Attribute Scores: Chicken or Egg Problem

Tadatsune

Registered Magic-User
Validated User
#11
Might needs a little more work. I'd say might is physical and psychological effect of said physicality.
Not really the intended focus of the thread, but OK. So, what, intimidation type abilities? I was thinking that persuasion/influence rolls would be based off W+P, but trying to physically intimidate someone would be M+W.
 

Tadatsune

Registered Magic-User
Validated User
#12
Note that if you systematically add 2 characteristics together, you create an implicit list of 30 broad aptitudes, one for each couple of characteristics (even though some will rarely be used in play).
I've become convinced that symmetry is the tool of the devil. Trying to come up with a reasonable example for every combination is a bit of a nightmare - meanwhile some combos are applicable to tons of situations.
 

Tadatsune

Registered Magic-User
Validated User
#13
I don't think there's anything intrinsically wrong with the second approach, only that it seems further removed from the idea of random stats plus interpretation (which was your starting point for the post) than the first.
Edit: I read you wrong the first time. Let me think about this for a minute.

Edit Edit: OK, I think I get it now - I believe I was thrown off by the difference between the order I presented the options and that on which you referenced them. If I understand what you are saying, the idea of randomly generating a number and then choosing traits based off that number is more in line with the spirit of this random char-gen endeavor than allowing the traits to adjust the overall generalized score., and makes that generalized score more abstract in the view of the player. Am I getting that right?
 
Last edited:

kenco

Registered User
Validated User
#14
Edit: I read you wrong the first time. Let me think about this for a minute.

Edit Edit: OK, I think I get it now - I believe I was thrown off by the difference between the order I presented the options and that on which you referenced them. If I understand what you are saying, the idea of randomly generating a number and then choosing traits based off that number is more in line with the spirit of this random char-gen endeavor than allowing the traits to adjust the overall generalized score., and makes that generalized score more abstract in the view of the player. Am I getting that right?
Yep. You said it much more succinctly though! :)
 

Tadatsune

Registered Magic-User
Validated User
#15
Yep. You said it much more succinctly though! :)
OK, thanks.

So, on an abstract level I agree with you. I'm not sure how well the theory holds up in practice, though. If I were just handing out a bunch of traits and then calculating a score based off that, then yeah, it might come off as a little weird conceptually. Let me further explain the system, so you can better understand what I'm talking about:

Everyone starts off with the same stats... say, 5s across the board. Then we roll (or pick) a character archetype. This system is meant for a dungeon-crawl style fantasy RPG, so we have three basic archetypes with different "aspects" (basically sub-classes, but intended to be more flexible). Depending on your archetype you get bonuses to certain attributes, namely +1 to Might and Fortitude for "warriors," +1 to Agility and Perception for "rogues" and +1 to Intellect and Will for "adepts." Your "aspect" might add a further +1 (not decided yet). Then you roll (or choose) an ancestry and a background, each which may provide further bonuses and penalties. Then you will roll a series of "traits" - from different pools - some of these will be linked to your archetype, others may be linked to your background/heritage, while still others will come from a two pools of general traits (one positive, one negative) which all characters draw from. Adding up all the bonuses and penalties will result in your final attribute scores.

The goal here is to produce "random" attribute scores that none the less maintain an identical sum across all player characters (so you don't end up with one guy with all 3s next to a dude with 18s in every category, a la classic D&D). Additionally, the hope is that we make all attributes important to all "archetypes" and "aspects" such that there are no "dump" stats, to prevent min-maxing players who elect to choose their various traits from having a significant edge over players who go with the default random selection method. (Some of this is inevitable, but the goal is to try to keep its impact to a minimum).

There's a bit of overlap in the system as currently conceived, so I need to think hard about where I want the actual numbers to come from - should a background or heritage give stat adjustments directly, or should that be handled by a pool of linked traits instead? I want to try to keep the system as streamlined as possible and to avoid number bloat if I can, so my aim is to reduce and consolidate where I can.

Back to the conceptual stuff. As you noted, in traditional random attribute systems, you generate a hard number and then the player is left to interpret what that number actual means in-world for their character. But this system is intended to take some of that pressure off the player - at least in the very beginning; where a player chooses to take their interpretation of their character after char-gen is entirely up to them. It's supposed to do this while simultaneously providing a certain level of mechanical heft to these various traits as well as providing Role-Playing queues for the player. Furthermore, the system doesn't tell you everything about a character's person and personality - just the key features, so you can quickly build from there. The benefits of the system are intended to be quick but meaningful character generation that can provide you an instant foundation for a one-shot character, a replacement character, or a character for a drop-in player who joins mid-game. Basically, it provides you with a jump-start - if you want that. If you don't, you can just go through the tables and choose whatever suits your fancy.

