Archive for Malifaux

Designing Leveticus

Levy

Leveticus was an incredibly difficult master to convert over to second edition. His playtest was long and rather painful. He was the only master who was tested publicly in both wave 1 and wave 2 and he saw more revisions than any other master. In the end it was all worth it, because I think we ended up with a very good result. However, his design was interesting and led us to some new design techniques, so I thought he deserved his own blog post.

In the first iteration of Leveticus, we tried to keep him as close to his 1.5 incarnation as we could. Unfortunately, in 1.5 he was a master with an incredibly diverse hiring pool, great damage, good summoning, and he hardly ever died. Combining all of those things into second edition was, to say the least, too much. This was our initial mistake, as I think it caused some perception issues right off the bat and led to a lot of trouble later on down the road.

Once we accepted that we couldn’t keep Leveticus as he was in 1.5, we had to think about how we were going to convert him over to second edition. I wanted to clean up some of his abilities, split his hiring options onto two limited upgrades, and tone down some of the more powerful aspects of his card so he could keep his flavor without being over powered. However, Mack had other ideas (this was still during Wave 1, so he was involved in Malifaux testing at the time).

Mack wanted to write down each of Leveticus’s abilities onto scraps of paper and then randomly draw them from a hat to see where they would go. For us this was a tried and true method, but I felt that Leveticus needed a bit more finesse in his design. The argument got heated, so I took it to Eric. As usual, Eric smacked me full on in the face and shouted, “Never question the hat!” This was, more or less, the response I expected, but I thought I had to try.

Unsurprisingly, Leveticus players were not pleased by the results of the hat method. I begged Eric and Mack to at least consider casting stones or reading the entrails of a goat, but they refused. Mack actually seemed pleased with the dismay on the forums. When I confronted him about it, he kept mumbling something about needing their tears to power his machine.

Quickly I realized that this master was not going to be ready for wave 1, so I pulled him and continued testing him through wave 2. During wave 2 Mack was distracted by the RPG so I was able to use a Magic 8 Ball like I had wanted to in the first place.

Well, that’s the story.

Although there is one thing I never really got closure on. I’m told Mack’s machine is till out there. Waiting.

 

Happy April Fool’s Day!

The Beta: Before and After

The wave 2 cards are finalized and available for download, so I thought it would be interesting to look at how far we came during the beta. For this blog, I grabbed the oldest iteration of each wave 2 master I could find and put it next to its finalized version. This doesn’t include upgrades or supporting models, but it gives a good general sense of how each master developed, and gives some of you a glimpse behind the curtain at iterations that never even made it into the public beta. It’s surprising how little some of them changed, and others were rewritten entirely. It just goes to show how much the beta accomplished! (I would suggest opening these with Adobe Reader)

Guild

Lucius Before the Beta

Lucius After the Beta

Hoffman Before the Beta

Hoffman After the Beta

Outcasts

Jack Daw Before the Beta

Jack Daw After the Beta

Hamelin Before the Beta

Hamelin After the Beta

Leveticus Before the Beta

Leveticus After the Beta

Resurrectionists

Kirai Before the Beta

Kirai After the Beta

Molly Before the Beta

Molly After the Beta

Ten Thunders

McCabe Before the Beta

McCabe After the Beta

Yan Lo Before the Beta

Yan Lo After the Beta

Shenlong Before the Beta

Shenlong After the Beta

Arcanists

Colette Before the Beta

Colette After the Beta

Ironsides Before the Beta

Ironsides After the Beta

Kaeris Before the Beta

Kaeris After the Beta

Neverborn

Collodi Before the Beta

Collodi After the Beta

Dreamer Before the Beta

Dreamer After the Beta

Gremlins

Mah Tucket Before the Beta

Mah Tucket After the Beta

Ulix Before the Beta

Ulix After the Beta

Wong Before the Beta

Wong After the Beta

 

The Language of Games

heiroglyphics

To a certain extent, I see rules as a language. The strength of having clear, well-defined rules is that two players who have never met before can come together and play a game without any arguments about what their models do.

This is why I will never condemn a player for wanting the rules to do exactly what they say. Only when players interpret the rules as they are written, not as they think the rules *should* be, can a game accomplish that level of clarity. This goes back to the old argument of Rules As Written (RAW) vs. Rules As Intended (RAI). A player who wants to play the rules exactly as they are written is a fan of RAW, where a player who tries to guess what the designer intended is a fan of RAI. The issue with attempting to guess another person’s intent is that not everyone’s interpretation of that intent will be the same, so house rulings and patches crop up, making it more difficult for gamers from two different groups to play a game together. In other words, as time goes on and more house rules are made, those two groups of gamers will slowly evolve separate rules languages until they are ultimately unable to communicate.

Now, I have no issue with house rules, so long as players recognize them for what they are. If it isn’t obvious, I’m more of a fan of RAW than I am of RAI, although I would never judge another person for how they choose to play the game. I just think that playing the game as it is written leads to the least amount of confusion.

This may be a bit of a surprise to some of you, after all, I wrote the rules (well, good portions of them, it’s a group effort). Why wouldn’t I want people playing the rules as I intended them? And haven’t I been seen on the forums calling for a “common sense” interpretation of the rules? Doesn’t all of this point to more of a RAI point of view?

Not necessarily. While I most certainly would like the rules played as I intended them, I find that when people guess at my intentions they are rarely correct (no hard feelings – I probably wouldn’t be able to guess about your intentions either). Also, in my experience, telling the difference between RAW and RAI can actually be more difficult than it seems. Rules As Written tends to have the bad reputation of being the most lawyeresque, strict, knit-picking interpretation of the rules possible. I would argue that this isn’t always the case and that, frequently, Rules As Intended arguments end up being far more strict and counter-intuitive. Let me give you an example.

