Game design hypothesis (generic rules systems)

[Ashtagon]

Hypothesis: any ruleset in which a maximally-advanced character with no "super" equipment (ie. only equipment available to the "commoners" of the relevant settings) or a "commoner" with the maximum "super" equipment available in all settings it is intended for (ie. don't count magic swords or assault rifles in a system intended to cover both high magic and James Bond spy thrillers), can have a bonus that equals or exceeds the maximum range of the variable dice rolls is doomed to fail the "can it be generic" test.

d20 systems generally fail because a fighter 20 has +20 BAB on a 1d20 roll (plus at least another +10 from other accumulated bonuses by that point). Skill checks are even more extreme.

This hypothesis also explains the oft-noted D&D "sweet spot" of 5th-8th level. An 8th level fighter can expect +8 (bab) +1 (magic weapon) +4 (Strength) +1 (miscellaneous buffing magic) = +14, which is pushing the upper limit of "half the rolled variable range".

Contrariwise, WEG's TORG system had base characters with attributes ranging from 8-14 (6-point range); skills typically raise than bonus a further 6 points at the high end (to a 12-point range) with dice rolls that typically allowed for a range of 20 points (the roll was in essence skill level +1d10 -1d10). It generally worked.

Opinions?

[SowZ]

I disagree. I don't think being generic has anything to do with characters being more reliant on die rolls than flat stats. Systems with broad power ranges that can exceed the max dice on the upper end and fall far below it on the lower end of power are best at being generic because GMs decide the power/level curve of any given setting.

Being generic means the game should have lots of options and many rules should be easy to add as well as take away without hurting anything. Further, generic systems should be built with custom abilities in mind. That is what defines a good generic game, nothing to do with dice rolls I think.

[Ashtagon]

Take another game --- GURPS.

Base attributes typically go in the 8-14 range, and with skill bonuses, reach a maximum useful value of 20 (thanks to various hard caps on bonuses and modifiers) --- a 12-point range. The 3d6 dice range has a maximum variance of 15 points --- bigger than the range of typical character ability.

Perhaps it isn't a fixed requirement for a rules set to be generic. But so far, every successful generic rules set has matched this criterion. d20 systems have only really "met" it by having individual settings house rule the system so much that they are effectively incompatible with each other; d20 Conan's "low magic fantasy" rules are not really compatible with core D&D's high magic fantasy, or with Mutants & Masterminds superhero rules, or with Spycraft; they are all mutually incompatible due to the amount of house rules present in each. Contrariwise, GURPS and TORG handle different settings, and are written to allow a character from one setting to play in another setting seamlessly (power levels may vary due to different buy-buy values in GURPS or availability of reality points in TORG, but even with D&D high fantasy, characters have different levels).

[SowZ]

Take another game --- GURPS.

Base attributes typically go in the 8-14 range, and with skill bonuses, reach a maximum useful value of 20 (thanks to various hard caps on bonuses and modifiers) --- a 12-point range. The 3d6 dice range has a maximum variance of 15 points --- bigger than the range of typical character ability.

But that's not the reason GURPS is good at being generic. I fail to see how that links to being generic. GURPS is flexible because the vast number of optional rules and mechanics that work fairly consistently within a variety of sub-systems. Besides, GURPS has rules for superhuman stats/skills that way exceed the numbers you listed for certain game styles and URPS would be a piss poor generic system if it didn't.

[Djinn_in_Tonic]

You haven't disproven the hypothesis.

The problem with the hypothesis is twofold. What is the "can it be generic" test? Without the actual test it's hard to quantify: for example, I consider D&D to be pretty good at generic fantasy, but you're clearly using generic to mean something else (I believe...am I right?).

Further, just because no "generic" system has quantity X doesn't mean no generic system can have quantity X unless we can postulate a proven reason for the lack of X in all such systems.

Remember: correlation does not imply causation. Even if we cannot find a single exception to the hypothesis, all we've done is proven that no existing "generic" game has a bonus that equals or exceeds the maximum range of the variable dice rolls.

Base attributes typically go in the 8-14 range, and with skill bonuses, reach a maximum useful value of 20 (thanks to various hard caps on bonuses and modifiers) --- a 12-point range. The 3d6 dice range has a maximum variance of 15 points --- bigger than the range of typical character ability.

This is only technically true: in actual play, about 90% of your rolls will be within a range of variance equal to or smaller than your value range. While outliers do exist, it's hard to claim that a bell-curve game is big on variance when the entire purpose of a bell-curve based system is to emphasis character skill over random chance.

d20 systems generally fail because a fighter 20 has +20 BAB on a 1d20 roll (plus at least another +10 from other accumulated bonuses by that point). Skill checks are even more extreme.

This hypothesis also explains the oft-noted D&D "sweet spot" of 5th-8th level. An 8th level fighter can expect +8 (bab) +1 (magic weapon) +4 (Strength) +1 (miscellaneous buffing magic) = +14, which is pushing the upper limit of "half the rolled variable range".

