Quote from: ggroy;730867Are these d100 rolls against a target number?  Or are they rolling against another player rolling a d100 "opposition" type roll?

Just a target number. You compare OV v. AV, and that gives you the target to roll against (My current proto table is roll low, because that makes the d100 odds clearer).

I thought about tooling around with some kind of exploding/imploding doubles trick, or even an inverse of my Drums of War trick.

On the other hand, I think if I define my range in the right spot, I can probably do just fine with a regular old d100 roll.

The other option that occurred to me is that if I set the TNs to include the possibility of explosion, then suddenly I'm not so confined by the range of any given dice. I could even steal the 2d10 w/exploding doubles from DC Heroes.

Quote from: ggroy;730872There's a simple minded way of going beyond 100 in a roll-under d% system.

The easiest case would be a player swinging a sword with A% success, and the target with a B% success at "parrying".

Essentially the idea is to preserve the relative difference between A and B, and "normalizing" it to the 0 to 99 range.  (Though this breaks down if the relative difference between A and B is greater than 100).

To take a concrete example over 100, let's take A=120% (attack) and B=90% (parrying).  One possible way is to "normalize" the higher number to "100%".

The attack vs. parrying rolls in this situation by decree could be made equivalent to:

attack roll ->  120% - (20) = 100%

parrying roll -> 90% - (20) = 70%

If one doesn't like the 100% "done deal" type of roll, one can knock off an additional 5% (or 1% or 10%, etc ...) from both attack and parrying roll success % figures.  (ie. Normalizing the higher number to "95%", (or "99%" or "90%", etc ...) respectively).



(IIRC, Mongoose Runequest did something like this).

My PCs have the option, to add a parry of 20% of their weapon skill onto their protection if they forgo an attack, + their parry skill, or a partial parry of 10% of their weapon skill+ 50% of their parry skill to protection and get a weaker attack/    What I love to do is add onto the above idea 0f attack vs dodge is making heavier armors actually harder to 'dodge/avoid' with, but letting them soak a variable amount of damage.  

And we use dividing dice for both damage and protection.
The Guy with Hardened leather?  Before skills, he's a base 16% to hit and protects ((17-d10)/(2d6/2)).
The Guy with Split Mail + a heater shield has a base 38+10=48% to be hit and a 24+16=((40-d10)/(2d6/2)) protection.  

So the first guy, before other skills, is much harder to hit, literally 1/3 the chance of being hit.  But he has a maximum of 16 protection points, and an average of 4 protection.  The second guy may get struck more, but he has a max protection of 47 and averages 12 points of protection.

(please note, it is a lower HP setting, so getting hit sucks.  And I also like giving shields their due, unlike most games)

Now, to understand this better, I'll bring out a few well known weapons.  You'll also get the idea quickly that people don't like being hit by weapons when they don't have armor.

A short sword/Gladius does 2d6+10/d8 damage, before bonuses or skills.  That's a max of 22 hits, and an average of 3 hits.  (but that potential damage, the high damage, is there, with a 12.5% chance of a 1 divider)

A Broadsword does d10+16/d6.  so it has a max damage of 26 before any bonuses, but an average of 6 hits, but with a much higher chance of rolling a '1' dividing die, 16.6%.

a 2 handed sword/greatsword does 2d8+17/d5.  so our max has moved up to 33, and our average damage has moved up to 9, with a 20% chance of a 1 divider.

So I like having a % chance to hit, mitigated by the avoidance % of the defender, with the protection+parry+combat strategy.

Here's another simple minded way of going over 100 in a roll-under d% system, using the same scenario of:  a player swinging a sword with A% success, and the target with a B% success at "parrying".

In this method, the opposition rolls would be:

attack = 50% + (A-B)
parrying = 50% + (B-A)

This system breaks down if the relative difference between A and B is greater than 50.


Essentially this isn't much more than "always fighting orcs" in the scenario where A and B are going up proportionally as the players advance in the game.


(IIRC, this is the basic idea underlying the opposition table in earlier versions of Runequest).

^

The above system in the previous post, can also be done as a single roll against a target.  (If one doesn't like opposition type rolls).

Essentially rolling-under an attack of 50% + (A-B) for a success.

The main problem with such a system, is if the DM doesn't want to players to know what the values of B are for various "targets".

ie.  The DM would have to do all these calculations on their end.

