How I’d Do It: Episode 7

This week, we’re going to cover some dry, boring theory.

Yaaaaaaay!

So, yeah. Our topic today is the delicate art of balancing effects. It’s not glamorous or tremendously exciting, but it’s something every card designer needs to know.

Now, of course, balance isn’t really a design job, it’s development. Designers come up with the concepts, and developers take those concepts and balance them. It’s not a huge deal if Design sends Development a busted card, because the oversight will likely be caught and sent back for tweaking. That said, if all balancing is left up to the development team, the set will take ten times longer to make. You can never be 100% sure at design-time whether an effect you make will be balanced, but you can usually be 70-80% sure. I mean, I don’t think anybody would have much difficulty deciding if a 1-drop with “Activate -> KO target character” is balanced or not. A large amount of balancing can be done right up front, and it’s only the final tweaks that really need to happen in the development stage.

There are a lot of factors to keep in mind when balancing an effect. I’ve broken them up into six main elements: Precedent, Cost, Restriction, Timing, Parity, and Context.

Precedent

The single most important balance element is precedent. Basically, Precedent assumes that whatever the developers have done in the past, regardless of your own opinions of it, should be considered “balanced”. Some pieces of it– Fartifacts, EomE, Beatdown– are better than the average, and some– Dr. Tzin-Tzin, Unus– are below. But taken as a whole, the historical body of card content is to be considered the average.

So, how do you use that? Well, it basically means that when you sit down to balance a card, the first thing you should do is to look at what’s gone before. For an easy example, let’s look at one of the most common card type in the game– +ATK to an attacker. Looking over the history of the game, you can see some common patterns:

• +3ATK is the “average”. In addition, a +3 pump can have an additional small effect, like adding Flight or moving the character.
• +2ATK is the most common pump. These pumps are usually stamped to a team, and most often serve to give Sealed decks a smaller taste of an effect that is generally going to show up on rares in constructed play. These effects tend to be more substantial and game-state-changing than the effects on a +3 pump.
• +4 pumps aren’t common. Where they do show up, they generally have either no additional effect, or an additional cost.

This kind of research is tremendously helpful when doing a first pass on a card– since we assume that all precedent is balanced, then to create a new balanced card, it should fit somewhere within the scale here. If you want to make a pump of a certain size, you can see what kind of additional effect you can get away with. If you have an extra effect in mind, you can see what size of pump would be appropriate for it.

And of course, this applies to every other kind of effect. If you want to make an exhaustion effect, a quick Precedent scan will let you know how big of a character to let the player exhaust, what kind of extra cost should be paid for it, and when to let it happen.

Of course, Precedent alone doesn’t tell you everything. If I want to make a +3 pump, but I want it to have a larger effect than that class of pump would usually be allowed, then I need to start adding detriments to it– some sort of downside to counterbalance the advantages I’m giving it.

For our first practical example today, let’s make a Green Lantern character. The GLs as a whole have many common effects, from reinforcement to pump, but today I’m going to pick one of them at random– Recovery. I’m also going to arbitrarily decide that this effect is on a character; so, the most generic starting phrasing of this becomes:

• Free -> Recover target character.

Alrighty, there’s a good start. But, it definitely needs some work. Precedent tells us that this effect, especially in the reusable form of a character’s payment power, should not be this open-ended and easy to use. Most character recovery effects are usable only during the recovery phase (requiring you to keep the recoverer alive through combat), or else they come with a hefty cost of some sort.

Let’s look at the latter first.

Cost

Cost is the most common and obvious downside that can be given to any effect. “Activate -> Stun target character” is really powerful, but “Activate, discard thirty cards -> Stun target character” is… less so.

The idea when setting up any cost is to see what the player is gaining, and then determine a fair compensation for that gain. For our first example, we’re getting a recovery effect. If we wanted to go with the easiest, most obvious equal transfer of cost for effect, we’d have:

• Stun a character you control -> Recover target character.

