I’ve been brainstorming the whole “attribute + skill” idea from last week and even though it doesn’t really fix the problems created by a freeform skill list, it does introduce the idea of encouraging players to think about how their characters are doing something instead of just what they’re doing. It’s kind of like the rule in newer versions of D&D where some weapons get a bonus for Strength and others get a bonus for Dexterity. My biggest reservation is that I don’t see an easy way to just add the idea to the existing rules. Any new bonus that gets applied to all or nearly all rolls is going to change the roll ranges and add new complications, so sticking attributes or something like it is going to basically mean yet another redesign of the character creation system and basic dice mechanic. Since this is still very much brainstorming, this week I’m just going talk out loud (and maybe incoherently) about some of the ideas that I’ve been tossing around. 

Why Attributes? 

The thinking behind leaving out attributes in the first place was that it was unnecessary because a character’s competence in fiction is more often a factor of the character’s role in the story (Hero Factor) than natural ability. So the assumption was that all characters are slightly above average in all of the things you would normally describe with attributes and the players who wanted their characters to be better or worse could do that with Edges and Drawbacks. But realistically, very few players choose “strong” or “smart” or “charming” or whatever when they’re creating a character. This was also true of the QAGS idea that Body, Brain, and Nerve were holistic and you could describe variations (a character who’s strong, healthy, and agile, but ugly or weak, sickly, and clumsy but good looking) with Flaws and Skills. Even if they did, when you think about it, it’s an even bigger “broad skills” problem than something like “Fighting” or “Educated.” Also, characters in fiction aren’t just defined by whether or not they’re a protagonist. Sure, Hero Factor goes a long way, but most characters are heroic in a specific talent department that’s not necessarily a skill: Tony Stark’s intelligence, The Bride’s fighting abilities, Ferris Bueller’s charm, whatever. 

There are also some practical reasons to re-introduce the idea of skills. For one thing, it’s kind of awkward to write rules when the players are constantly told to vaguely “make a roll,” meaning that the players roll d12+Hero Die + whatever Edges apply. It’s just seems weird because we’re used to “make a strength roll” or “make a history roll” or whatever, but you can’t really say that when strength and history are just two possibilities from an non-existent and infinite list of freeform skills. Aside from avoiding the awkward “make an appropriate roll” constructions in the rules, attributes can help prevent situations where everyone’s making the same roll (since they all have the same Hero Factor and nobody has relevant skills) whenever everything about the situation screams that certain characters should have a better or worse chance of success than others. Tyrion Lannister should have a better chance than Podrick or realizing that someone’s trying to lure him into a trap, but if we make the (admittedly wrong) assumption that they have the same Hero Factor, Tyrion only gets a bonus if he has some appropriate Edge like Detect Lies or Read People or something. A perception-type attribute could fix that problem. 

I’ll circle back to this when I start talking about rules changes (possibly next week), but you may have noticed that a lot of the Trademarks are kind of a roundabout way of creating characters with high attribute scores. Brawler is a character with a high Strength, another is high Dexterity/Agility/Whatever, there are several variations that are high Brain or Charisma, etc. A few (like Monster Hunter and A Very Particular Set of Skills) don’t fit the bill, but a lot of the existing Trademarks are really just high attribute scores in disguise. I didn’t realize it when I added Trademarks, but if you look over the list of sample Trademarks it’s pretty obvious. 

What Attributes? 

If I decide to create an attribute-like trait type, the big question is “What are the attributes?” Before getting into how detailed the attributes should be (Is there a holistic Body attribute like in QAGS, or multiple attributes that describe physical ability, like Strength, Dexterity, and Constitution in D&D?), there’s a question of what the attributes are defining. I think there are three main categories to choose from:

• Traditional: These are the kinds of attributes most gamers are used to, where each attribute kind of describes the character’s natural ability in whatever areas the game designer thinks are important: Strength, Reflexes, Charm, Endurance, Perception, and so on. The biggest advantage here is that it’s familiar and checks off all the “why have attributes?” boxes I talked about before. 
• Super-Skills: In this scheme, the attributes are more like broad skill groups than raw talent. So you might have things like Combat, Mechanical, Covert Ops, Magic, or whatever. One pro of this option is that you can customize it to the setting: a fantasy game has a Magic attribute, cyberpunk has Hacking, etc. The biggest con is that since some characters would zero in some Attributes, the most logical way to do it would be “Pick X Attributes,” which removes the standardization and default roll perks of having attributes in the first place (and may necessitate rules for dealing with a zero in the attribute). It’s also less clear-cut which attribute applies (is lockpicking Mechanical or Covert Ops?), which could be a good thing or a bad thing. 
• Style: This is somewhere between the two previous options, with wishy-washier traits like Forceful, Sneaky, Athletic, Precise, Charming, etc. I like it because it’s the most straightforward way of making the attribute about how the character is doing the thing: Intimidation is Forceful, a well-reasoned argument is Precise, flattery is Charming, etc.  That’s also a con because putting what stat to use in the players’ hands has an equal chance of leading to creativity or ridiculousness depending on the players. It also, once again, removes the standardization that makes attributes handy from a design perspective. And it's kind of a broad stroke approach that can be weird if you’re not playing a genre where broad strokes work (The character with a high Sneaky has a pretty good chance for rolls involving petty theft, international intrigue, sneak attacks, or computer hacking). 

There’s also the option of mixing and matching the three categories (like Star Wars d6, which had some traditional stats mixed in with things like “Technical” and “Fighting”), which is very tempting. With clear definitions and rules for dealing with overlap, it could be set up to allow for some of the benefits of all three options. Of course, it could also result in a huge, confusing mess, but at this point the whole system is a huge confusing mess. Luckily I’m still managing to convince myself that working through the messes is the key to making the game work the way I want it to. This may be an indication that my brain is broken. 

After I realized that the path I’d been following with specializations didn’t work, I started considering alternatives. One idea (with many possible variations) that kept coming back to me was mixing fixed and player-generated descriptors: A small group of very broad fixed skills at the top level, with freeform narrower skills below them. A tweak on the general idea that I also really liked was to derive the stat for the big skill from the smaller skills. A commenter in the playtest group suggested something along the same lines (his suggestion was to limit the Edge number based on the number of Tags). The common theme of thinking of things from the opposite direction (starting narrow and going broad) seems worth exploring. 

Before we start, let’s cut out the third level of specialization. Even in the previous examples, they were often either arbitrary or meaningless. For example, how do we differentiate between Swordsmanship (Inigo’s Tag) and Fencing (Inigo’s Focus)? Unless our combat system is extremely detailed, we mostly take the player’s word for it. There might be a few situations where the GM can say “that clearly doesn’t count as fencing,” but they’re going to be rare.  Therefore, there’s no need for mechanical differentiation, assuming that example applies across the board. 

What about a skill with more clearly-defined degrees of specialization? For instance, there is a clearer differentiation between History (Edge), 20th Century History (Tag), and World War II History (Focus), right? Sure, there are some ambiguous cases because history is messy (does the Focus apply to events of the time period that aren’t obviously related to the war?), but if you’re rolling to see what your character knows about D-Day, it’s clear that the Focus applies. The next question is, from a game mechanics standpoint, should we care? I’m not sure we should. From a design perspective, there’s a slippery slope, with each level of specialization begging the question of why there’s not another level of specialization. If we drill down to World War II History, why allow a further specialization in the European Theater, further specialized to Operation Barbarossa, then even further to the Siege of Leningrad? From a fiction perspective, we usually don’t care. We’re going to expect the same kind of knowledge from a historian in a book or movie regardless of whether they’re introduced as “Professor of History,” “World War II Historian,” or “Siege of Leningrad Expert.” The generic Historian knowing an obscure fact about Leningrad or the Leningrad Expert knowing an obscure fact about the French Revolution isn’t going to ruin our suspension of disbelief. 

So let’s go back to my “The character sheet is not the character” mantra and remember that we don’t need to quantify everything in mechanical terms. After a certain point, specialization is a matter of characterization, not stats. The differences between our historians is a matter of roleplaying: The Leningrad expert’s office has more books about Leningrad than the generic historian, he’s more likely to be a member of the Siege of Leningrad message board, and if a situation reminds him of a historical anecdote, chances are good it happened in Leningrad. The GM can give the player things like automatic successes or situational bonuses for specializations if she feels they’re warranted, but those are basically cases of GM fiat that don’t need a special mechanic. 

