skill specificity; in which I come full circle

So how general should a character’s skills be? This is a battle I have fought on ever side of but I think I’m getting close to an answer I like.

Traveller (Tabletop Game) - TV Tropes

At first (say 1978, playing Traveller) I would have said “very specific” because that felt realistic to me. I suspect I felt something else I didn’t have words for yet as well, but I would. But skills were so specific in Traveller that there wasn’t even a general rule for determining success — pretty much every skill had its own paragraph describing how to use it. That’s mighty specific!

As I matured, or at least grew older, and started to think about game design I started to appreciate the verisimilitude of generalization — or at least categorization. How much difference was there really between a submachine gun and an automatic rifle? Did they really need different skill? Even if they were different, don’t you think you could pick up the skill pretty fast if you already knew something similar? So at this point I was digging the Traveller Mercenary (Book 4) abstraction (available only for military careers though, so in a sense this is just a bucket of skills with a new name) of “combat rifleman”.

As time and experience moved forward I came to really like generalization. In degrees. Eventually I’d be very happy with “Violence” as a skill that subsumed all of that combat shit. Sure, there are different skills involved in real life usage of fists, a knife, a rocket launcher, but verisimilitude was no longer my goal, at least not through strict simulation. I felt (and still feel, for these games) that if you the player want your character to be great at rocket launchers and shitty at barfights, then don’t get into barfights. Or don’t use Violence when you’re in one. Or use a rocket launcher. That is, you can fabricate this background detail of specificity yourself. The machine doesn’t need to know it in order to crank out resolution effects.

I still like this though it’s no longer an axiom of taste but rather just another kind of game I like to play. I’m long past claiming only one game or one kind of game will satisfy me. There are too many cool things out there to sit in a corner with that one you liked when you were 12 and deny all others.

So recently we’ve been playing some classic Traveller. Here’s my character:

Brad (Mickey “The Wrench” Doberman)

Mickey is distinctly below average in all respects, but remains the hero of his own story.

UPP: 465766

Cutlass-1 Brawling-1 Mechanical-2 Ship’s Boat-2 Vacc-1

TAS Membership, Pinnace to call my own (The Stephen Foster)


That’s not a lot of skills. You’d think it would benefit from a ton of generalization. But you have to look at all features of the character, especially skills, in ways other than just how they mechanically operate in play. They are also defining characteristics — they don’t just tell you how the character will engage mechanically but also who they are.

See, Mickey sucks. He’s below average in every stat but intelligence, where he’s perfectly average. He’s not gifted by birth at anything — he’s weak, he’s slow, he’s got a nasty cough that won’t go away, he dropped out of high school, and his parents were nobodies. But on the other hand, that’s a lot of information! How come he’s so weak and sickly? Maybe the fact that he’s a Belter has something to do with that. Maybe it’s the cigars he loves. How come he dropped out of school when he had the brains to pursue it at least a little further? What kind of nobodies were his parents? Why?

Mmm, derived detail. That’s my thing. I love random values and then trying to make sense of them. Lots of people prefer to throw up their hands and say “that’s ridiculous”, “that’s impossible”, “that’s unbelievable” but I always prefer to look at stats and say “I wonder so hard how that could be that I am now making up a plausible reason”. Weird stats are stories.

Since I started by talking about skills, look at Mickey and think about that.

But let’s look at something simpler. I recently rolled a character up with a random generator and got a character with Rifle-2, SMG-2, and … Hovercraft-2.

Look at that hovercraft.

Now here’s a skill problem I have confronted many times and usually my perspective is poisoned by a need for verisimilitude rather than a consideration of what else it delivers to the game. So anyway here it is. If you can drive a hovercraft, what else can you drive? What can’t you drive?

This is the essence in finding a category or generalization: looking at “hovercraft” and deciding what acceptable siblings it might have.

And then we have to wonder, if you don’t have Hovercraft as a skill, what would happen if you tried to drive one? Could you do it badly (a kind of gradated skill, like throwing a ball, where anyone can give it a shot and get some kind of result) or can you not even try (a qualified skill, like driving or surgery, where there’s a minimum knowledge you need to even figure out where to start)?

But that’s not what’s interesting about Hovercraft as a skill. It used to be but it matters less to me than the power of it: who is this person who only knows about hovercraft? Where do you come from where this is something you know to the exclusion of all other vehicles? Who is this Hovercraft Person?

And then in play, when I play Hovercraft Hero, look what happens. Whatever we are doing in the story, if I want to exercise my agency I need to make it about either gunfire or hovercrafts. Everything that happens in game, I am steering towards gunfire or hovercraft. Or both.