It’s simple, and it’s… mostly balanced. While you can use this effect to maintain a better quality of non-stunned characters, you’re still stuck with the same quantity. Now, that said, this is a fairly boring implementation of this effect. One of the most exciting things about VS is that most effects aren’t strictly symmetrical– forcing a discard doesn’t simply cost you your own discard, and forcing an exhaustion doesn’t simply cost you an exhaustion. VS has a truly remarkable system to determine egality– it’s almost like having a currency exchange of common costs. For example, the following things have been shown to be roughly equivalent; ie, if a card is balanced with one of these effects, it’s probably balanced with any of them.

• 1 exhausted character
• 1 discarded card
• 5 endurance paid
• 1/2 a character KO’d (ie, you’d expect roughly TWICE the effect for this)
• 2 locations replaced
• etc
  

This is where the real heart of cost-balancing comes in– balancing “Discard a card –> Target opponent discards a card” is easy, but figuring out if “exhaust a character -> Target opponent discards a card” is balanced is trickier. This is where game knowledge and finesse come in– you need to think about what that cost means within the context of a game. A discard cost isn’t a huge burden, since you can plan around it; but a forced discard is a big deal, because if you time it right, you can completely derail your opponent. So even though an exhausted character is roughly equal to a discard when they’re both costs, it doesn’t balance out quite the same way when they’re on different sides of the arrow.

Let’s go back to our GL character above. We determined that “Free -> Recover target character” was way too open, and that “Stun a character -> Recover target character” is a bit too bland. As my third arbitrary decision about this character, I’m going to make him a 4-drop. Alright, that gives us some useful information, because it lets us do a pretty focused Precedent check to see what kind of costs should be associated with our effect.

One drop below, we have 3-drops Xorn and Dr Sivana, who costs you absolutely nothing, but require the recoverer to survive combat, which gets to be fairly difficult as the turns wind on. One cost above our guy, we have Blink and Illyana Rasputin– one still recovers at Recovery, but doesn’t need to survive herself. The other doesn’t have any permanent cost, but needs to sit out the combat phase. Right at 4, we have Alan Scott, who activates to recover an on-team guy, and Chthon, who activates to recover, but only if he’s hidden.

So… 3-drops only work during Recovery. 4-drops recover anytime, but need to exhaust for it. 5-drops recover with almost no cost. So… what do we do for our recovery guy? It would seem the baseline is:

• Activate -> Recover target character.

But that’s too open. It would work on a 5, but at 4 it’s a bit too easy. Do we want to add an extra cost?

• Activate, discard a card -> Recover target character.
• Activate, pay X endurance -> Recover target character with a cost of X.
• Activate, exhaust another character you control -> Recover target character.

Hmm. I don’t know about you, but this isn’t really working for me.

Balancing with costs alone is tricky. It leads to a need to balance an effect for both single-use (good enough if the character dies right away) and long-term use (weak enough not to massacre the opponent over four turns of use). It also gets tricky because the higher up you get on the “effect size” scale, the less relevant pure costing is– costs can often be completely circumvented (KOing freebies, playing card draw to power discards, exhausting hidden 1-drops, etc), so if all you’re working with is cost, you’re in danger of having an effect that’s far too easy for a dedicated combo deck to abuse.

This is why costs are very frequently combined with a second kind of detriment:

Restriction

Restrictions come in all sorts of flavors:

• Use/play only during the [X] phase.
• Use only once per turn.
• Target character with a cost of [X].
• Target character with the [X] team affiliation.
• Use/play only if [X].
• While you control [X]
• If you have more [X]
• etc

A nice metaphor for thinking about cost and restriction is this: costs only work in full dollars. You can make a $2 cost, or a $3 cost. But they’re strictly integers– you can never perfectly balance a $2.50 effect purely with costs. It’ll either be too cheap, or too expensive. In that case, you use restrictions– not to change the cost, which is still going to be $2 or $3. But with the addition or removal of the right restrictions, you can adjust that $2.50 effect until it’s bad enough to cost $2, or good enough to cost $3.