Then scheme I’m currently thinking of is probably going to require some changes to terminology (and possibly to other parts of character creation), so for now let’s just call them Level 1 Traits (the artist formerly known as Edges) and Level 2 Traits (kind of corresponding with Tags). Since the L1 Traits are fixed, I want to keep the number of them limited, so they’re going to need to be broad, effectively fitting into the “Attribute” slot that so far Cinemechanix has avoided (fortunately, I think this method avoids some of the problems that caused me to leave out Attributes in the first place). Level 2 Traits are freeform, but the assumed level of specificity is similar to what I’ve always thought of as the “default” breadth of Edges: History, Occult Lore,  Swordsmanship, Firearms, Breaking and Entering, Persuasion, etc. There are a few ways to “enforce” this default without forcing the player to choose from a list:

• By Example: Every setting should have its own list of sample traits, and as I mentioned last week most players will choose a lot of their skills from these lists. I can use the list to set the tone. 
• GM Veto/Forbidden Traits: The GM can veto any trait that she considers too broad, and there should be a section in the GM section with guidelines, warning signs, and examples. There may even be a list of expressly forbidden overly broad traits. Also, all L2 traits have to be more specific than the L1 traits (which shouldn’t be hard). 
• No Descriptor Fuckery: If a trait is a more specialized version of a default-level (example) trait, the character gets the bonus in the same situations he’d get the bonus for the default-level trait. So the Leningrad Expert would add his full bonus to any history roll, the fencer would get his full bonus for all sword-fighting rolls, etc. If the relationship isn’t clear-cut, the GM still award a partial bonus. I’ve got a feeling that developing the skills system will suggest more specific guidelines for this general concept. 

So without trying to figure out the L1 Traits just yet, the basic idea is that players buy their L2 Traits using character points and assign them to an L1 Trait. The number of L2 Traits tied to an L1 Trait determines the character’s score in the L1 Trait. When you make a roll, you only get to add a single L1 trait, but you can stack L2 Traits (including those that aren’t under the L1 Trait you’re rolling). Let’s assume that our L1 Traits are “Fighting” “Knowledge,” and “Athletics,” that each point in an L2 Trait gives you 1 point in the relevant L1 Trait. For the time being, we don’t care what the bonuses mean. Here’s a sample character:

Fighting 3

• Swordsman +2
• Firearms +1

Knowledge 1

• History 1 

Athletics 1

• Acrobatics 1

Here are how a few rolls would play out:

• Shoot the bad guy: Fighting + Firearms (+4)
• Identify what kind of gun the bad guy is using: Knowledge + Firearms (+2)
• Name the members of Motorhead: Knowledge (+1)
• Swing across the room on the chandelier: Athletics + Acrobatics (+2)
• Flying kick the bad guy: Fighting + Acrobatics (+4)
• Answer a question about 15th Century fencing: Knowledge + History + Swordsman (+4) 

The Attribute + Skill construction isn’t new (it’s used in d6, Storyteller, and lots of other systems), but I’ve always like the Storyteller-style version where the skill wasn’t tied to a particular attribute. Feng Shui has section that describes what a skill means where it specifies that buying a skill includes several elements like ability (you can do the thing), knowledge (you know about the thing), connections (you know other people who are into the thing), and probably some others I’m forgetting. The unlinked version of Attribute + Skill fits that perfectly: When you want to shoot a gun, you roll Fighting + Guns; when you want to identify a gun, you roll Knowledge + Guns; when you want to buy a Gun, you roll (let’s say) Social + Guns. This gets rid of the overdefining of skills that some systems do (Firearms only covers shooting; you need Repair to fix your gun when it jams and Connections to buy a gun semi-legally). 

My version allows the benefit of the unlinked variation of Attribute + Skill construction, but having them linked in another way (Skills determine Attributes) allows players to use their skill selection to help define their character by their choices of where to put particular skills. If you put your Baseball under Athletics, you’re a player. If you put it under Knowledge, you’re a fan. Making Attributes a derived stat based on skills also gets rid of my biggest complaints about attributes: the tendency for them to overshadow other traits or result in counter-intuitive characters who have enormous raw talent (a high attribute) that they never develop (the player doesn’t buy any Dexterity-based skills because he’s got such a high Dexterity he doesn’t need them). 

There are a lot of details that need to be worked out in order to turn this basic concept into a usable addition the rules, but I like the idea enough to keep playing with it until it either works or hits a dead end. 

