In Nomine Second Edition: What have we learned?

[dataweaver]

Originally Posted by Rocket Man

Beth said what I had in mind, and far better than I would have. So I'll just add one summary passage:

GURPS is a game built for humans, into which celestials can be plugged. In Nomine is a game built for celestials, into which humans can be plugged. That in itself explains a lot of the difference.

Yeah; but I don't think that the disparity between humans and celestials needs to be as great as it is in order for In Nomine to be In Nomine. Contrast this with In Nomine's concept-driven design philosophy vs. GURPS' accounting-driven design philosophy, which I consider to be a much more fundamental difference between the games.

[dataweaver]

Personally, I'd like to see an IN2e play up its concept-driven design philosophy even further. For instance:

Make Roles the primary source for Skills. Each Role would have a list of Essential Skills and a list of Support Skills (representing skills that effectively define the role and skills that are important but not central to the role, respectively). These lists would generally be short — typically one Essential Skill and two or three Supporting Skills — but there would be no hard upper limit: it's based on what makes sense for the Role rather than what's "balanced". You would get the Essential Skill(s) at the Role's level, and you'd get the Support Skills at half the Role's level.

In this approach, humans would have Roles, but only for the Skills. In fact, a variation of this would be to say that the Essential/Supporting Skills is unique to humans, and Celestials have to buy those Skills independently. Doing so would interfere with the next possibility, though:

Another reason for doing this is that it helps lay the groundwork for "abbreviated characters": when you're dealing with a minor NPC, you could get a reasonably well-defined character by specifying his Forces and his Roles, and disregarding Characteristics and Skills entirely. Where needed, Characteristics could be assumed to be twice the appropriate Forces, and Skills could be assumed to be equal to or half of an appropriate Role. There would still be a number of "fiddly bits" to deal with (Attunements, Artifacts, Songs, and Vessels, IIRC); but every little bit of (optional) streamlining helps.

[Archangel Beth]

Originally Posted by dataweaver

"The symmetry of 5s"? This is the first I've heard of that; what is it, and why is it important?

"Brainfeel," mostly. It's the starting-values and ultimate values. Angels are based on 3s -- 3 Forces in each realm as the sort of "average," total of 9, max of 18. Humans are based on 5s -- 5 Forces for an adult mundane (breaking into 3*2 for a Soldier), 15 for a max human.

Threes are the whole d666 thing -- especially the 111. Fives are HAIL ERIS, ALL HAIL DISCORDIA... Er, ahem. Fives are just... human. Five fingers on the hand. Counting by fives.

It's brainfeel.

Perhaps that's a result of the way that the translation was done, and a 4e-compatible translation would do a better job; after all, many of the changes that took place in 4e were the direct result of lessons learned when writing GIN.

First I've heard of that claim! O:D


But I do think the main hurdle is the granularity of the d666. In a 2d6 system, it's harder (than in a 3d6 system) to make something "noticeably stronger" without being overwhelmingly so. The bell-curve is what it is. You can tweak with more deadly guns, wimpier vessels, human skill bonuses, Extra Points For Humans (that's another quick-and-dirty way to make humans better; give them more points for Skills), general roleplaying bennies... But the ranges just... compress differently. (I remember this being something we discussed a lot for Songs, because of how hard it is to do a Song in IN, versus the spell mechanics (which were clearly analogous) in GURPS.)

I don't actually find GURPS (well, GURPS 3rd edition) point accounting to be much of an issue, personally -- but then, I was co-author of GURPS IOU, and am very comfortable with doing stuff where fiddly bits are pitched out the window at need. O;>

Hmmm. I wonder if re-working from a base of http://e23.sjgames.com/item.html?id=SJG37-0032 might be an interesting exercise. It'd probably have to be a hybrid, and use 3d6, but it seems like it ought to be possible to wed Really Simple GURPS to the IN concepts somehow. ...I'll have to think of it sometime when I have more brains.

[dataweaver]

Originally Posted by Archangel Beth

"Brainfeel," mostly. It's the starting-values and ultimate values. Angels are based on 3s -- 3 Forces in each realm as the sort of "average," total of 9, max of 18. Humans are based on 5s -- 5 Forces for an adult mundane (breaking into 3*2 for a Soldier), 15 for a max human.