That’s basically what restrictions do– they add a limit to when, whether, how, or how often you can use an effect. An effect that can be fired once per turn is dramatically different to one that can be fired repeatedly. A 4-drop character with “Activate -> Stun target 3-drop” is dramatically different than one with “Activate -> Stun target 5-drop”.

Restrictions and costs can be used interchangeably– if a cost feels too hefty, you can move the balancing burden into heavier restrictions instead. And since, as mentioned above, restrictions give you much finer control over balance than costs, making that exchange might ultimately grant a more correct balance.

I think we finally have enough tools to make our 4-drop Green Lantern. Costs alone didn’t quite “do it” for me, mostly because it’s difficult to do anything really interesting with them. But with a combination of cost and restriction, you get a lot more room to maneuver. Keeping our Precedent check in mind, here are some random ideas that I think would work:

• Activate -> Recover target stunned Green Lantern character.
• Activate -> Target opponent chooses a stunned character you control. Recover that character.
• Activate -> Whenever target non-defending character becomes stunned this turn, recover it.

And we don’t even need to stick with Activate. Precedent tells us that it’s the standard, but if we want to lighten the load a bit and not force an exhaustion on the recoverer, we can add a heftier restriction on the back end to balance it out.

• Pay X endurance -> Recover target stunned character with a cost of X. Use only during the recovery phase.
• Pay 1 endurance -> Recover target stunned character if it was the target of an opposing effect this turn.
• Discard a card -> Recover target stunned character at the start of the recovery phase this turn. Use only during the build phase and only once per turn.

Ooh… I like that last one. As with anything, it would need to be tested to ensure that it’s not too abusable, but that seems fairly good to me. Endurance payments aren’t very “Green Lantern-y”, but the team does have a few card draw options, which translate well into allowing discard costs.

I think we’ve got a winner.

It’s a close analogue to Illyana, our 5-drop exemplar, but it has a narrower effect, and forces you to call your shot in advance. That second part is a major thing– beyond the simple “cost/benefit” balance we’ve been using up until now, it adds another axis to the measurement: Time.

Timing

Fan card designers don’t think about timing as much as they should. Looking at most custom sets, you’ll likely see a heavy weighting toward payment effects– cards that can be used, if not at the owner’s leisure, than at least at his discretion. You know all those power effects that end up on real cards that have you cursing, “GRRR! Why is this triggered?! Why can’t I fire it NOW?!”? Those powers that would be perfect for the combo you’re building, if only they weren’t build phase-restricted?

I’ll tell you something about those cards– they were designed by someone who knows what the hell he’s doing. Fan designers don’t put painful timing restrictions on cards, because we hate playing those effects. You can tell a pro from a hack, because pros have the stomach– and the balls– to make cards that make you cringe… and yet make you want to play them anyway.

Putting effects on a trigger is an extraordinarily powerful restriction. It may not always seem that way; a character who triggers at the start of combat is always going to get his effect, more or less. But think about it: wouldn’t you prefer to hold 7-drop Ronan’s replacement effect until… “whenever”? So that the opponent can’t just wait until it’s safe to start flipping things? Or, look at 6-drop Holocaust– I’ve recruited him almost a dozen times in my Unaffiliated deck over the past few months, and I’ve gotten his game-breaking power to fire… once.

Putting effects on fixed triggers takes control away from the player. The effect is immediately a little worse– ie, a little less broken– because the timing of it is fixed. The effect might fire too early or late to be helpful, or a clever opponent may be able to ensure it never fires at all.

Let’s use timing to build our next card. Since timing is a fairly heavy restriction if set up correctly, let’s go for a much stronger effect this time: returning a plot twist to hand.

(And the crowd went: “Ooooooohhhhhhhhhhhhh…..”)

As usual, let’s drop the most open phrasing of the effect to get started:

• Free -> Return target plot twist card in a KO’d pile to its owner’s hand.