In this thread the following question came up: Does Lilith need to be able to draw Line of Sight (LoS) to a Silurid in order to target it with a Ca Action?

The reason for this question was that Lilith and the Silurid have the following abilities:

On Lilith: Master of Malifaux: This model is not affected by severe or hazardous terrain and does not need LoS to target models with Charge or Ca Actions.

On the Silurid: Silent: Models cannot ignore LoS or cover when targeting this model.

Taken at face value it would seem that Lilith would indeed need LoS to target the Silurid, as her Master of Malifaux ability is negated by the Silurid’s Silent ability. This makes the most intuitive sense, because otherwise there is seemingly little reason to have the Silent ability at all.

On the other hand, the argument was made that Silent stops models from ignoring LoS restrictions, but Master of Malifaux simply states Lilith does not need LoS. In other words, Lilith doesn’t ignore LoS, she simply doesn’t need it, therefore she can target the Silurid. At first, this would appear to be the proper RAW interpretation of the rules. However, I would argue that it is not.

This entire argument is predicated on “ignoring” a restriction being different than “not needing to adhere to” a restriction. Although slightly different wording is used, they do indeed mean the same thing, and that’s the important part. Because they mean the same thing in the literal sense of the English language (although not necessarily the exact wording used) the two rules do interact with each other, and Silent will effectively negate Master of Malifaux, forcing Lilith to draw LoS as normal. The best argument against this interpretation is that if the designers had intended these two rules to interact, they would have been sure to use the same language; in other words, it’s a Rules As Intended interpretation (and an incorrect one) which is the more strict and counter-intuitive of the two arguments.

I think I have shown how trying to guess the designer’s intent can actually lead to the less intuitive conclusion. But it begs the questions, why don’t we just always use consistent wording to avoid this? And when does strict wording matter?

Strict wording matters when you are dealing with game terms. Game terms are definitions which are strictly outlined in the rulebook and which have a different definition than the usual English definition. For example, in Malifaux, “killed” is a game term. When a rule says that a model is “killed,” it does not mean the model’s heart stopped beating, it means the model is removed from the table and drops any applicable markers. This is very different than the English definition, so we spell it out in the book, and it means something entirely different than other terms (like sacrifice). Placement is another game term (call out box pg. 51), because in Malifaux “placement” is different than “movement” (when in the literal sense placing something is generally going to involve moving it). Reducing damage is different than preventing damage, etc. When the rules use a specific game term, they are speaking very strictly about that term, and that term will be defined in the book. However, when the rules use the English language, they are referring to the literal definition of that language, but the exact phrasing isn’t as important (this is why “ignoring” a restriction is the same thing as “not needing” a restriction).

Alright, now we know when to adhere to strict wording and when not to. All that said, what about the first question, why not just use perfectly consistent wording in all rules which the designers intend to interact? Why use “ignore” and “does not need” at all? Why not just choose one and stick to it? Part of the issue is sheer manpower and deadlines. With enough people and enough time, we could certainly do a better job of this. But, even with infinite time and manpower, there are still very good reasons to rely on the English language definitions as opposed to making everything into a game term. I give an example of this in the thread I linked above:

This is hardly the only example in the rules (NOTE: speaking about the Silurid/Lilith debate), it’s just the one the forums have latched onto for the moment. For example, a model which is inflicting damage usually “deals” damage. A model which is taking damage usually “suffers” damage. Ideally, these would have been the exact same term. But it would have lead to weird sentences like this:

“When another model is suffering damage, this model may discard a card to force the target to suffer 1 additional damage.”

Compare this to:

“When another model is suffering damage, this model may discard a card to deal 1 additional damage.”

You can see the economy of space there, if nothing else. And, before people pick this apart and start trying to show how they could have used the exact same wording while saving space AND making sense, keep in mind this is only one example. You would have to do that hundreds of times, no two the same, and screwing it up even slightly means nobody understands the rule.

The dealt/suffer wording in regards to damage is just one example. And it’s one which could have a weird rules interpretation, if the RAI route is taken. For example, the Armor ability states:

Armor +1: Reduce all damage this model suffers by +1, to a minimum of 1.

This could lead to the argument, “If the designers had intended Armor to reduce damage from this *insert action or ability here*, they would have used the wording ‘suffer’ instead of ‘dealt.’”

Lightsaber

Clearly this is not the interpretation we wanted (nor is it the correct one) so why did we leave ourselves open to it? Two reasons: grammar and card space.

Depending on which model is the subject of the sentence, the appropriate word changes. For example, if the model taking damage is the subject, then “suffers” is the appropriate word to use (i.e. this model suffers damage. The model being damaged is the subject). However, if a model is inflicting damage, then “deals” is the appropriate word to use (i.e. this model deals damage. The model doing damage is the subject). Keeping the wording consistent for all situations would have involved some creative writing to change the subject of the sentence, which would have led to some strange (but totally consistent) wording. Unfortunately, this wording makes less sense in the context of the English language, and English wins, because it’s what we speak. Ignoring the language we speak makes this game a lot less accessible to new players.

Economy of space is equally important. Ultimately, a card is a pretty tiny space in which to write a complex rule. You have no idea how many hours I have spent tweaking, rewording, and adjusting kerning to make the rules fit. That extra line or two does, absolutely, matter.

And, at the end of the day…nobody has ever (to my knowledge) made this argument about Armor (yes, I know, they will now. Shut up.) The rules are intuitive, players get it. And that counts for a lot.

Armor cat

Armor cat laughs at your rules arguments.

I think I have made my position about as clear as I can, but there is one more question I would like to address. Why don’t we just add a few extra lines to make the rules as absolutely clear as possible? Well, first is manpower and deadlines, of course I wish we were better at this. Even with unlimited resources, we still run into the card space issues I mentioned above. However, there is one more reason we don’t always do this: Every line is a line that can be misinterpreted.