I think the first option presented suits this system well (in fact, the more I write about it the more that belief is reaffirmed). My only reservation was the possibility that you might end up with, say, a "weak" Warrior with a negligible strength score, but none the less having "mighty thews" or some-such. I'm not so much worried about the mechanics of this as the potential cognitive dissonance - your "halfling barbarian" might have "mighty thews," but they're mighty for a mighty for a halfling and don't change the fact that he's a little guy who spent most of his life studying to be a scribe before running away and getting adopted into a tribe of wildlings. That sort of "contradiction" is perfectly fine with me, but I'm wondering if it will cause a problem for some players. Thus the second system, which proposes to create a tiered system of traits keyed to the various attributes - think something like an expansion of Fallout 4's perk chart, where you need a certain level of attribute to unlock a certain perk. There are however, some serious issues with this - like, how do I randomize this? (not sure its possible) how do I balance different tiers of perk? Shouldn't a player with higher might be able to not only choose "better" might perks, but more of them? I think what you'd need to do is eschew static situational bonuses in favor of bonuses that improve based on the underlying attribute - that way lower tier traits would be in line with higher tier ones and you could then randomize selection. Frankly, that seems like a lot of work - and potentially unbalancing complexity - for what seems like a marginal benefit.

Or at least, that's what I'm currently thinking.

P.S. Thanks for your replies - this board can be a bit slow, so they are doubly appreciated. Just writing about stuff is very helpful for clarifying things.
 
Last edited:

kenco

Registered User
Validated User
#17
...the hope is that we make all attributes important to all "archetypes" and "aspects" ...to prevent min-maxing ... (Some of this is inevitable... keep its impact to a minimum).
That's an important consideration, and well thought through.

It does make me wonder why a Warrior archetype would get +STR, +FOR. If a Warrior with +PER, +WIL is just as effective overall, why would the 'archetype' select for STR and FOR?
...I need to think hard about where I want the actual numbers to come from - should a background or heritage give stat adjustments directly, or should that be handled by a pool of linked traits instead? I want to ...keep the system as streamlined as possible ...
If a given background always gave you the same set of traits every time, then it would automatically have an associated set of attribute adjustments (which you would pre-calculate for player convenience). That seems straightforward and eliminates a possible round of rolling traits: therefore faster.
...this system is intended to take... pressure off the player... while... providing ...mechanical heft to ... traits...[&] Role-Playing queues.... [it] doesn't tell you everything about a character... just the key features, so you can quickly build from there. The benefits ...quick but meaningful character generation ... provide ... a one-shot character, a replacement character, or a character for a drop-in player who joins mid-game. Basically... a jump-start - if you want that. If you don't... go through the tables and choose ...
It sounds just fine for that. I always appreciate a system that makes it quick to generate a character, and I enjoy grappling with a random one. As I said earlier, I'm not a fan of really long lists of specific traits, each with its own definition etc.; but that's a very personal preference.

What about gear? In some games (e.g. early D&D) gear choices are vital for some (or all) characters, and picking them is the main time-sink in the character generation process.

Does the system generate starting gear? Or is that simply not an issue for your system/ genre?
I think the first option presented suits this system well (in fact, the more I write about it the more that belief is reaffirmed).
Agreed, based on your description of the second option, below.
My only reservation was ...you might end up with... a "weak" Warrior with a negligible strength score, but ... "mighty thews" ... I'm wondering if it will cause a problem for some players.
You have a good point about the muscly halfling. You can also avoid some strange-seeming possibilities by how you name your traits. If the trait is called 'small' or 'short', then it clashes less obviously with 'mighty thews'.

No doubt it will cause a problem for some players! But not for many players, I think.:)
...the second system... proposes ...a tiered system of traits keyed to the various attributes... where you need a certain level of attribute to unlock a certain perk.
This seems more complicated than it needs to be. Your score in a given attribute could determine how many + and how many - traits you get for that attribute. Why complicate it by adding the tiers/ permissions?

If your ONLY concern with the 1st system is a few players being worried about odd-sounding combos, this 2nd approach sounds like a lot of work to solve a minor problem.
... serious issues...how do I randomize this? (not sure its possible) how do I balance different tiers of perk?
Yeah. I'd drop the tiers and make all traits/ perks equal (as in System 1).

If you STILL want to limit permissions based on your attribute score (even though they are all 'worth' roughly the same), you can offer a list of traits for each die roll/ table entry. You must always take the first trait in the list for which you have the required attribute score. E.g. One table entry might be: Might Thews (7)/ Muscly (6)/ Quick (5)/Healthy (2). If you have 7+ STR, you get Mighty Thews. If you have 6 STR you get Muscly. If you have 5 STR you get Quick. Otherwise you get Healthy.

