Balancing the Balance

Balancing a game is arguably the most difficult and time consuming phase of design. When refining the mechanisms and trying to reach an Alpha and Beta state, you can grab new testers, test once, and gather the data you need to progress. However, with final balancing, not only do you need your mechanisms to not move at all, but you need to attempt to create controls in a realm full of variables to isolate and identify what’s out of whack.

We are in the balance stages of Hocus. We haven’t changed our mechanisms for about 6 months now, which means we’re in the fine print of balancing. Our testing matrix is quite complex, even for a simple game, due to a few factors:

  • 2-5 players can play. The game has a different texture if you’re playing with 2 players versus 5.
  • There are 9 different Spell Books, which means there are tons of permutations
  • Our game has a luck factor, due to the cards you're dealt and draw
  • Our game has a skill factor, due to the decisions you make in an ever adjusting situation

We’ve progressed through a few levels of balance. I’m going to discuss some of our efforts for Hocus, then for each case immediately broaden it to a higher level so that it’s useful for other. Essentially, I want this to be useful for all!

The Killer Hand: As we’ve written about prior, throughout the course of Hocus, we’ve found that by altering the probabilities of the game ever so slightly, and giving players further means by which to alter their chances, certain hands become far more probable. This is dangerous when those hands are things like a Full House, which is supposed to be one of the best hands in the game.

There are a few ways in which we ultimately balanced the killer hand for Hocus, including:

  • Limiting hands to 6 possible cards. With so much player agency, 7 cards is too many.
  • Adding a timing element to the game. A good strategy is to expedite the round such that those with a solid hand cannot profit too much.
  • Forcing players to pay for their own points. If multiple players think they’re in contention, a Pot can get quite big quite quickly. However, if everyone knows Bob has the killer hand, or is likely to, they’ll leave Bob to it. See bullet #2 on why this is problematic.
  • Letting players dynamically define what good is. If you give someone full reign over a community, they’ll build a Straight Flush, Full House, or powerful Flush or Straight. However, if you muddy their plans, suddenly a high Pair can be viable. It’s really about playing the board and not following along with someone else’s plans.

Your game may not have concerns with a killer hand, but there may be incredibly powerful cards about which people are worried, or certain strategies that seem very potent. The solution might not always be to fix every number so that things are perfectly balanced, but give players agency to balance things themselves.

I’ve heard, for example, that Old Friend in Last Will is too powerful. Old Friend gives you a bonus action every turn. However, if you observe, the person who goes absolutely first to get Old Friend will rarely get other worthwhile benefits. And, if you obtain cards such to deny them hefty combos, you’ll find Old Friend isn’t quite the deal breaker  you think at first glance.

One downside of a player agency driven solution is that in an age where people might not play your game a second time, they may not see that it is in fact balanced. If people just play once and don’t begin to dig into the game, they’ll leave with a bad first impression. You’ll have to evaluate if it’s worth the risk.

The Dominant Spell: Hocus has 9 Spell Books. All but one of them have 3 unique Spells that complement each other, and the 9th has 2 Spells used to manipulate a small deck of bizarre and wondrous cards. While it isn’t a CCG by any means, there’s a lot of content and many permutations here.

Throughout development, we’ve seen several cases where certain Spells would be used repeatedly by players. The idea is that sometimes it should appear to be the best option, but not always the best option. So, how do you preserve a card’s potency and intent without removing all of its teeth?

We utilized a few tactics.

Cost: Cards and time are the economic resource of Hocus. If you make someone discard cards, or draw, but do not allow them to advance the game state, they must choose to forego other opportunities and risk losing a window to use their cards by taking a Spell.

Synergy: Every Spell stands on its own, but some are clearly and obviously tied to a partner. If you do Spell A repeatedly, that’s fine, but until you utilize Spell B, you won’t see the full power of your battle station.

Time: I mentioned this before and I’m going to mention it again. If you allow players infinite time to experiment, dig, sample things, and pry, eventually, they will find the thing they want and win. However, if you give players a ticking clock, and provide incentives for others to push it, you’ll find that time waits for no player. Limited and unclear number of actions is a beautiful way to curtail a potent option.

Opposing Spells: We’ve deliberately seeded Hocus with abilities that can dominate a particular aspect of the game, but little else. Sure, you put a lot of big stuff in the Community, but the Illusion Wizard did as well...face down. Sure, you have three Pockets to choose from, but the Alchemist manipulated your Pot such that it’s of little value. The game is deeply interactive and everyone is intertwined. If you let someone do whatever they want in a vacuum, they’ll misbehave. If you force them to deal with their neighbors, interesting things happen.