I almost want to get that tattooed on me somewhere. It would just be awkward quoting myself.

Every line is a line that can be misinterpreted.

-The dude who’s elbow this is

I could write paragraphs about how each individual rule works. But somewhere buried in that paragraph would be a line that is either ambiguous or which people really want to mean something else, and a whole new argument starts. When writing rules, you just need to keep them succinct, clear, and then encourage the player base to interpret them with a grain of common sense.

 

After thought: how many pages are the official Magic the Gathering rules? If you don’t already know, take a guess before you click.

Running The Malifaux Beta

beta

There are two key skills to game design; predicting human behavior, and running the numbers. Games are played by humans, so predicting how they will react in the game (and how they will feel about the game) is absolutely key. That’s not something you can figure out just by calculating statistics, because people don’t always make the logical choice (or even play a game for the same reasons). On the other hand, you don’t want a system that can be broken and manipulated, so you also need to run some hard numbers and know your statistics. This is important both so that you can make sure that the most intuitive choice is also the most statistically viable choice, and to make sure that six months after release someone who does understand hard statistics doesn’t release a net list that dominates your meta.

A single game designer only needs one of these skills; but a game will need people with both skill sets to function. I’d like to think I’m capable of doing the tight rope walk between both of these skills, I would argue they’re pretty much the only two things I’m good at. Luckily, running a public beta falls very heavily into one of these areas: predicting behavior.

The past four months of public beta have been a hell of a ride. If a beta like this goes bad, it can go VERY bad. Every time I mentioned to another game designer that I was running a public beta, I always detected a little bit of a cringe. All too often it can just turn into a non-productive hate fest that nobody wants to talk about once it’s done. I think we can all agree that was not the case for the wave two testing, but why not?

First, I need to give credit to the community. You guys are awesome. The Malifaux community is one of the most mature, respectful gaming communities out there, and I don’t know if I could have pulled this off with another fan base. Also, I need to give credit to the hard working moderators who kept an eye on things.

That said, we are all gamers. We were all on the internet. And we all showed up to the beta to argue a point. Even with an amazing community, that can be a recipe for disaster, so I needed to get ahead of the game. I needed to provide an environment where the great community we had could live up to its full potential.

First and most importantly, I set a schedule and I stuck to it. Updates would come once a week, every week, on the same day. This is something I instituted in the Showdown playtest, insisted on continuing during the wave 1 beta, and kept up here. Updates for this beta came every Tuesday. Even the day was no accident. I poled a number of people behind the scenes and found that Monday was the most convenient day for a beta update (followed by Tuesday). Unfortunately, Monday didn’t work for me, because most playtest games are played over the weekend and I don’t even see them until Monday, so Tuesday became update day. This gave me enough time to input the data, and gave the player base enough time to digest the rules before their games on the weekend. I stuck to this schedule rain or shine, posting updates on Christmas Eve and New Years Eve. I did post two updates at odd times in the morning the day before my daughter was born and the day we came home from the hospital (yes, we spent a week in the hospital, that was fun) but that was pretty much the only thing I let get in my way.

I kept to such a strict schedule because a predictable schedule is probably the best behavior management tool you can ever have. When people don’t know when things will happen, they get antsy and, worse, they get bored. People *want* to be productive but, if they can’t have that, they’ll always settle for a little destruction. I can’t point to examples of this in the beta because I didn’t let it happen but, after five years with the school district, I can assure you it would have (and no, I’m not comparing you to children. It’s just how people are. I’m the same way. How many horror stories start out, “So, I was bored one day…” Internet flame wars are no different.)

My next most powerful tool was allowing people to feel heard. I accomplished this in a number of ways. The first and most effective way to allow people to be heard in the beta was simply to put in the changes they were asking for. How many of you can proudly point to an ability and say, “That was my idea!” That’s an awesome feeling, even if you were just pushing for a slightly lower/higher stat. I say that this is the easiest way because…it’s the entire point of the beta. People showed up to influence the game, allowing them to see that influence kept the beta a positive experience. It’s what I wanted, and it’s what the players wanted (efficiency!). Now, that’s all well and fine when the ideas people were pushing were good. Obviously, that can’t always be the case. However, more than once I put a change in I knew would probably be removed just so people could tell I was listening, and try it out for themselves (and, hey, sometimes I was wrong and those things stuck).

Of course, sometimes I reached a point where this wouldn’t work. The suggestions were too varied or there wasn’t enough time to test things I knew might be dead ends. In these instances I jumped in, brain stormed with the community, and did a “mid-week update.” Basically, I tried to distill what people were saying and actively worked with them to come up with something when there wouldn’t normally be an update. This was beneficial when there were a lot of voices going in different directions. It focused the discussion and made it more productive, rather than simply allowing fifty different lines of argument to continue until there could be an update (which is obviously a recipe for negativity).

My final tool for allowing people to feel listened to was to simply…respond to them. I know this may seem obvious, but you’d be surprised how many professionals just don’t respond to emails or private messages. I tried to always respond (admittedly, at the height of the beta, with my daughter being born and spending a week in the hospital, I probably missed a few). Of course, I didn’t just wait for people to reach out to me. If I wanted a better understanding of where someone was coming from, I messaged them.

No matter how you do it, letting people know that they are being listened to is incredibly important. Because if they feel like they’re just yelling at a wall, they’ll get frustrated, worked up, and then just start yelling at each other.

My next tool was preventing escalation by hyperbole. Escalation by hyperbole is something you see a lot on the internet. One person comes in saying, “Yan Lo is a little weak.” The next person responds, “I’ve never seen him lose!” Cut to ten pages later and one guy claims Yan Lo shoots lasers out of his eyes and levels Tokyo on the weekends, and the other guy is claiming that Yan Lo is the worst model ever made and if he goes to print like this he will melt his models down into a point and jam them into his eye because at least then they’ll be able to deal damage. This sort of progression is terrible for the beta, both because it gets people insulting each other, and because it seriously dilutes feedback. I combated this sort of escalation in two ways.