The original reason for adding Tags and Focuses (specialization) to the rules for Edges was to allow character to better describe their characters with more specific traits without being penalized compared to players who go with more generic skills. After a few modification, it seems to be working. The character who wants to be a master swordsman can now use a Tag and Focus to get a slight advantage in a swordfight over the character who puts the same number of points into a generic Melee Edge in exchange for a slight disadvantage if they’re using another weapon. 

So far, I’ve been assuming a certain hierarchy of Edges, Tags, and Focuses. Specifically, the examples I used last time assumed a broad Edge (Melee Combat) with more specific Tags (Swordsman/Unarmed Combat) and even more specific Focuses (Fencing/Wrestling). That works fine with a set trait list for each level, but if players were required to pick from a list, the question of broad vs. specific Edges wouldn’t exist in the first place. I could just build the list so that all of the Edges are “balanced” (whatever that means). With freeform trait selection, do Tags and Focuses just spread the confusion over three levels instead of one? 

Player A chooses “Guns-->Pistols-->Colt Desert Eagle”

Player B chooses “Ex-Marine-->Combat-->Guns” 

Well, shit. Assuming both players put the same number of dice into their Edge, Player B gets a lot more bang for the buck. There are probably a few cases where the GM might deny a bonus to Player B but still allow Player A at least the basic bonus for the Edge (for example, if they’re using something the Marine obviously wouldn’t have training in, like black powder rifles, or something), but those cases would be rare. Most of the time, Player B gets the same bonus with any gun he picks up that Player A gets only with a specific model. On top of that, Player B gets the Tag bonus for brawling, knife fighting, and other forms of combat and an Edge bonus for anything he can convince the GM was part of his Marine training. 

It looks like the specialization rules are a failure, but that’s partially my fault. This whole time, I’ve been thinking in terms of “broad vs. specific,” with the assumption that more specific was better, but if you go back a couple of post you’ll find that the I started out using the phrase “more relevant to the character.” If we take that to mean “which Edge (and attached Tag/Focus) tells us more about the character?” Player B looks a lot more coherent. We know he’s veteran (probably a combat veteran) who’s either naturally a good shot or keeps in practice. All we know about Player A is that he likes guns enough to have a favorite. Of course, that might not always be the case, and there's no guarantee everyone will choose specialization. Here are three monster hunters built with 30 points: 

Monster Hunter A:

Fighting Monsters 2→ Guns→ Shotguns

Occult Lore 1→ Monsters→ Werewolves

Investigation 1

Monster Hunter B

Monster Hunting 3

→ Combat → Guns

→ Monster Lore → Werewolves

→ Investigation → Corpses 

Monster Hunter C: 

Monster Hunter 6 

Assuming Hero Factor 4 and a Trademark Die for both players, Monster Hunter A would roll d20+d10+d4 to shoot a monster with a shotgun. Monster Hunter B would roll d20+d12+d4. Monster Hunter C would roll d20+d12+d6. Pretty much the opposite of what happened when I was looking at one skill where everyone more or less played fair.

Well, again, shit. I can come up with a few ways to maybe incentivize Monster Hunter B to act more like Monster Hunter A (limiting the number of Tags per Edge, requiring a minimum Edge rating for a Tag, etc.), but none of those prevent the player for just going with boring-ass option C instead.  

So, the new rules don’t solve the problem. I’m not convinced they’re completely useless (they still fix the Princess Bride problem we started out with, I think), but they’re just as exploitable as Edges with no specialization. The real problem, of course, is that freeform trait selection is by its very nature exploitable. Fixing the skill list would fix the problem, but I’m awfully attached to letting the players choose their own. The way I see it, there are three main options here: 

• Ditch Tags and Focuses and go back to just Edges. 
• Keep Tags and Focuses because they’re kind of useful even if they don’t fix the problem they were supposed to fix. 
• Come up with a fixed list of (probably relatively broad) Edges and let the players come up with their own Tags and Focuses. 

The first two will work fine with a GM who’s willing to step in and veto overreach, especially if there’s a solid sample list to choose from. In my experience if you give players a list, 90% of PC traits will come from the list. I used to think of that as a bug, but maybe it’s a feature. The third option is better from the “let the game system do the work” perspective, but feels overly restrictive to me. There’s a fourth option of fixing the options for Tags and Focuses as well, but I can’t imagine choosing that one. 

