I actually wrote this a while back, but for some reason I never got around to posting about it here.

But since I'm actually having a roulade for dessert this evening, I thought I'd post a link to a semi-generic game I wrote recently called, simply enough, ROULADE.

https://drive.google.com/file/d/0B3hWDeEiNI5aRWF1QjVqY25lTU0/edit?usp=sharing

It was sort of written with a host of inspirations in mind, an odd hybrid of DC Heroes, Fudge/FATE, a hint of GURPS (or rather lighter games of the same type), and probably a bit of Risus even.  It's got some cool ideas in it I like:

• The layer/scaling system
• A universal framework for action and conflict sequences
• Freeform chargen that's still very flexible.
• A humorous and surprisingly handy design metaphor involving food

It's also got some stuff that might not be so popular:

• A table-based resolution system (though there's an optional formula method)
• Technically no involuntary character death
• It's point-based chargen
• I'm not sure if the action sequence framework won't over-simplify

I'm still considering it kind of a work-in-progress side project, so I'm kinda curious to hear any feedback you'd like to give. It's free, and it's only 36 pages, so it should be a quick read, and I'm happy to answer any questions about it.

Something I think this still needs maybe is some kind of 'Maneuvers' or even 'Stunts' mechanic for action and conflict sequences.

As it stands, the structure of the action resolution system is very flexible, but the actual mechanical acts involved are pretty much boiled down to rolling the dice for the thing you want to do, and hoping you don't fail. This isn't necessarily bad: it kinda puts the onus on the players to describe the action well, with the mechanics just there to figure out who succeeds.

But I'm thinking some kind of tactical option system might be needed for those who want a sequence to be a bit more mechanically interesting. At the very least, some kind of 'try a special thing, maybe get a bonus' would be in order.

Any ideas?

Also, I've long thought of dropping the table system. The structure of it barely justifies the use of a table, and while that's enough for me, and I like tables, I also feel like something faster might suit the rest of the game better.

Trouble is I couldn't really think of one that still gave the clear difference between layers. The 4d6 curve actually works really well for what the layer system is trying to do, making layer difference just right in the sweet spot of 'fuckin hard/easy' without either being impossible, or not impossible enough. And using column shifts, more or less, helped to differentiate talents v. training in the mechanics.

I just haven't been able to think of anything that pairs up right with my design goals.

Sort of weird. So acting value is ignored for higher-layer attacks, opposing value is ignored for lower-layer attacks? I don't know if the difficulty change is enough for going up or down layers.

If I understand correctly:
Vanilla-1 vs. Strawberry 1 = L+2 = 16+   --same odds as an opposing value of 3 in the same layer ?

Vanilla 1 vs. Chocolate 1 = L+1 = 15+ = same as an opposing value of 2 ?


This seems to cause weird things to happen with e.g. the Difficulty rules - if you have a low acting value an Inconceivable (LA +2) task is easier than a Incredible Task (Difficulty of 8 at the same layer). Or if you have a very high rating, an Effortless (LA-2) task could be harder than a Very Easy task (opposing rating of 1).

Quote from: Bloody Stupid Johnson;730794Sort of weird. So acting value is ignored for higher-layer attacks, opposing value is ignored for lower-layer attacks? I don't know if the difficulty change is enough for going up or down layers.

Essentially, yes. If you're attacking something higher in layer than you, you're swinging out of your weight class: think Batman punching Superman.

If you're attacking someone lower than you, you're punching down: Superman is now punching Batman.

They're essentially in totally different leagues, power-level wise, so in Batman's case the only thing that's gonna save his ass is all that training he's got (probably lots of +3s). Natural capability (ie. SPECS and Talents), aren't really relevant at that point because no amount of simple talent is enough for me to punch a human tank like Supes.

Such is the idea anyhow. It's not meant to be truly impossible, just really hard, and the 4d6 curve is sharp enough that it more or less hit the odds I liked. The idea was to allow Batman, Spiderman, and Superman to interact in the same system, while still able to meaningfully affect each other, but also without the math going bananas.

QuoteIf I understand correctly:
Vanilla-1 vs. Strawberry 1 = L+2 = 16+   --same odds as an opposing value of 3 in the same layer ?