Now, let’s run a Precedent check to see how this has been handled before at various points on the curve. Rama Tut from MOR bounced a plot twist for 2 resource points, but you needed to control Doom to make it happen. Since the only way to do that on 2 was by also drawing a location, and the additional reality that there’s very little of worth to bounce on turn 2, it’s pretty restrictive. Also on 2 is Namorita, who is almost strictly better than poor Rama– invest 2 points, and you can return something whenever the heck you want. On turn 3, you’ve got Floronic Man– no restrictions, but 3 full resource points invested. He seems a little worse than Namorita, and mostly he is, but it’s slightly counterbalanced by the fact that he only goes the KO’d pile, not the Removed From Play zone, so you can theoretically reuse him later on. At 5, we have Garth/Manta, who get you a returned card once per turn for a trivial endurance payment. At 7, we have Aquaman, who can get you up to three for no cost, simply for having characters next to him.

What we can see from this Precedent check is that returning a plot twist, while a fairly rare effect, isn’t terribly expensive when it does show up. At 2 and 3 it costs you resources, but how often does your 2-drop survive into a critical turn anyway? By 5, it’s become a throwaway effect. By this analysis, I’d say our sweet spot for card return is turn 4, but we just did a 4-drop, so let’s try a 3 instead.

You can see an obvious balance of costs vs. timing in the low drops. Floronic Man and Namorita outright cost you resources to use their effects, but in exchange, you can decide when the effect goes off. This can be tremendously powerful– the longer an opponent can keep them alive, the better the odds that they return something game-breaking instead of a random 1-cost pump. Rama shows us the impact of timing– if we want to keep our resource investment, then we need to take the returned card right away on 2, or we need to underdrop later. Either way, the effect determines your tempo for you.

So if a 2-cost character with a timing impact only gets to fire once… let’s see if we can balance a 3-coster that triggers repeatedly.

• At the start of the combat phase, return target plot twist card in a KO’d pile to its owner’s hand.

Even though the timing is rigid, it seems too good– it’s a once-per-turn like the 5-drops, without the cost. Granted, the timing might screw you up (since you can’t use it to double-Beatdown an attack), but it still seems a bit too open. So, where else can we move the trigger?

• At the start of the recovery phase, return target plot twist card in a KO’d pile to its owner’s hand.

That’s a little better– it makes sure you need to invest something in the power, since you’ll need to protect the character to keep it around. I could probably live with that effect, but I’m going to keep looking– the first card we made is pretty heavily tied into the GLs’ stall mechanism, and I’d like to see if I can make this work for their other theme, the fast-paced rush.

• Whenever this character enters combat, return target plot twist card in a KO’d pile to its owner’s hand.

Ooh! I think we’re onto something here. By using a combat trigger, we’re obliquely borrowing another piece of Precedent: Emil Hamilton <> Ruin. Ruin has a devastatingly powerful effect, but it’s counterbalanced by two related items. First of all, it triggers on combat, so unlike the Ahmed Samsarras and Umars of the world, you can’t just let the owner of the power sit back in (relative) safety and fire the effect every turn. In order to function, Ruin needs to be exposed to danger. I think we’re getting really close to an effect I’d be happy with, but there are still some balance issues. First, it’s not quite a fair exchange– you’re almost guaranteed to get at least one use out of it, so a free return may not be entirely fair.

The second issue brings us to our fifth element of balance theory. Because, while the effect is good, and it’s balanceable, I’m worried that it’s too much power in the wrong hands.

Let’s fix that.

Context

Context is an important balance tool that asks questions, not about the effect you’re making, but about who you’re giving it to. The simplest example of this is to look at the “Activate” cost. We’re so used to seeing that keyword that we sometimes forget how complex it is. “Activate”, when given to a 1-drop, might as well read “Free ->”. Exhausting a 1-drop is not a huge deal. But the exact same keyword, printed on a 7-drop, had damn well better be nearly or more useful than simply swinging with that 7-drop’s 16ATK. When you Activate someone, that is, most of the time, the only thing that character will be doing for you that turn. They can’t reinforce, they can’t attack, and they can’t exhaust for other effects. This isn’t a huge deal with a 1-drop, but losing the opportunity to swing with your 5-drop on your initiative because you wanted their activated power… hurts.