First, I put a lot of emphasis on battle reports. This kept the discussion on hard facts (VP differences, models killed, etc). Hyperbole is a lot more difficult with the truth sitting in front of you. It also gave either side an “out” if they got into a hyperbole war. They could always just politely say, “Well, I’d like to see the battle report.” Of course, keep in mind, I’m only commenting from a community management point of view, battle reports were incredibly useful for a myriad of other reasons (which made them all the more efficient).

Secondly, I put a stop to the escalation when I saw it, and I never allowed any one thread to carry on too long. Usually, if I locked a thread where this sort of thing was happening, neither party bothered starting a new thread. The battle ground had been closed, and both parties wandered away. Granted, not every thread I closed had this problem, but any thread that goes on too long runs this risk, so I made a habit of pruning them. (This also made it easier for me to read feedback. Again, we’re back to efficiency. If you want to run a public beta by yourself, efficiency is something you’re probably going to want to keep coming back to).

Another great tool was recognizing passion for what it was. Sure, a lot of the time that passion was frustration pointed at me. And on occasion I got frustrated right back, and I had to say sorry once or twice. But, at it’s core, everyone was there for the love of the game. If I could take that passion and let people legitimately help influence the game, I won twice. I had one less person unhappy with me, and one more person helping me.

Finally, I encouraged people to give me feedback privately if they wanted. This allowed a number of people who weren’t comfortable with the forums to have a say. Opening this line of communication was very important. It allowed me to get feedback I may have otherwise missed out on while also preventing other, less productive, communication.

For most of the beta, I tried to maintain my distance. I didn’t want my opinions influencing feedback. If I came in and told people how to use a model, that’s how they would use it. Unfortunately, the players who get that model a year from now won’t have me sitting at their table telling them what to do. So, to a certain extent, I needed to let people struggle to see where the system broke. On the other hand, I sometimes needed to pop in and say, “test this.” Or, “calm down guys, here is the story…” When I worked for the school district I had a teacher who told me, “Everything is a moment in time, and you will need to make a judgment call. You won’t always be right, but you do need to make a call. Learn from it either way.” I definitely heard that voice in my head a lot during the beta.

Of course, there was a lot more to the beta. I didn’t even touch on…game design. Or analyzing data, or managing my time, or, well, a lot of things. But I think this beta was unique in how positive it was, and I wanted to spill some light on how I worked towards that. And, again, most of the credit goes to you guys. I provided an environment where the community could be productive, but you’re the ones who did it.

Anyway, this whole process was a great experience, so I want to wrap this up with a sneak preview of one of the first masters on which I had some say on the art direction and character.

Meet Ironsides

(Click image to enlarge)

Ironsides_Art_Malifaux

 

Sleep Now In The Fire…

Resource Design In Malifaux

resources

Today’s post is on resources, and how to design resource mechanics for Malifaux. By resources, I mean something that will be spent when using an ability in such a way that it limits the use of that ability. For example, discarding a card to take an action, etc.

Fundamental Resources

There are a few resources which I qualify as “fundamental resources.” These are resources which basically every model in the game uses. There are three fundamental resources:

Cards: The cards which make up a player’s hand can be spent on almost any model.

Action Points: The action points which dictate how much a model can do throughout the course of the game. I have a whole post on viewing AP as a resource.

Wounds: The number of wounds a model can take before it dies, which can also be spent by certain models on some abilities.

Without these resources, the game simply wouldn’t function. While using them as a limiting factor for certain actions is common, relying on them too much can make the game collapse (I’ll get into this a bit later).

Secondary Resources

Secondary resources include anything which can function as a resource, but isn’t one of the fundamental resources. This can include almost anything from a condition on a friendly model (Chi on Yan Lo) to a condition you put on an enemy model (Poison for McMourning) to things like scrap, corpse, and scheme markers.

Crew Mechanics

Many crews are themed around a specific resource. McMourning is themed around poison, Colette is themed around scheme markers, etc.

However, a fundamental resource can never be used as a crew mechanic. If a fundamental resource is used, there will be too many other models which interfere with the functioning of the crew. This is something we have tried a number of times before, and it has never been pretty.

For example, Jack Daw went through a period in the beta where he used the cards in hand instead of wounds to stay alive. Basically, he only had one wound but could always discard a card to prevent any incoming damage. This was neat and unique, and a throwback to 1.5 Jack Daw. But it had the severe limitation of Jack Daw having to save his hand to stay alive rather than cheating fate, which hampered the rest of his crew. Additionally, any model which allowed Jack to draw (or forced him to discard) cards caused huge fluctuations in his survivability. For these reasons, we needed to abandon this mechanic.

Similarly, Kirai went through a period of using her wounds as her primary summoning resource. Of course, opposing models were continually trying to take this resource away from her by attacking her, which both made her easier to kill and weakened her summoning. On the other side of the spectrum, run of the mill healing abilities made her summoning too powerful. Again, this caused huge swings in her power level. (Now, to an extent you could say that Kirai still uses wounds as her summoning resource, but I would argue that she is primarily designed to use Seishin as her resource with wounds as an option. But that’s a topic for a Kirai design post.)

Using a fundamental resource as a crew theme resource always causes wide swings in power because so many models interact with them. Additionally, playing with them too much can cause issues for all of the models, not just the ones playing in theme.