It’s a mistake to remove powerful, big abilities from your game. It’s incredible fun when players feel like they are cheating, but doing so well within the bounds of your expectations. Being powerful is fun. However, you need to think about all the methods your game provides you to limit things.

Do you have a timer in the game? A way to force people to make choices?

Do you have any resources? A cost? These can be official resources, like food in Imperial Settlers, or subtle resources, like cards in hand, Victory Points, or, well, TIME.

Do you have interaction? Man is the greatest foe. Though every number may be mathematically sound, many games like Magic and Netrunner and Innovation show that interaction and cunning opponents are far more interesting solutions.

Intermission: The cluttered and plodding content

One of our best testers, Marguerite Cottrell, sent us a fantastic video a few months ago that provided some really wonderful insights we ourselves had missed. Essentially, she noted that Spell Books that did X tended to be better than Spell Books without. She also noted that every Spell Book tended to do a major and minor thing, but a few Spell Books didn’t really have a major thing.

Wonderful, insightful, and very good. The lesson is to find someone who can examine your game from a high level. Or, you yourself need to break out the spreadsheet and find ways to categorize and quantify your content.

However, as we rushed to balance leveraging these insights, my friend and tester Matt Yang noticed how much slower the game had become. In order to give everyone the identified X, we had doubled the content of almost every single turn in the game. Players now had to make 2 decisions that were deeply involved.

Oh dear.

We recognized that we needed to maintain nice, quick pacing for our game and balance. We thought way back to the original intent of the X and remembered that it was intended as a catch up in very very specific situations, namely, when spells had a very niche use that might not seem immediately valuable. We had then crept out from that to add this catch up to things that didn’t really need it, then to everything.

The lesson is to have a purpose for every decision you make. Remember why you did a thing and what problem you’re trying to solve. Keep that in mind as you apply that tool to other situations. Often, you’ll find that a fix for one problem is inappropriate for another and the consequences can lead to a major revision.

The Advanced Inclusion: Because we’re foolish, or wanted to have a ton of content in the game, we opened up our final 3 Spell Books to be a bit wonky. Whereas the first 6 Spell Books more or less just manipulate the various elements of the game in different ways, our final 3 Spell Books introduce new mechanisms and complexity.

Here, we have to balance a few things! Firstly, once someone learns how to navigate the complexity, are these Spell Books viable in competition? Secondly, how do we allow the game changing mechanisms but still keep them within balance? Often, abilities like this are very controversial. If you look at the Japanese faction for Imperial Settlers, you’ll see many players say “How the hell do I win with them!” That’s me. You’ll also see players say “Oh the Japanese are so powerful.”

Some of the divide here is simply due to the fact that the faction is so different from the others. Remember that perception is a big part of balance. If players are convinced it isn’t, no matter what you can prove otherwise, well, it isn’t. When introducing new mechanisms in cards, factions, and so forth, be sure to keep them simple enough that the learning curve does not adversely affect this perception.

The core backpressure of exception based design is accessibility. In this case, you’re almost less worried about balance at the outset and more concerned with: can my players get this and utilize it in a way that is compelling and competitive. When you introduce new mechanisms, keep that learning curve in check first.

The Control Experiment: Try to find ways to create control experiments. Give your best playtester (there’s always someone who is a super good player) the weaker abilities and see how they do with it. See if they can craft strategies and come out ahead. Give your weaker players some of the stronger, more apparent abilities and see if they can come out ahead. Keep track of what wins, and how often, and whether they win with the same strategy or there are different options.

You can also give your best player the most difficult content to leverage. For us, those are the exception based Spells. You can observe the difficulty ramp as it passes from your best to worst players.

Be sure to test with the same group over and over and to test the abilities about which you’re concerned against new ones. Think back to your high school science class. Pursue the testing methodically and take notes. Easy things to track include:

  • Final scores
  • Ways in which points are earned (and using what tactic)
  • Abilities used and how often
  • Reactions when certain abilities are used: Do people feel like it was fair? Unfair? Excitement? Frustration?

Hopefully some of this is useful as you enter the balance phase for your own game. It is difficult every time I encounter it, so I always learn with every try.

Comments

Cannot underscore a control group hard enough. Get the baseline experience understood before tweaking, and note when a change becomes the new baseline you want to maintain. v2.1a that stuff.