primary focus and a secondary focus, and that’s it. Everything else they either have zero skill, or some baseline. This simplifies the balancing act significantly. However its effect on gameplay might be too strong. Combat always comes up, so a combat-oriented Role is good. But someone specialized in Hacking? Surely you can’t hack as many things as you can fight, especially in a post-catastrophe city lacking power and civilization.

Other Roles like the Doctor seem natural to be in the game, but we don’t want a whole Healing skill just for them. Perhaps you can trade-off a primary or secondary focus for a special ability. But what’s that worth and how do you balance them? We quickly return to an earlier situation.

What’s seems clear to me is that we need to go to one of two extremes: worry about thematic skills, keep it simple, and balance just by feel; or at the other extreme, equate skills to some sort of survival value and balance the numbers with iterative re-evaluation. The first is easier to put together, but can easily lead to dominant strategies or situations. The second is more difficult to get right, has exponential complexity with every skill I add, but has the best chance of being optimized via simulation.

To be honest, I’m still working on this. I want a clean, simple design, but I want enough options to provide more gameplay. I want things to be thematically agreeable, but be balanced when you look at the numbers. Hopefully I can get this scripting language written so I can just dive in and see what I come up with.