Of course, there are lots of models that use cards and/or wounds as a resource, and that’s fine. But it is a question of degree. A single model which has a zero action that requires a discard to function isn’t going to bog down an entire crew. In fact, minor discard abilities help to even out the luck inherent in drawing a hand, as low cards can be used to make these abilities work if you drew a bad hand (or, conversely, if you drew a really good hand, you now have a tough choice about which high card to discard). I like these abilities because they even the playing field a bit. But a master which revolves completely around one of these resources is going to become a black hole that the whole game has to revolve around.

In my experience there are two kinds of resources that work very well as a crew themed resource:

The first are resources that are obtained in a manner that allows the opponent a resist. For example, McMourning can summon after he applies poison to enemy models. Poison is McMourning’s themed resource. However, the enemy models will always have a chance to resist being poisoned. To an extent corpse markers fall under this category as well, as the most efficient way of getting them usually involves killing enemy models.

The second are resources that are totally unique to that crew. Take Chi on Yan Lo as an example. Chi generation is not limited to being gained by killing enemy models (although that is an option). However, because Chi is so specific to Yan Lo, I never need to worry about another model having an ability that grants Yan Lo an insane amount of Chi. Every model that gives out Chi was tested with, and intended for, Yan Lo.

The only exception to this rule that I can think of is Colette. She uses scheme markers as a resource, which are neither specific to her crew nor gained through conflict with the enemy (although I would still consider them a secondary resource). The catch here is that the things Colette does with scheme markers are, generally, a bit less powerful than the things other masters can do with their resources. Resurrectionists can bring in brand new models, Yan Lo can gain all sorts of upgrades, etc. Colette tends to use scheme markers for movement and damage mitigation, areas that other models often have built in. This makes her a very tricky and very unique master. However, it also means that I have paid very close attention to her in the beta, as she took on a mechanic I knew would be difficult to balance. But, I think we pulled it off.

Anyway, those are my notes on how to apply resource consumption in Malifaux design. The easiest way to design a broken model is to make it revolve around one of the fundamental resources.

 

 

 

The Problem With Summoning Part 1

handcomingoutofgrave1

Summoning is a major part of Malifaux, but it comes with some unique issues. There are very few (if any) other table top war games where actively putting new models on the table plays such a major role. Generally, all of your actions are focused on removing the opposing models from the table. While summoning makes Malifaux all the more diverse and unique, it creates some issues which other similar games never need to worry about.

I mentioned in another post that AP should be viewed as a resource. In this view, the winner is the player who most efficiently turns his or her Action Points (AP) into Victory Points (VP). This makes actions which either remove opposing AP, or add friendly AP, very powerful. Removing opposing AP is the more common of the two, and is most easily seen in killing models. The vast majority of models in the game have the potential to remove opposing models from the table (or are specifically designed for that purpose).

Summoning friendly models is the logical opposite to killing enemy models. Both forms of action hinge on changing the AP differential between you and your opponent in your favor. If you’re only looking at the AP differential, there is very little difference between deleting an opposing model and creating another friendly model. So it would seem that, on the surface, these two strategies are inherently balanced. Unfortunately, that’s not the case.

To remove an opposing model, you almost always need range and Line of Sight (LoS). This inherently implies that there is a positioning element, which allows your opponent to react to you; either through positioning (running away) or counter attacking (killing your stuff first). Additionally, the opponent will generally have an opportunity to resist (because actions that can kill opposing models almost always require an opposed duel of some kind). This further adds to the counter play.

Summoning, on the other hand, can be accomplished in a vacuum. There is nothing inherent about summoning which requires positioning or opposing models to pass duels, etc. You can just do it in a corner without worrying about what the opponent is doing. Imagine if Lady Justice could sit in a corner swinging her sword and cheating in high cards every so often while removing opposing models from the table. From an AP differential standpoint, unchecked summoning is just as powerful. And, in addition to being too powerful, mechanics which don’t encourage interaction are simply less fun and more frustrating.

Of course, I am only talking about how summoning plays inherently. There is nothing specific about its design which forces player interaction and positioning, but that’s precisely the sort of problem we go about solving when we design summoning mechanics. This is why summoning actions usually require another resource (such as a corpse or scrap marker) which lessens the power of the ability, adds a positioning element, and forces interaction as the best way to get those resources is for models to start dying. I don’t think any of our wave 1 summoners are too powerful (and their beta counterparts are shaping up), I’m simply noting what I look out for as a designer. There are a number of other creative solutions which I will talk about in a future article, but for now I’m just laying out the problem.

Malifaux Activation Order

Today’s post will be another that takes what we intuitively understand about Malifaux and breaks it down explicitly, along the same lines as the article on Viewing AP as a Resource.

Order of Operations In Malifaux

operations

Mastering activation order is one of the most important skills in Malifaux. However, it can be tricky to get right. Do you activate the model which is most likely to be killed if it waits until later in the turn? Or should you activate the model which is best able to kill an opposing piece before it activates? Maybe neither one is the answer and you should activate the model which is about to place a valuable scheme marker or put up a defensive aura.

In my experience there are four really pressing reasons to activate a model early:

1) To get one more activation before the model is killed.

2) To kill an enemy model before it can activate.

3) To buff friendly models or debuff enemy models that are yet to activate.

4) To score VP (either before the model is killed, or before the method of gaining VP is denied by the opponent).

Remember, winning the game is all about Victory Points (VP) so the best option is usually the one which grants VP in the most efficient way. This would make option four appear to be the most appealing and, while this is often true, it isn’t necessarily always the case. Action Point (AP) efficiency is incredibly important, so often activating a model which will be killed if it doesn’t activate soon enough can be the best choice. For me, the most important model to activate the soonest is usually going to be one which combines a few of these options.

For example, maybe you have a model which will be killed if it does not activate as soon as possible, and that model is also in position to kill an enemy model before it activates (combining 1 and 2). Or maybe the model is in a position to score some VP which may not be available later while also taking an action to generate an aura which buffs the rest of your crew. Look at the board and remember that, once the game has started, all that matters is a model’s positional value. It doesn’t matter how many stones you spent on it or whether or not it’s your master. If it is in the most valuable position, it should be activated first. And, when possible, activate models which fall under more than one of the above categories.