Right now I’m leaning toward option 2, but it’s entirely a gut reaction because I like the idea of specialization now that I’ve been writing about it for a few weeks and don’t like fixed trait lists. Objectively, though, there are good arguments for all 3. I might have to keep talking may way through it here to make a decision. Whether I do or not, I’d love some feedback. If you’ve got thoughts, let me know on social media or the playtest group. 

Last week I talked about the specialization rules for Edges that I’m hoping to start playtesting soon. Just as a reminder, there are 3 levels of Edges: The Edge itself, which gives you a number of die bonuses equal to the rating and cost 5 points; A Tag, which gives you an extra die bonus and costs 2 points for the first, 4 for the second, etc.; and a Focus, which lets you add a bonus die equal to your (unmodified) Hero Die and costs 1 point (and is limited to 1 per Tag). The idea behind specialization is to reward (or at least not penalize) players who take Edges that better describe the character rather than Edges that are more useful. There were two lingering questions from last week: (1) Does it just encourage a different kind of power gaming; and (2) Do the specializations actually solve the problem if you keep skill selection freeform? 

In order to attempt to answer the first question, let’s look at the die rolls. The table requires some explanation: 

• I’m assuming here that everyone has 15 points to put into their Edges. 
• Inigo/Fezzik are our specialists, and have focused on fencing (Melee 2→ Swordsman→ Fencing) and Wrestling (Melee 2 → Unarmed Combat→ Wrestling), respectively. 
• Westley #1 is an all-purpose badass who puts all his points into Melee Combat. 
• Westley #2 has decided to specialize in both Fencing and Wrestling Tag/Focus pairs for his Melee 1. 
• Westley #3 had decided to power game and figured out a way to get both specializations while still getting 2 Edge bonuses. He took Melee 1→ Swordsman→ Fencing but also took Athletics 1 → Unarmed Combat→ Wrestling. Since bonuses stack, assuming he can convince the GM that Melee Combat applies to Wrestling (obvious), Athletics applies to Fencing (reasonable), and Wrestling applies to hammer fights (maybe the biggest stretch), he gets to add the bonus for both Edges to all rolls and gets the benefits of Tags and Foci when fencing or wrestling. This requires 1 point more than the players have, but Westely #3 is a weasel so if he can find that point somewhere he’ll just cheat and hope the GM doesn’t check his math. 
• The “Hammer Fight” entries show how the characters do in a fight where their Tag and Focus doesn’t apply. 

The first two Westleys work out about as expected when compared to the Fezzik. The generalist (Westley #1) has a slight advantage in a fight with anything but swords and wrestling, but suffers about a 3-point disadvantage when he’s fighting a specialist with their chosen weapon. The double major (Westley #2) has an advantage over the generalist, but still isn’t as good as the characters who chose to specialize in just one thing. He’s also at an disadvantage when he gets into a hammer fight with anyone. The optimal use of character points is to specialize in one thing, which is what I was hoping for. The weasel (Westley #3), as expected, is a problem. He’s evenly-matched with the specialist and has a marked advantage over the other Westleys. 

Looking at the numbers, the benefits of specializing are significant enough that I don’t think we need a “price break” for the first specialization. You get higher rolls with a specialization than you do with an extra point in the main Edge. Also, there’s no need to increase the price for additional Tags (and Focuses); The dwindling benefits compared to someone who spends the same number of points to specialize de-incentivizes specializing in a bunch of stuff within a single Edge. After a certain point (if my quick math is correct, 3-4 specializations, depending on Hero Factor), the guy with a bunch of Tag/Focus pairs even loses ground to the generalist. Also, as Westley #3 shows, you can get around the escalating cost for multiple Tags by choosing overlapping Edges and/or having a weak-willed GM. 

So how about we make the cost of a Tag/Focus pairing the same 5 points it costs to increase your main Edge rating by raising the cost of both by 1 (3 for a Tag, 2 for a Focus)? That makes the Tags and Focuses more about characterization and less about saving a few character points. It also helps solve the question of how to fit Tags and Focuses into the quick-start character creation rules, since you can just trade in one of your points for Edges for a Tag/Focus pair. 

That took longer than expected, so I’m going to save the 2nd question for next week.