Vanilla 1 vs. Chocolate 1 = L+1 = 15+ = same as an opposing value of 2 ?


This seems to cause weird things to happen with e.g. the Difficulty rules - if you have a low acting value an Inconceivable (LA +2) task is easier than a Incredible Task (Difficulty of 8 at the same layer).

Hmm. Took me a moment, but I think I get what you mean. And that is kind of a problem (one I can't believe didn't occur to me).

There are ways to explain that effect thematically, but they'd be justifications. IE. beating someone competent in your own class is still going to be harder than beating someone incompetent of a higher one.

From a certain point of view though, I can see how it makes more sense for higher layer OVs to essentially start with a target number that's already higher than an in-level 8.  the 4d6 range doesn't really give room for that though. More argument for an alternate die mechanic, maybe?

QuoteOr if you have a very high rating, an Effortless (LA-2) task could be harder than a Very Easy task (opposing rating of 1).

This on the other hand doesn't appear to be true, since the opposing value in that case is still going to be coming from the leftmost, lowest value column.

An AV Ch8 vs. an OV of Ch1 needs a 7+, but an AV of ChX vs. an OV of VaX always needs a 6+, because the OV doesn't matter if acting against something below your layer.

You're right that fighting someone of a lower layer works, my bad.
Maybe rolling up as you say. Difficult.

It would work as long as the chance of success vs. [layer+1] is more than the highest same-layer opposing value. With d100 you could get less granular than with the 4d6 and just have only a few percent shift between each category, without definitely needing any rolling up.

If you do start setting target numbers above 100 and using rolling up, then the % chance of breaking the layer might end up just equal to the rolling up percentage, unless someone has a skill. Depending on how much you want skills to help out, you could have them increasing chances of rolling up, or give them extra shifts/plusses, depending on how much you want to help vs. same-layer vs. higher layer.

(EDIT: sorry just to elaborate on that, for instance in rolemaster you would roll up on 96-100 by +d100. If you had a target number of 150 say, then you need to roll up to get there so skill helps very little - a +50 skill only increases your actual chance of success from 2.5% to 5%. The lower the rolling up, the more skill helps.)

Here's the prototype table I came up with:



Obviously, if I switch to this, a few things will have to be redefined slightly, mainly the effect of Training and Talents. Possibly I can go back to making better use of the column shift idea from earlier drafts.

Here's version 1.1, with the new resolution table and a few other fixes. No maneuver/stunt system yet because I haven't figured out how I want to do that.

https://drive.google.com/file/d/0B3hWDeEiNI5aNnctbEx6bERWeFU/edit?usp=sharing

New table's good. I've had a very quick look at 1.1 again. You're probably still modifying (and a core mechanic replacement is always pretty gruesome), but I did notice that in 3.4.2 (resisting strain) and 3.4.7 (healing powers) there's references to the old mechanic e.g. extra healing for 1s/6s rolled.
Also, how does layer affect strain limit?

On stunts, I'm assume you'd still roll against target OV to do most of them (maybe a different OV value for some if its targets an unusual stat, whether wrestling or taunting), perhaps with a bonus or penalty? DC Heroes could be a good place to steal from, you could have a manuever like 'going for broke' that's an action test penalty but a damage test bonus (which could cover either called shots or 'power attack' or whatever), a multi-attack option, some special effects like trip/disarm in place of damage?

Quote from: Bloody Stupid Johnson;730936New table's good. I've had a very quick look at 1.1 again. You're probably still modifying (and a core mechanic replacement is always pretty gruesome), but I did notice that in 3.4.2 (resisting strain) and 3.4.7 (healing powers) there's references to the old mechanic e.g. extra healing for 1s/6s rolled.

Yeah, I just noticed that.

I'm not 100% sure I know what to replace it with either. It was a lot of hard thinking that lead to the original solution in the first place. And the d100 doesn't give me the kind of introspection the old rule did, so I'm gonna have to come up with a new one, I guess.

Maybe Damage - Resistance, min 1, with a doubling on doubles?

QuoteAlso, how does layer affect strain limit?