Threes are the whole d666 thing -- especially the 111. Fives are HAIL ERIS, ALL HAIL DISCORDIA... Er, ahem. Fives are just... human. Five fingers on the hand. Counting by fives.

It's brainfeel.

Ah. Not something that concerns me terribly much; but that's just me.

Originally Posted by Archangel Beth

First I've heard of that claim! O:D

It was mentioned that G4e's new Possession rules were written the way they were because of the experience of trying to write up Vessels and Body Hoppers in GIN without them.

Originally Posted by Archangel Beth

But I do think the main hurdle is the granularity of the d666. In a 2d6 system, it's harder (than in a 3d6 system) to make something "noticeably stronger" without being overwhelmingly so.

Conversely, it means that a small tweak might be all that's needed.

Originally Posted by Archangel Beth

I don't actually find GURPS (well, GURPS 3rd edition) point accounting to be much of an issue, personally -- but then, I was co-author of GURPS IOU, and am very comfortable with doing stuff where fiddly bits are pitched out the window at need. O;>

Yeah; I've always figured that Archangel Beth's Citadel looks suspiciously like the campus of IOU…

On the point of your comment, though, that's great advice for a GM; but in terms of game design, one of the things that I like about In Nomine's game engine is that there are fewer fiddly bits to throw out.

Which is where your musings about an In Nomine/GURPS Ultra-Light hybrid comes in. That said, GUL uses a straight 3d6 mechanic and the four standard GURPS attributes, both of which are marks against it as an In Nomine substitute (in my book, at least).

[Jason]

Originally Posted by dataweaver

Make Roles the primary source for Skills. Each Role would have a list of Essential Skills and a list of Support Skills (representing skills that effectively define the role and skills that are important but not central to the role, respectively). These lists would generally be short — typically one Essential Skill and two or three Supporting Skills — but there would be no hard upper limit: it's based on what makes sense for the Role rather than what's "balanced". You would get the Essential Skill(s) at the Role's level, and you'd get the Support Skills at half the Role's level.

I rather like this idea; I may adapt it for a house rule for my own game in the future. It always bugged me that a celestial could have lived hundreds of years, but not have the skills you'd think would have come with multiple human-length lifetimes. If they've been embroiled in celestial affairs on Earth this whole time, though, and not actively living human-like lifetimes, that's another matter.

[dataweaver]

A fairly radical idea to consider: a variant of the d666 mechanic that is perhaps both more flexible and more in the spirit of the game than the current mechanic.

The die pool consists of three unique dice: a Divine die, an Infernal die, and a Check die. Roll all three, then compare the Divine and Infernal dice to each other and keep the smaller of the two. If you keep the one that's aligned with your side of the War, add it to the trait being tested; if you keep the one that's opposed to your side, subtract it instead. Compare the modified trait to a target number set by the GM: if the modified trait is at least as high as the target number, you succeed; look at the Check Digit to see how well you did. (Use the existing guidelines for Divine and Infernal Interventions to establish whether the Divine die adds and the Infernal die subtracts or vice versa.) Divine and Infernal Interventions are unchanged.

Commentary:

Note that without Interventions in the system, a 111 and a 666 would produce midline results in terms of the chance of success (the equivalent of rolling a 7 under the current d666 model), as opposed to the current model where a 1+1 is the best possible roll and a 6+6 is the worst possible roll. This has some minor implications in terms of the impact of Interventions on the game, in that demons aren't going to be turning a worst-possible roll (6+6, CD 6) into a best-possible outcome and vice versa, nor will angels turn best-possible rolls into best-possible outcomes. Both will take mid-level rolls and turn them into best- or worst-possible outcomes. There will be a slight bias in favor of angels, in that Divine Interventions will upgrade a CD 1 with a best-possible result whereas demons who roll Infernal Interventions will replace a CD 6 with a best-possible result.

Statistically, the opposed dice method is equivalent to a 2d6 roll, except that it's centered on 0 rather than 7. It should be obvious how this is more in the spirit of the game than the existing d666 die mechanic. The flexibility comes with the fact that the target number is a distinct value, a "difficulty rating" that effectively measures how good of a rating you need to have in order to have a better-than-average chance of success. Set the default target number at 7 to get the same odds of success as the current mechanic produces; drop it down to 6 or even 5 to make characters more successful across the board. (I'd go with 6 as the default target number, myself: better "brainfeel" and still produces reasonable results.)