Of course, activating as soon as possible isn’t always the best option in Malifaux. In fact, activating later in the turn can often reap some great benefits. On the message boards, the term “out activating” is often thrown about. What people mean by this is that their crew outnumbers the opposing crew, allowing their last few models to activate without the opponent responding.

In my opinion, there are three prime reasons to hold a model back and activate it last:

1) Activating last allows you to put the model into harm’s way and make its attacks without fear of opposing models activating afterwards and killing it (this is especially good on models which can move out of harm’s way after attacking).

2) Activating last allows you to score VP more easily (for example, placing Scheme Markers near an enemy Master after it has activated when you have the Spring the Trap scheme).

3) Activating last allows you to put some good buffs on the model before moving in and attacking.

Just like when deciding which model to activate first, you need to keep VP efficiency in mind. You may want to hold your master back so that it can attack safely, but if a 4 stone minion is in position to Spring the Trap, it may be the most valuable model you have. And, when you can hold a model back which fulfills more than one of these categories, that is usually a good choice.

It is also nice to see what your opponent does before committing your models. Don’t just look for the best opportunity; try to guess what he/she is doing. Anticipate what they may have planned after initiative is flipped on the next turn and use your late-turn activations accordingly. Does it look like your opponent is set to plant the last marker for Line in the Sand as soon as initiative is flipped? Can you stop it? At this point you are determining the flow of the game and that is very powerful: use it to your advantage.

Once again, this was probably a lot of stuff people already understand intuitively, but spelling it out can often help. Let me know if you found this useful.

Balancing Models In A Dual Faction Environment

Second edition has introduced a lot of cross over between factions; new mercenaries, new dual faction models, and new hiring rules. Not to mention the options that upgrades add. But, with all of the potential combinations, how do we maintain a balanced environment? I think, on the whole, most people agree second edition is a relatively well balanced game, so how did we do it?

I’ll get into a few specifics down below, but the most important thing is a shift in philosophy. The goal is not to create a balanced system by hammering down problems as they pop up like some infernal game of whack-a-mole; the goal is to create a system that is capable of correcting itself.

SCOTT-PILGRIM-Mary-Elizabeth-Winstead-with-her-THOR-Hammer

Some people look at a game system as a machine that requires constant maintenance. But, over the years, you’ll end up with rubber bands holding the whole thing together and duct tape keeping the fender on. I see a game system as a living thing. My goal is to give it the tools it needs to survive and then let it correct itself.

So, when we go about balancing Malifaux with all of those models switching factions, everything needs to be looked at through that lens. Accept that the system is larger than you. Accept that you won’t catch everything. Trying out every single potential combination is simply impossible unless we want to beta test for years. Instead of worrying about the whole, we worry about the parts. A healthy system is created from the ground up, so first we want to make sure that models are balanced individually. Maybe, during the beta, nobody ended up testing a crew with Seamus, a Showgirl, Killjoy, and Taelor all together. However, even though that specific list wasn’t tested as a whole, so long as each of the models in it is balanced, we will probably be fine.

This is the first lesson: balance models individually. We can’t balance the game by saying, “Well, Resser masters are weaker than those of other factions, so we can give them stronger minions.” (This is an example I am making up, by the way, Resser masters are fine). If we follow that line of thinking, what happens when other factions start hiring Resser minions? Similarly, we couldn’t say, “Arcanists have weak 4 stone options, so we can give them stronger 7 stone options.” (Again, making this up). This has the exact same issue. What happens when Arcanists can hire 4 stone models from other factions? What happens when that 7 stone model ends up elsewhere? Not to mention, this would probably just lead to Arcanist players spamming 7 stone models.

I’m not saying that the factions need to lose their identities; they all still have distinct areas of strength and weakness. Guild have more range options, Ressers have more summoning, Neverborn have more movement, etc. What I am saying is that each model in a faction needs to be balanced as an individual. If all of the parts are working, the crew you create from those parts should work as well.

Of course, this doesn’t cover everything. We’re all gamers; we’ve all seen what happens when there are two models/cards/whatever with rules that work individually but, when combined, create some stupid infinite loop or other silliness. How do we avoid that? Well, models which are capable of being broken like that tend to have some warning signs. As a designer, you need to learn to recognize those warning signs and work them out even if you haven’t found a specific model which causes a broken combo. Generally, anything which copies an action from another model can prove problematic, so we keep a very close eye on these abilities. That’s why they frequently restrict the use of copying abilities which mention a model by name or using triggers. Anything which adds suits can be problematic as it can result in unlimited triggers. Models with these abilities are usually closely watched (for example, the Daydream can add suits but, being the Dreamer’s totem, we know exactly which model he is adding suits to. Somer can add suits, but he can only do it to gremlins and pigs so we don’t need to balance all mercenaries around this ability, etc).

Finally, the last thing to watch is simple language. Anything which is too open ended can end up being a problem. Let me give you an example. Here is Shenlong’s Burn Like Fire action from the beta a week ago:

(1) Burn Like Fire (Ca 6 / TN: 10 / Rst: Wp / Rg: 6): Target enemy model suffers 1/2/3 damage. If the target has a Condition, this model may choose to gain that Condition.

Here is his Burn Like Fire action from this week:

(1) Burn Like Fire (Ca 6 / TN: 10 / Rst: Wp / Rg: 6): Target enemy model suffers 1/2/3 damage. End all instances of the Defensive, Focused, Fast, and Reactivate Conditions on the target. This model gains all Conditions ended in this way.