Interesting. The reason I'm most familiar with (and the one that I, as a long-time player and homebrewer for 3.5, hold to be true) for the existence of the D&D "sweet spot" is that it is between levels 5 and 8 where the Wizard stops dying if breathed upon heavily (and has enough spells to contribute in more than 1-2 encounters before becoming a commoner with a crossbow), but the Fighter's contribution to a given combat encounter hasn't yet been rendered meaningless by 5th+ level Wizard and Cleric spells. The fact that one more level added onto that 5-8 range (which means little for the Fighter's +1 to hit: that's a 5% increase in effectiveness, which won't smash open a system) adds the encounter-ending/encounter-avoiding likes of Plane Shift, Slay Living, Scrying, True Seeing, Cloudkill, Teleport, Contact Other Plane, Hold Monster, and Dominate Person is, I believe, more likely to cause the disintegration of the "sweet spot."

Opinions?

FATE is a great generic game where the expenditure of FATE points (which are fairly common) and a high skill can easily put you over the default -4 to +4 bell curve or just smash the bell curve into effective nonexistence via re-rolls. d20 Modern can be turned into a generic system fairly easily, and has the sort of high rolls that D&D has. Deadlands' core system, if stripped of the supernatural elements, makes for a great generic system for grittier games, and has no real "bonus" structure to speak of. World of Darkness also has no real "bonus" structure, and the core book can be used for generic games quite easily.

However, I can't really bring up counter-examples without knowing what you mean by the "can it be generic" test. Can you elaborate?

[CarpeGuitarrem]

Might have a point to it, even if it's not a perfect test. The correlation could be on to something.

Generic systems can emulate a broad range of genres.

The systems you talk about cover a specific range of success: you always have a chance at anything, and you never have an automatic success at anything. We'll call this the "Possibility Range", for lack of a cool term.

Most genres overlap with the Possibility Range on some level. If a system uses the Possibility Range, therefore, it has overlap with most genres.

Fate, actually, isn't a completely generic system. It rides off of a presumption of frequently successful protagonists plus a very narratively dramatic scale of action. You have to significantly retool the core assumptions of the game to do something other than adventure fiction.

[erikun]

Counterpoint: D&De3 is not a generic system, and the faults in D&D3e are not about it's nature at trying to be a generic system. The faults are all internal to the system itself, and about the pointlessness of some options compared to others.

Counterpoint: Mutants & Masterminds can work as a generic superhero fantasy game, or a rather decent generic game. You can have bonuses exceeding +20 on rolls at high levels, but this does not break the game and does not cause it to function any worse than it does at lower levels, primarily because the range of target numbers is appropriately high.

Counterpoint: I could design a d% roll-under system where the target number represents the chance of succeeding at a task. We could implement bonuses of +100% or more, and target difficulties in the 101-200 range that represent superhuman (or highly specialized) challenges. I fail to see how this makes such a system unusable for generic play.


I'm not sure what you mean by a generic system, and am not sure what you mean by stuff like "maximally-advanced character" and "super equipment". However, I'm pretty sure that it is entirely possible to have a generic system which bonuses larger than the dice range. The system just needs to be designed to handle such variables in it.

[Quellian-dyrae]

I'm...actually inclined to the exact opposite. Or at least, I'd say that a generic system where the stats cannot exceed the range of the dice has a significant weakness, because it doesn't do well at emulating high-power concepts. I mean, assuming I'm understanding what you mean with generic systems correctly (basically, a game where the rules base can be used to emulate a particularly wide range of genres and character concepts, yeah?)

A generic system where you roll 1d20 and the maximum bonus is 18 means there's no way for an opponent to be, say, so tough that you can't hurt it*. Which means the system can't accurately model, for example, an unarmed soldier attacking a modern tank. Or a common crook attacking Superman, or whatever.

*I'm assuming things like DR are counted as stats for this purpose...if they aren't then I've missed the point of this thread entirely .

[Ashtagon]

I think the reason that having the base stats possible for a hero for all of the viable settings not exceed the dice range being needed for a generic system is that otherwise, it is impossible to cover "gritty" concepts. The reverse is not true. While it is easy to add genre-specific bonuses to infinity and beyond (well, to unreasonably large numbers anyway), the reverse is not true. It is extremely hard to remove bonuses that exist in the core rules in order to make a genre-specific version and still have characters essentially transferable from one genre to the next.

Imagine a D&D game in which BAB ranges up to +10 for fighters characters at level 20 (and up to +0 for wizards). That certainly allows for a grittier style of D&D. But conversely, if you want to make a "high fantasy" setting, they can "find" all manner of magic items, mystic rituals, secret training, or whatever is needed to add on extra bonuses to their rolls. There's no easy way to do the reverse if you want the standard D&D high fantasy rules used for gritty (short of playing e6 and effectively cutting out 2/3 of the game's level progression rules).