Bonuses and penalties can either be dropped in favor of a more freeform approach (i.e., the GM assigns a target number based on his assessment of the overall situation) or they can be reversed and applied to the target number: so what is currently a +1 bonus to your roll instead becomes a "-1 difficulty" to the target number. Or keep them as is, with a universally fixed target number. The first option minimizes the "fiddly bits" involved, but puts extra weight on GM judgment calls; the second and third provide more guidance for the GM, but require more fiddly bits.

The main drawback to this approach is that you need to be able to distinguish the three dice from each other. If the only dice you have to work with are scavenged from board games and the like, this could be an issue. Conversely, it opens up another potential revenue stream for In Nomine, in the form of In Nomine "vanity dice" packs: the Divine die could be white with gold pips and perhaps a stylized "angel wings" motif (or go with the symbols of six Choirs in ascending order, excluding the Malakim); the Infernal die could be black with bright red pips and possibly some sort of "demon wings" or "horns and tail" motif (or the six non-Lilim Bands in ascending order). I'm less sure what the check die would look like, other than being unadorned and having a distinct color scheme.

[Rocket Man]

As far as tri-colored dice, I use that already -- white, black and red dice, thanks in part to an old Axis & Allies set.

[Omegonthesane]

Dataweaver - your description of the mechanic fails to convey the information correctly - although I am guessing it was your intent that a roll where Divine and Diabolical dice came up the same value would be a +0 instead of a +Something to the attribute being tested.

I feel the suggestion of changing the d666 from 2d6+Check to (Divine or Diabolical) + Check adds little to nothing. As you say, it is indeed statistically equivalent to a 2d6 roll centred around 0, so all you have actually achieved is to reduce all the randomly generated values by 7. In exchange for this, you require that all three dice, rather than only one, should be distinguishable; and that the work required to interpret a roll go from "add 2 numbers [the 2d6], compare 2 numbers [the total and the TN], report check digit" to "compare 2 numbers [the Divine and Diabolical Dice], add 2 numbers [the qualifying dice and the attribute], compare 2 numbers [the total and the TN]", adding an operation to every roll that was not needed.

The suggestion of changing to a roll-over rather than roll-under model has more merit. In terms of operations, nothing is lost; in terms of probability, the system only needs to have numbers changed rather than any deeper reworking in order to implement such a change. It would not be a great idea to add or subtract things from the target number in any circumstances - it is easier to add or subtract things from the total of the dice in front of you than it is to add or subtract them from the not-in-front-of-you number that your roll is to be compared to. Even if a +4 to the roll is, in fact, equivalent to a -4 to the target number that must be met or exceeded by the roll.

As it happens, the vanity dice can be made regardless, and were I inclined to get such things in general I'd definitely go for such a dice set.

[dataweaver]

Originally Posted by Omegonthesane

Dataweaver - your description of the mechanic fails to convey the information correctly - although I am guessing it was your intent that a roll where Divine and Diabolical dice came up the same value would be a +0 instead of a +Something to the attribute being tested.

Yeah; I figured that that was obvious. If you're supposed to keep the lowest one and there is no lowest one, you don't keep anything — and you don't do a +0 either; you just skip straight to comparing the trait to the target number. But if you want it spelled out, that can be done with a single, short sentence: "on a tie, discard both dice."

Originally Posted by Omegonthesane

I feel the suggestion of changing the d666 from 2d6+Check to (Divine or Diabolical) + Check adds little to nothing. As you say, it is indeed statistically equivalent to a 2d6 roll centered around 0, so all you have actually achieved is to reduce all the randomly generated values by 7.

Where it really shines is in the aesthetics. It just feels more apropos to a game about angels and demons for the dice mechanic to feature a clash between the divine and the infernal. And reducing the randomly generated values by 7 is no small thing, as it produces more pleasing target numbers: trait vs. 7 for the "opposed dice" model, contrasted to trait + 2d6 vs. 14 in order to switch the current model over to a roll-over mechanic. (And if you go with needing to beat the target number rather than just tying with it, you can go with the more elegant target number of 6 with the opposed dice method. Making an equivalent change to the trait +2d6 approach wouldn't get you anything, as 13 is no more pleasing of a number than 14 is.)