See the difference? This week’s version is much more restrictive, spelling out exactly which conditions he may gain. Now, did I find some broken combo involving this? No, but I acknowledge that there very well could be one I’m not seeing, if not now then in the future. Keeping the wording tight keeps the game healthy.

The other thing which you really need to watch is the number of debuffs available to a crew. Although watching the whole system is impossible, you can track how many models have an ability capable of debuffing (lowering the stats of other models around them) that may be taken outside of their native faction. If a crew ends up with too many of these models, it can make for an unfun game. This is why Montresor lost mercenary, and Iggy is being watched very closely.

If you follow all of these rules, you should have a relatively healthy system. But I mentioned earlier that I see a system as a living thing, and a living thing is going to need some tools to survive. The final piece of the puzzle is creating a few answers to potential problems which are available to absolutely everyone. This is where the mercenaries really shine. Each of the named mercenaries adds something unique which can be very useful from a meta perspective. Let’s have a look at them.

Johan, in addition to being an efficient model, has Rebel Yell which can remove conditions from a model. Conditions are a huge part of second edition, and it’s very possible that a player could run into a situation where they really need to get rid of some of them. Having a model available to every faction with this ability really helps level the playing field.

Taelor has Welcome To Malifaux (my personal favorite action that I designed in Wave 1, just as a side note) which gives her some good anti-summoning tech. Summoning is another big part of second edition and I think it’s necessary for all crews to have an answer to it. Keep in mind, this isn’t something we put in to answer a specific problem. We didn’t look at the meta and say, “Nicodem’s summoning is too powerful, so we’ll let everyone use Taelor. Problem solved!” That would be terrible design. However, we acknowledged that playing against a summoning crew can be a big shift in thought for people new to Malifaux (after all, most wargames only remove models from the table, they don’t come back on). And, in a small playgroup where a player may end up facing the same crew over and over again, it’s good to give that player an answer if they are getting frustrated. It’s also good from a tournament perspective if a player is familiar with the meta in their area; everyone always playing summoners? Pest control is on the way!

Hans can snipe upgrades off of models. Now, he is definitely the most controversial of the models listed so far. People seem to be worried about his ability to remove upgrades from masters (since upgrades are usually a big part of a master’s playstle). While this is a valid concern, we were aware of it when designing Hans, and Leader models can mitigate this ability by discarding cards. I have never personally seen a Leader have an upgrade removed by Hans unless they felt the cards were more valuable (in which case, they weren’t too attached to the upgrade anyway). People are more worried about the idea than about how it actually functions. However, perception is important, and with Hans filling this role, I have no intention of adding more models with this capability. Alright, controversy out of the way, Hans plays a role just as important as Johan and Taelor. Upgrades are a big part of second edition, and it’s good to have an answer to them to shake up the meta a little bit. Hans was never designed to take upgrades off of Masters; he was designed to take upgrades off of enforcers and henchmen (and he’s good at it). Like the other two, Hans isn’t an answer to a specific problem. However, if a player finds themselves facing an opponent who takes the same list over and over again, he can be used to shake things up. Sick of your friend always taking Decaying Aura on Bete Noire? Hans is your man.

That’s all I have for right now. Just remember the goal. The goal is not to create a machine and keep it running forever, the goal is to create a living thing that takes care of itself.

Swimming Against The Tide

swimming-against-the-tide

Comparing two models can be tricky, because when you do it, you need to consider everything the model can do and not just a single ability. It’s human nature to pick the two things which are most alike and compare them side by side; however, this is dangerous from a game design stand point.

This has been cropping up with Leveticus in comparison to the hanged. The Hanged has an action (Whispers From Beyond) which is very similar to the Unnatural Wasting action on Leveticus’s Pariah of Iron upgrade, except The Hanged doesn’t need to make a damage flip where Leveticus does. This led to a number of people requesting this action get a buff just because it wasn’t as good as the one on The Hanged. So, I posted this in the Leveticus thread:

I’m not worried about Whispers From Beyond on The Hanged. The Hanged being able to do something better than Leveticus does not make Leveticus worse. Nekima has a weak melee damage of 4, but that doesn’t invalidate Lilith as a melee piece. At the end of the day Leveticus gets one more AP to use the Action, has a longer range, and has a higher Ca with the opportunity to cast it on a +. Also he tends to be able to last longer to get more use out of it. Of course, all of this is academic, because I am not concerned with comparing Pariah of Iron with the Hanged; I’m concerned with comparing Pariah of Iron with Pariah of Bone.

Comparing models is difficult, because a single aspect of the model can’t be cherry picked and compared to something else, the model needs to be considered as a whole. But that’s not how people tend to operate.

At the end of the day, a good model can have a bad ability; it doesn’t make the model any worse, it just won’t come up that much. However, sometimes when a good model has a situational ability, that causes people to lower their perception of the model as a whole. It’s an unfortunate quirk of human nature, but since we tend to design games for humans, adhering to human nature is pretty important.

A good example of this was the Hunter ability on Samael Hopkins in the wave 1 beta. Hunter allowed Samael to ignore cover while within 6″ of his target. This is situational enough, but he also had Visions of Flame which was a better ability so long as the target had burning. However, Hunter may have come into play every once in a while. Because Hunter wasn’t very appealing, people got stuck on that and under valued Samael as a whole. All sorts of replacement abilities were suggested (which would have made him broken, since the rest of him was perfectly fine). Ultimately, I just removed Hunter and didn’t replace it with anything. Once the sub-par ability was removed, people focused on the better aspects of the model and I haven’t heard a single complaint about him being under powered, despite the fact that he is technically a smidge less powerful than when he had Hunter.

Human nature is an awkward thing to work around sometimes, but fighting against is is like swimming against the tide. Human nature wins.

 

Viewing AP as a Resource

Today I will be talking about viewing Action Points (AP) as a resource. Doing so doesn’t change much, but it helps to explain logically what most players already understand intuitively. Spelling things out like this can sometimes help point us in the right direction when we have a difficult in game decision.