@erikun: I define a "maximally-advanced character" as a player-character (not necessarily the same as an NPC; some genres expect NPCs to be more powerful, or to be less powerful) with the highest amount of level progression possible for conventional games, by whatever standard is used to measure level progression. For D&D, that means level 20. For WFRP 2e/3e, that means a character with 10-15 fully completed careers. For GURPS, that's much harder to pin down; GURPS Dungeon Fantasy might man a 500-point character, while Supers might mean a 5000-point character.

I define "super equipment" to be equipment not routinely available to the "commoners" of the setting. For high fantasy, that generally means powerful magic items that are expensive but not unique and that a PC could reasonably expect to get hold of. In a spy thriller game, it might mean a laser pistol prototype or advanced eavesdropping tech. In a sci-fi setting, it might mean an armoured combat spacesuit (normal civilians use unarmoured space suits) and plasma rifles. Basically, could a middle-class "commoner" walk into a store and buy it with a month's wages or less.

[Quellian-dyrae]

I think the reason that having the base stats possible for a hero for all of the viable settings not exceed the dice range being needed for a generic system is that otherwise, it is impossible to cover "gritty" concepts. The reverse is not true. While it is easy to add genre-specific bonuses to infinity and beyond (well, to unreasonably large numbers anyway), the reverse is not true. It is extremely hard to remove bonuses that exist in the core rules in order to make a genre-specific version and still have characters essentially transferable from one genre to the next.

Imagine a D&D game in which BAB ranges up to +10 for fighters characters at level 20 (and up to +0 for wizards). That certainly allows for a grittier style of D&D. But conversely, if you want to make a "high fantasy" setting, they can "find" all manner of magic items, mystic rituals, secret training, or whatever is needed to add on extra bonuses to their rolls. There's no easy way to do the reverse if you want the standard D&D high fantasy rules used for gritty (short of playing e6 and effectively cutting out 2/3 of the game's level progression rules).

The inverse of that argument would be saying there's no easy way to do awesomely powerful archmages in D&D (short of starting off at level 14 and effectively cutting out 2/3 of the game's level progression rules).

Having a wider range of power levels does not prevent gritty play, any more than it prevents heroic play. It may consolidate those two styles to a certain range of levels, but that doesn't make the game worse at being generic. I'd say it is actually more useful for a generic game if the power levels are clearly defined. That way, just because the game can do armored tanks, it doesn't mean you might accidentally get one when you're trying to play mundane humans.

EDIT: That being said, it is nice if the range of power is such that there's room for advancement within a given power band, so you don't wind up faced with the choice of stagnating mechanical development versus pushing outside of the intended power level of the game. But that's a fairly different issue.

[erikun]

I admit to being quite confused right now. What do you mean by "gritty" exactly? Because I'm inclined to call a gritty campaign as one with high challenge and low resources to rely on. I mean, you say that D&D3e can't be gritty and then turn around and say that E6 is gritty. This seems to completely invalidate your claim; clearly, you can play D&D3e as gritty.

Or do you mean that you cannot play a gritty game while using the full ruleset? While this is true, it doesn't really state much. Unless your system is so incredibly restricted that gritty play is the default core assumption, you're going to be tossing out options to make it so. You don't allow every option in the core Gurps rulebook when playing a gritty campaign.

[NichG]

Its interesting that you're tackling this from the point of view of 'generic' versus non-generic. I ended up addressing this in a homebrew system I'm using for my current campaign, but for completely different reasons. Maybe they're related though.

A bit of history - my gaming group has a tendency to like games that cover a huge range of power-scales over the course of the campaign. We've each played in games that were basically commoner -> Nobilis in scope, and it has kind of become a hallmark of our campaigns.

The problem is that in D&D, everything is dominated by modifiers at the high end. I've had a DM who addressed this by having characters switch over to rolling a d100 instead of a d20 at a certain point in the campaign, but this is a pretty rough transition (The sudden +40 to all checks is kind of extreme). Systems like World of Darkness are better about this, but there's a practical issue in that we don't want to roll and count up 100 dice for this or that - it scales mathematically but not in play.

So to address this I made a system where you roll a die and multiply the result by a 'skill rate', before adding your ranks in the skill. The skill rate scales with the absolute skill rank/10, so the math isn't too bad, and the variance is always the same fixed fraction of the mean.

The thing I think is odd is, we did this in order to preserve the core dice mechanics across transitions between power-levels and styles of game. That said, there's no reason if you just wanted to do a campaign with a single style and power-level, you couldn't just focus in and prevent advancement outside of that window. For example, in D&D if you want gritty you could run a Lv3 campaign where you never get beyond Lv3. So if you're just looking for a certain 'static' feel from a generic system, I don't think this is necessary.

On the other hand, if you want advancement to feel a certain way, something like this consideration becomes a lot more important.