Originally Posted by Omegonthesane

In exchange for this, you require that all three dice, rather than only one, should be distinguishable; and that the work required to interpret a roll go from "add 2 numbers [the 2d6], compare 2 numbers [the total and the TN], report check digit" to "compare 2 numbers [the Divine and Diabolical Dice], add 2 numbers [the qualifying dice and the attribute], compare 2 numbers [the total and the TN]", adding an operation to every roll that was not needed.

Technically yes; although as I pointed out above, the second step is skipped if the Divine and Infernal dice tie. In practice, I've found that this approach runs just as smoothly as the 2d6 method does: I've used it in play extensively, and the only complaints that I've ever had about it have been from people who haven't tried it and assume from the more involved description that it's more difficult than it really is. (Well, that and a few long-time gamers who were already familiar with 2d6 and didn't like leaving their comfort zone.)

The need to have three distinct dice is a potential problem, as I've already acknowledged; but it's a fairly minor one, especially if you're already having to distinguish two ways (chances are that if you have two kinds of d6s in your home, you probably also have a third kind). In fact, in a pinch, you could even get by with using a single die: roll it once as your favorable die, and a second time as your unfavorable die, replacing the first with the second if the second is lower. Only roll it a third time if the check die's value matters (i.e., if you succeed or if the first two rolls indicate the possibility of an Intervention). Mind you, I wouldn't recommend doing this; it's much easier if you have three distinct dice to roll. But the lack of such needn't be a show-stopper.

Also note that changing the system from roll-under to roll-over likewise adds to the complexity of the roll: you end up adding three or four numbers (the two dice and the trait or traits) together instead of adding just two (i.e., the two dice). Granted, it doesn't add much complexity; but then, neither does the opposed dice approach.

Originally Posted by Omegonthesane

The suggestion of changing to a roll-over rather than roll-under model has more merit. In terms of operations, nothing is lost; in terms of probability, the system only needs to have numbers changed rather than any deeper reworking in order to implement such a change.

Actually, neither change requires a deep reworking of the rules. Both require you to rewrite the way that the d666 works, which can be done within a single paragraph; any other changes come from the switch to roll-over.

Originally Posted by Omegonthesane

It would not be a great idea to add or subtract things from the target number in any circumstances - it is easier to add or subtract things from the total of the dice in front of you than it is to add or subtract them from the not-in-front-of-you number that your roll is to be compared to. Even if a +4 to the roll is, in fact, equivalent to a -4 to the target number that must be met or exceeded by the roll.

You're assuming that the player does all the math; and to the extent that he does, it does indeed make more sense to adjust the total of the dice than to adjust the target number. But when the GM is doing the math (such as when he's accounting for factors that the character and possibly the player aren't even aware of), the reverse is true: it's easier to adjust the target number that's in his head or in front of him rather than the total of the stats and dice that aren't.

And if you want to go with a bit more of a freeform approach in a high-trust game, it's often much easier to have the GM just pick a target number based on his gut feeling as to what it should be instead of going through a bunch of additions and subtractions to come up with a number which may or may not reflect the actual difficulty of a task. Realistically, two factors that are each rated as a -2 penalty could conceivably overlap enough that they really ought to be treated as a net -3; conversely, you might have two such factors that compound their problems, resulting in the combination realistically being worth a -5 or worse. I'm not saying that In Nomine (or any RPG) should attempt to model this level of detail; I'm just pointing out that cumulative modifiers aren't always as trustworthy as one might suppose.

And in terms of assigning difficulty numbers on the fly: you'd need to pick a target number between 9 and 19 to cover the range of results that come from adding 2d6 to an average characteristic. With opposed dice and needing to beat the target number for success, you'd need to pick a target number between 1 and 11. It wouldn't be much of a stretch to assign the target number by asking the GM to rate the task's difficulty on a scale of one to ten, something that people do all the time.

[robkelk]

Originally Posted by Omegonthesane

Dataweaver - your description of the mechanic fails to convey the information correctly ...

Oh, good.

I have no idea what he's trying to say.