When selecting your crew, you have a number of AP which will be available to you throughout the game. An average crew crew can probably get about 85 AP over the course of an average game (8 models x 2 AP = 16 +1 AP for the master = 17 x 5 turns = 85). This varies depending on the number of models, additional actions the models can take, etc. It’s good to know about how many AP your crew can generate in game.

The point of the game is to collect Victory Points (VP). So, ultimately, the winner of the game is the player who most efficiently spent his or her AP to gain those VP. The more AP it costs you to get any single VP, the less viable that objective is. For example, an objective that requires you to cross the board is good for a crew with a high walk stat (as they will use fewer AP to cross the board), but another objective that involves killing models may be less appealing to the crew if they deal low damage (as this objective costs more AP for the crew). This may seem obvious, but as I said earlier, this is mostly an exercise in expressing logically what most people grasp intuitively. “Keep to your strategies and schemes,” is common Malifaux advice and what people are essentially saying is, “make sure you accomplish as many VP as you can with as few AP as possible.”

Of course, the number of AP a crew generates in a game is not a set number. While 85 might be the maximum potential for most crews, this number is going to go down as the crew loses models throughout the game. This brings me to my next point: not all AP are created equal. If a model is very survivable, it is much more likely to survive the game, and thus it is likely to be able to generate every AP possible. Because of this, survivable models will generally have less potent actions.

On the other hand, a model might be very good at causing damage.

“Broke into the wrong goddamn rec room, didn’t ya you bastard!”

Rec Room

In this case, the model can deny your opponent AP by removing their models. Additionally, in kill-based objectives, the model can generate VP at a more efficient rate.

During the transition from 1.5 to second edition, one of the major shifts which we made was to reduce the number of specific AP in the game. Specific AP comes from things like melee expert which allows the model more attacks, or nimble which allows it a free walk action, etc. These abilities were found on the vast majority of models with a cost of 6 or greater. However, these models were generally tough, good at killing things/accomplishing objectives, and they generated extra AP every turn on top of it. One of the disadvantages of fielding a crew of nothing but the most expensive models is generally that you will be getting fewer AP every turn, although the ones you do get will be more valuable. In 1.5, most of the models could both generate numerous AP and use them very effectively (and the expensive models which did not have specific AP very rarely saw the table). By reducing specific AP in second edition, crew selection has become much more difficult. People generally want a few heavy hitters, but they need some good objective runners as well. Hard choices are a good thing for a game.

Of course, we didn’t want to swing the other way either. In the early M2E beta test, there was a great amount of concern about “swarm” lists. In other words, lists which ran as many cheap models as possible in order to maximize AP. However, this is generally balanced by the quality of the AP those models generate. And the key faction which is best at doing this (Gremlins) generally pay for their extra AP by way of damage. While this mechanic is very fluffy for the gremlins, it also serves a very important balancing function; when gremlins suffer damage to generate more AP, they are essentially risking the AP they can generate later in the game in return for AP on the current turn.

Abilities which increase a crew’s AP potential during the game are incredibly powerful, so they are usually relegated to Master-level actions. The most common form of this is summoners who can spit out a number of smaller models to accomplish various objectives. These often prove some of the most difficult abilities to balance. The earlier in game the model is summoned, the higher its maximum AP output becomes. This is why there are often restrictions on summoning which make doing it in the early game restrictive or more expensive (i.e. requiring corpse markers which won’t drop on their own until later in the game, or in the case of Molly having to be near the enemy). Coming up with mechanics which restrict the early game more than the late game can be very difficult and is one of the reasons summoning is hard to balance. Another tool we use to balance summoning is having the summoned model take damage, meaning it is less likely to achieve its full AP potential.

Finally, there are even ways to transfer AP from one model to another (for example, with Obey you can spend one AP with one model to allow another model to take an action worth one AP). This is very useful if you have a big, heavy hitter with whom the AP would be more valuable. In other words, the heavy hitter has a limited number of AP, so other models funnel theirs over to him.

Of course, Obey-type actions aren’t just used to give AP to heavy hitters. They can also be used to allow models in critical locations to take interact actions or drop scheme markers. This brings us to another important aspect of viewing AP as a resource: positional value. We have already established that not all AP are created equal, but up to this point we have only explored how this is true because some models have better stats than others. However, the model’s position on the board also plays a part. A bayou gremlin may be a cheap model with low stats, but if it is in position to place the last Scheme Marker for Line in the Sand and score some VP for you, it may be the most valuable model in your crew. In other words, the value of a model will change as the game changes, and this is something which I call positional value. Again, this is understood intuitively by most players, but it can be useful to spell out.

This is another tricky thing about balancing summoners. Frequently, summoners place models right in the middle of the action, increasing their positional value. A normal model might need to take three walk actions to get to the center of the board, but if a model is just summoned right there, it has essentially gotten 3 AP worth of positional value right off the bat. This is a big reason why summoned models are slow and may not interact the turn they come into play.

A lot of players talk about a model’s effectiveness by referencing its soulstone cost (i.e. “My 8 soulstone cost Coppelius held up 13 soulstones worth of models. Pretty good!”). While this isn’t a bad rough estimate of a model’s effectiveness, the thing which I place the most value on is its AP efficiency; how many AP does the model need to spend to get me a VP? While Coppelius in the example above may have done a good job, he’s not going to win the game for you if your opponent has a 2 soulstone gremlin running around throwing scheme markers everywhere. In other words, don’t ignore the positional value of models. Once the game has started, how many soulstones the model costed you is totally irrelevant; all that matters is VP.

I hope this wasn’t too dry or too obvious for you. Leave a comment and let me know if you found this post interesting, and what sort of posts you would like to see in the future.