The second reason that Ruin’s power is reined in, is because it’s situated on a 3-drop. Whenever Ruin uses his power, he’s going to be stunning. And great as he is, I’m rarely going to keep him over my 4-drop. So what this means is that Ruin is likely only going to fire twice, and then he’s going away. There are several implications of this. First of all, he’s small enough not to be worth keeping over the next drop. 3-drops average 3/3 lower than 4s, so keeping a 3 over a 4 is rarely helpful. You need a full-sized pump to close that gap in the next turn’s combat. Whereas, if Ruin was a 4-drop, he might actually be worth keeping. 4-drops only have an average drop of 2/2 under 5-drops, which is a much more reasonable gap. That’s a gap you can close with even a wussy +2 pump, or even just some smart powering-up. On a 4-drop, the bonuses of Ruin’s power would make him worth keeping over a 5. On a 3-drop body, he’s not.

But on the flipside, Ruin is still large enough. He’s large enough not to be able to stun while attacking down, which can frustrate you when you need his search. He’s large enough to constitute a large investment of points, especially considering that we know he won’t be around very long. The context of Ruin’s power is flawlessly balanced; one drop higher or lower, and he’d become an Ahmed-level threat. But where he is now, he’s the perfect balance of big and small to give you a proper bang for your buck without making him completely abusable.

Context can rear its head in ways other than cost. 5-drop Mr. Sinister from MVL has a bit of context balance that I guarantee you’ve cringed over if you’ve ever played him: he doesn’t have Range. His power is crippling to certain decks, but with his purely average stats, it’s not terribly difficult to shut down. Of course, the standard response to this is to hide him, to shove him in the back row behind your next-biggest guy, and hope that his bodyguard will keep him safe. While this is an excellent move, it presents a big problem– you’ve just paid 5 resource points for what amounts to an ongoing plot twist. You don’t get any offensive combat power for your 5 points, because Sinister can’t attack from the back. So if you choose to hide him off-init, you’re giving your opponent the option of simply ignoring him, since he’s not a threat that they need to take out.

Context is the issue that caught my eye above with our second card design. I like the “enters combat” trigger, but on a 3-drop, it seems a little too good. While Ruin’s power seems to be almost the same thing, it’s subtly different– first of all, our power doesn’t require us to stun. As we just mentioned, most 3-drops are big enough not to stun when fighting 1s and 2s, so if you’re fighting a swarm player, a 3-drop could become an endless supply of Savage Beatdowns. Also, while Ruin’s power makes saving him from damage with large DEF pumps a pretty stupid maneuver, ours is more open, and actually becomes a pretty viable object of your defensive attentions. And, let’s face it– it is a LOT easier to keep a 3-drop alive than a 2-drop or even a 1. What that then means is that it’s not very hard to keep this dude alive until your important swing turns– 5 and 6– where he can start recycling some very powerful plot twists to guarantee a win.

So, what does all of that mean for us? It means that he’s probably a bit too big. By the same token, I don’t want to drop him down to a 1-drop, because that makes him trivial to replace on a later turn if you’re forced to lose him. So, I think his sweet spot is 2. With the cost adjustment, and an additional cost to the return (since, you’ll recall, I decided that a guaranteed freebie return was just a little too good), I think he’ll be just about right. A discard would be fair– a card for a card– but I’m not going to use that, because I’ve got one last balance method that I haven’t brought up yet.

No, instead of a discard, his final power will go in a slightly different direction.

Parity

“Oh, Spud. You sly devil, you.”