These two things are what makes a game go: who am I and what will I do? Specificity makes these a puzzle with some powerful clues. I am Hovercraft Hero. I am going to fucking hovercraft this problem.

13 thoughts on “skill specificity; in which I come full circle

  1. That’s really stimulating! I think perhaps ttrpgs tend to do things the wrong way around – general skill, with specialisation, whereas, as you indicate, the specialism is what’s interesting.

    In reality, most skills belong to families or have strong non-specialised aspects. For example – having spoken to one or two ex soldiers – there’s a difference between gun combat and marxmanship. “How hard would it be to hit those guys on the roof?” / “For a trained soldier, automatic.” The point being that the abstractable skill is really all about getting that shot off and not getting shot doing it.

    So, suppose skills were always expressed in terms of specialism – which is after all what a PC will focus on anyway?

    Your Rifle, Sub Machinegun and Hovercraft guy stays the same, but:

    – Rifle is a specialism of Fire and Manoeuvre. He also can function with, say, a LMG, or a range of weapons.
    – SMG is a specialism of Close Quarters Combat. He can also use pistol, grenade and in the initial clash, bayonet.
    -Hovercraft is a specialism of… Skimmers? He can also use a grav car and hoverboard.

    In this model, mechanically, your specialism might be rolled using 2D6, whereas straying from it would either take much more time, or else be more random, e.g. D12.

    Something like that?

    Liked by 1 person

    1. Yeah I think the observation admits to all kinds of possible design choices to exploit it. What you describe is pretty cool and aims at “realism” in a satisfying way.


      1. It would certainly make it possible to have minimalistic card style character sheets, similar to, say, games like Elder Sign, while making the skills non-generic. The challenge would be to make the specialisms – and hence specialist weapons – mean something mechanically. I’d rather have the differences be qualitative and situational, e.g. an LMG gives you bonuses on defence (effect of suppressive fire) and lets you take multiple shots when you have a clear view of the targets.

        I think “realism” has playability benefits because it makes the system predictable and easy to remember and navigate.

        Liked by 1 person

  2. There’s a useful aspect of specific skills in terms of world building. For generic space world you don’t need it, but to tell a story in a setting that is not familiar – say the weird 1920 Vanuatu, skill specifity helps everyone understand what’s important in the setting.

    Liked by 1 person

  3. Classic Traveller is not a good system anymore. Great for 1978, but after 40+ years of game design and innovation we have better games. Stars Without Number and Offworlders to name just two that follow in Traveller’s footsteps. Still it is worthwhile to look back now and then and see if there is anything we overlooked, or reflect on how we have changed.

    Liked by 1 person

    1. I have to disagree — we’ve been running a classic Traveller game for almost a year now and have been having a great time with it. There are other games and we like them too, but I think people overlook some of the things that make Traveller continue to reliably deliver.


      1. Are you REALLY running Classic Traveller? Or are you running Brad Murray’s Traveller on the bones of Classic Traveller?

        Liked by 1 person

  4. How do you feel about the handful of skills that leap scope to become classes? Admin is the equivalent of having a Modern Soldier skill that let you shoot, handle military chains of command, do tactics, drive and repair AFVs.

    Liked by 1 person

    1. Generally speaking I think they fit the scope of the rest of the rules. We have detailed rules for positioning and injuring people and detailed combat skills. But we have no rules for Administration beyond the skill. There’s a symmetry there that works.

      Liked by 1 person

      1. If these skills were acquired via a point buy system, the points would reflect “impact on play” rather than a realistic simulation. It might take years full time to acquire Admin 3, but in-game, it is no more – probably less – interesting than Gun Combat 3, which realistically would be acquired through part time training.

        I wonder, though. A “Specialism points to general skill” model could preserve the simulation.


      2. I’m not sure the simulation is the interesting part. The interesting part is that Admin is fine with broad coverage because the ruleset that covers Admin is tiny. Pistol has tight coverage because there is a detailed combat system that makes subtle distinction between different weapons. The degree of “impact on play” they have is determined both by how much impact they have via the rules and (far less quantifiably) by how much impact they have via the scenario.

        How we got to Admin-3 doesn’t really interest me. It’s just where we are and we can make up any story we want about how we got there. This attitude might be coloured by the fact that I’m uninterested in advancement systems though.


      3. Ah, that’s the nuance I was looking for!

        The Traveller character generation system creates interesting characters of approximately similar in-game agency. The profession skills work fine in that context. However, if one did have an advancement system, then things like Admin and Scientist shouldn’t normally get to advance from adventuring.

        Liked by 1 person

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s