A given trait need not be given the same number score every time it appears (so that sometimes a 5 STR character might be Muscly). A given trait might even appear in more than one Attribute table.

This might have the interesting effect of granting you the same trait more than once... you'd have to design accordingly. Or if you would otherwise get a duplicate, you might get bumped up or down in the list. So if you have 5 STR in the above case, the first roll on that result gives you Quick, but the second gives you Muscly (because you already have Quick).
Shouldn't a player with higher might be able to not only choose "better" might perks, but more of them?
More of them, yep. In fact, that's what I thought you were originally suggesting... (see above)
...eschew static situational bonuses in favor of bonuses that improve based on the underlying attribute - that way lower tier traits would be in line with higher tier ones and you could then randomize selection. Frankly, that seems like a lot of work - and potentially unbalancing complexity - for what seems like a marginal benefit.
That's a pretty interesting solution, but I agree that it's probably not worth the trouble.

Cheers

K
 

Tadatsune

Registered Magic-User
Validated User
#18
It does make me wonder why a Warrior archetype would get +STR, +FOR. If a Warrior with +PER, +WIL is just as effective overall, why would the 'archetype' select for STR and FOR?
Well, it's partially a matter of perspective - i.e. how much you value a particular, combat specific role vs. other, different combat roles and non-combat activities. Again, I think some min/maxing potential is unavoidable. Increased Might and Fortitude means increased capacity to deal damage and to resist death and injury, increased stamina, etc. +1 M & F is both mechanically and thematically appropriate for pretty much any Warrior type - including those that make heavy uses of Agility or Perception (Duelists or Marksman, for instance). However, the mechanical benefit means you can indeed potentially max these out with targeted choices to create a Warrior that hits exceptionally hard and is very difficult to kill.

The hope is that we can mitigate the impact of this sort of min/maxing by making other traits valuable to a "Warrior" such that there is a real trade-off being made. So, reduced Will means failed saves against fear or domination effects, lower chance of surviving death saves, reduced stamina pool/capacity for second winds, lower leadership bonuses, reduced capacity to resist challenges/control berserker rage features, &etc. Out of combat, it means less capacity for social influence, less capacity to resist temptation, etc. Lower agility or perception means lower dodge/guard ratings, reduced ranged accuracy, lower stealth and acrobatics skills, lower initiative and awareness, etc. Lower Intellect means reduced access to knowledge skills, bonuses from tactical features, resistance to illusions, etc.

Now, it's overly optimistic for me to think that I can completely mitigate min-maxing potential, but the hope is that I can get attribute interdependence to a point that the advantages of min-maxing are marginal compared to what you give up.
 

Tadatsune

Registered Magic-User
Validated User
#19
What about gear? In some games (e.g. early D&D) gear choices are vital for some (or all) characters, and picking them is the main time-sink in the character generation process.

Does the system generate starting gear? Or is that simply not an issue for your system/ genre?
I'd like to do randomly generated "starter-kits" for players to grab-and-go, but I'm currently struggling to integrate this with my proficiency system. The game uses a broad proficiency system which breaks down weapons into major categories - "Axe and Bludgeon," "Blade," "Spear and Poleam," "Marksmanship" and "Unarmed." Under the current concept, acquiring a level of proficiency (from 1 up to a max of 5) gives you +1 to hit with all weapons in the category, as well as an "expertise" trait of your choice which would grant you additional benefits (such as access to new tactical maneuvers) to specific sub-categories of weapons. So, a character with a level of proficiency in "Blade" would get to choose between "long blade" and "short blade" expertise. If you select (or roll) a particular proficiency/expertise combo in generation, you get a free weapon for that category - but determining the exact weapon will probably be left to player choice. I'm not sure how to make this approach compatible with a theme-based starter package - such as a "wandering pilgrim" package with robes and pilgrim's staff that might be available to an Adept.

One possibility is to reverse the process, such that the equipment package you roll retroactively determines your weapon and armor proficiencies. I'm not entirely sold on that approach, though.
 

Tadatsune

Registered Magic-User
Validated User
#20
This seems more complicated than it needs to be. Your score in a given attribute could determine how many + and how many - traits you get for that attribute. Why complicate it by adding the tiers/ permissions?

If your ONLY concern with the 1st system is a few players being worried about odd-sounding combos, this 2nd approach sounds like a lot of work to solve a minor problem.
You make a very good point. It would be easier to simply control the number of traits rather than trying to tier them. That said, I am strongly leaning toward the other approach for stat generation.
 
Top Bottom