It doesn't.  Since it's such an abstract value already, and since the ability of a character to even take Strain is already heavily affected by layer, I'm not sure it actually matters much. It's just an average that's pretty easy to define, basically, whereas factoring in differing layers between the two would be quite sticky indeed.

QuoteOn stunts, I'm assume you'd still roll against target OV to do most of them (maybe a different OV value for some if its targets an unusual stat, whether wrestling or taunting), perhaps with a bonus or penalty? DC Heroes could be a good place to steal from, you could have a manuever like 'going for broke' that's an action test penalty but a damage test bonus (which could cover either called shots or 'power attack' or whatever), a multi-attack option, some special effects like trip/disarm in place of damage?

Well, conflict is already pretty abstract, so more specific effects don't seem to suit.

What I have been thinking about is sort of like your first suggestion: trading OV for RV in Attacks and vice versa.

Basically the idea is that you can declare a maneuver (describing it of course), and can thus trade between the two values on the coming attack. But the opponent, if they have yet to act, can forgo their next action to counter-maneuver, making an immediate check against the attacker's skill to cancel the effect.

The counter effect though, would give something of an advantage to slower characters.

I've uploaded a fixed 1.1. The old 6s/1s rule has been replaced with a simple success/failure threshold trigger.

And now for 1.2, I've had a go at replacing the Rule of 20 with an actual table for resolving Strain: https://drive.google.com/file/d/0B3hWDeEiNI5aMnowMmFHRmFjRTQ/edit?usp=sharing

It's a bit of a hack, and I guess since it's no longer used for everything, the UAT is no longer all that 'universal', but it also removes some other kludgy bits and I don't have to rely on stupid dice tricks for damage results anymore.

So I'm actually in the brainstorming phases now of a new project which I think is going to be the flagship game for Roulade, or at least born from it.

As such I've been thinking a bit more about further revisions and refinements to the system.

For starters, talents and training suck. They feel tacked on and overly minor. I mean, I like the concepts they exist to replicate, but I feel like the implementation is awkward. Bonuses like this just don't work well in a system this granular.

So the proposed change now is to change out training from a bonus, to a rating, essentially turning it into a proper skill. It would be cheaper than SPECS, and essentially provide a way to be good at one specific thing, better than one is at the original stat.

Talents are trickier. I'm seriously considering dropping them entirely. In intent they were supposed to replicate advantages/perks/etc., but they just wind up feeling piddly. I thought perhaps of trading them out for 'tricks' of some kind, essentially working in pre-formed stunts, able to trade rating in one thing for another for certain kinds of actions (like -1 AV to hit, for +1 EV for damage). But I'd kinda like there to be a mechanic for just doing that stuff anyway.  

I'm also thinking I need to at least optionally make the Strain Limit pool additive instead of average, to make combat less lethal, and I'm not 100% sure I like how Destiny damage soaking works now either. I think maybe it should just cost 1 Destiny on a fail, but I'm not sure.

A question on the table (and I like tables):

You have L+1 and L+2 listed on the action table under acting value, yet the chances are less than for 1-8, while L-1 and L-2 under opposing value are easier to act against.

Assuming L+X refers to scales higher than 1-8 (so L+1 would be L+2 in relation to L-1), then the numbers should show an easier chance of success than for 8. The table would work if L+1 and L+2 were moved to the opposing value columns, with the understanding that the acting agent is always operating at their scale, and the opposing agent is scaled in relation to the acting.

So if a dragon 4 is L+2 compared to a dwarf 6, the dwarf would roll on the 6 row against L+2 and the dragon on the 4 row against L-2.

I think it would be easier to use the table with this change.

*****

A consideration in favor of the 4d6 table (though I would use 4d6-4, or 4d6 reading the 6s as 0 for a 0-20 range) is that it makes it dead easy to use bonus and penalty dice. So the same dwarf 6 with dragon proof armor and a spear of dragon slaying would be rolling 6 dice and keeping the best 4, while the dragon 4 would be rolling 6 dice and keeping the worst 4.

L +/- X references are always from the POV of the acting character.

So an acting character attacking someone above their scale uses the L+X row, because the target is their Layer plus however many, thus lowering the chance of success. Ditto the strain table.