Parity brings us full circle to the discussion that opened the article: equal exchange. The same way that 1 discard = 1 draw (more or less) for you, 1 draw for you will more or less be equal to one draw for the opponent. The theory goes: anything is fair as long as it affects everyone equally. I can draw a card as long as you can; I’ll KO a character if you do it too.

But be warned: assuming that Parity equals balance is a very dangerous trap to fall in, because equal treatment is NOT always fair. In fact, it almost never is. Criminal Mastermind lets both players draw cards, but the opponent doesn’t WANT any more. Madelyne Pryor KOs “all stunned characters”, but this “balanced treatment” inherently favors the player with the fewest stunned characters– which, lo and behold, is very easy to do with Underworld’s all-concealed curve. Dirty Tricks makes combat harder for all players, but while this would hurt a rush player, a stall player relishes it. And while Shriek and Carnage give +1 or +2 ATK evenly to the entire board, that bonus inherently favors the attacker, and the player playing a swarm build.

Parity is only an illusion of balance. It will always favor the player who knows it’s coming, and the player who determines when it happens. You can’t simply slap “each player” on a card and assume it’s fine. In order for Parity to truly create balance on its own, it needs to be presented in a form where both players can manipulate it.

Of course, with all of that said, sometimes you want false balance. Madelyne Pryor isn’t truly fair, but if she was, she wouldn’t be worth playing. This is the final kick in the pants to this entire treatise on balance: PEOPLE DON’T BOTHER PLAYING WITH COMPLETELY BALANCED EFFECTS. Oh, sure– they’ll play a balanced 4-drop if that team has no alternative. But they’re only playing that team because the 3 and 5 are the craziness. In the end, VS is a contest of balanced imbalance– my guy is imbalanced in Way X, your guy is imbalanced in Way Y. Over the long term, X and Y create an equal amount of advantage, so even though our guys are not balanced on a micro level (“This guy’s too good for a 3″), every deck is going to have something that’s just as good (“That’s okay, my 4’s just as broken.”), so it’s mostly a wash.

Don’t you love it when I invalidate an entire article with a single sentence?

Way back at the top, I pointed out that “Discard a card -> Target opponent discards a card.” is a boring effect. I said that pure equal exchange– just like pure parity– may end up balanced, but it’s still likely to be pretty boring.

But, that doesn’t have to be true.

Ultimately, the point of all of this “balance” crap is to make the game fun. Broken cards cause a degenerative game, which quickly ceases to be entertaining. Balance ensures that nobody is guaranteed to lose from the moment they sit down, which means that you’re rarely consigned to a depressing thrashing.

However, all the rules and guidelines and principles I can spit out will never trump your raw, gut feeling that This card freaking rocks. It may break every rule of common sense and fly in the face of precedent, but that doesn’t automatically mean it’s a bad card. “Pure exchange” ends up being boring because most costs in this game are paid using the same half-dozen currencies– if you fall into the pure exchange trap without putting in the extra work to do it interestingly, your cards probably will suck. The idea I value most highly when designing cards– and I think UDE has a very similar rule– is that every card should be at least a little bit interesting. It doesn’t have to rock your socks off, but it should be different enough to at least require one synapse to fire for it to be used. Enforcing a “soft” self-restriction against pure exchange and other lazy traps really helps to ensure this core value is upheld, because it means you’ll never be allowed to get quite lazy enough to make that “Discard -> opponent discards” character.

But that’s only my rule. I am very lazy, so I need to set up these blocks and guards for myself. I need to identify the areas where my designs are weak, and absolutely forbid myself from repeating my sins. But don’t let my rules completely restrain you, by any means. Because as with all things, no rule of card design is absolute. Equal exchange can be done, and done well. Coup d’Etat proved that– there’s always room for a “tired old mechanic” to be combined with something unprecedented. If you find that the Precedent rule adequately covers every card you make, you aren’t making cool enough cards.

I’ll finish today by quoting one of the finest human beings who never lived.

“Go ahead– Take Chances! Make Mistakes!”

“AND GET MESSY!“

Peace out.

-Spud