(Some thoughts on levels of abstraction in game design. This doesn't exactly fit in anywhere earlier in the thread, so bumping. Its design related, but bordering on the art of GMing in places. Thanks to Gleichman, Arminius, chaosvoyager and also Justin Alexander for discussion behind much of this post.)

A game mechanic is an abstraction of something in the game world (unless its a story-driven mechanic that exists at a purely 'meta' level). Resolution of a game world event may be resolvable either based on world details and commonsense (DM 'fiat'), or via referring to game mechanics (the abstraction).

There can be conflict between the two different approaches, for instance:
*use of Search checks (skill + die roll) to find objects vs. use of player description of how they search.
*use of what players (as characters) say, vs. using a dice roll (i.e. Charisma or diplomacy check or reaction roll).
 **a character might have a high Stealth bonus, but across open ground that would require a justification of how to apply this, given a lack of cover, or against a target definitely staring in a particular direction.
*the newbie D&D player may describe to the GM how they're going to mangle an NPC in combat, when what they can do relies on random rolls, and when HPs are largely abstract). Their action then has to be translated into game terms (unless the game resolution system is dropped entirely for that action) then the game system results translated back to details to tell the player if they succeeded.

There are sometimes hybrid approaches where a character gets a roll, but at a bonus/penalty for the specific details, or where a character defaults to game mechanics unless they perform a specific player action (e.g. 'I search the room' vs. 'I search the cloak on the hatstand in the far corner'). A die roll (game mechanics) can generate details which the players can then interact with - e.g. PCs fail to open a door (bad Str/open doors roll) and the GM says its hinges are rusted, so they use lamp oil to attempt to oil them.

Problems occur when different game mechanics are outputting at different levels i.e. if a result of an ability is more specific than the game systems allow for elsewhere, meaning it needs translation into mechanics ("an unseen servant can exert 20 pounds of force"). Likewise if a power gives a bonus without describing how it works -  a power that gives '+X Diplomacy' without a description of how it operates (raw skill ranks, 3.5 warlock power "Beguiling Presence") is hard to use if a GM is using what a player says only to determine NPC reactions without a roll, whereas a power that specifies '+X Diplomacy because the PC is now super attractive" has an effect either way. Abilities can be designed to complement a more detailed approach, either by specifying how they work exactly, or via otherwise working with the approach (the 2E D&D Riddlemaster for example gives the player an extra 'try' with the players first unsuccessful try is retconned away, or double the time limit); the 2E D&D Etiquette skill uses a dice roll, but with success or failure of the roll being used to let the DM feed a player information on what they should do in-character - use of the correct noble title or the shrimp fork.

Use of the detailed approach in GM adjudication is very common in old-school games, since many tasks will fall outside the scope of what the written rules cover. The adventure Dungeon of the Bear for Tunnels and Trolls (1982) comes to mind - it has rooms which describe traps using real-world descriptors as an adjudication aid e.g. cubic footage of water in the octopus room (to explain how poison in the water would be ineffective) or tonnage of rocks that come down at other traps, etc., to better adjudicate how effective various player tactics may be (despite other parts of T&T, e.g. combat, being extremely abstract).

A desire for more detailed rules in an RPG is perhaps driven in part by a desire to reduce the conflicts between results using abstractions and detailed adjudication (i.e. converting strange player requests to stab someone in the face into game mechanics becomes easier if there are hit location and called shot rules). To an extent more detail generates more problems, but more specific ones, such as hit location leading to having to adjudicate if characters can reach particular limbs or of partial cover.  
The approach of using detailed results is somewhat back in vogue presently in e.g. Dungeon World ("Narrative Positioning") - it has mechanics which are more comprehensive than most old school games, but deliberately vague. However, for the most part game design has become more and more abstract, with more attempts to increase simplification and handle of more situations by a single die roll, such as in 4E D&Ds use of 'skill challenges' to handle a majority of non-combat events. Another factor is the trend toward narrative mechanics.

Translating between concrete and abstract
Mechanics can represent things that are fairly tangible in the game world (Strength, number of arrows you own), or that are hard to measure, multifactorial or intangible (Wisdom, morality, how much an NPC likes you). Inconsistencies are more likely with 'near' abstractions (the Str-6 wizard tears down a door the fighter failed to open earlier), while the GM just being stumped as to how to interpret a value for something is perhaps more likely with intangibles (a PC loses only 20% of their Neutral Good alignment, thanks to a ring of Spell Turning reflecting most of the BBEG's Morality Undone spell). In the second case, inconsistencies are harder to spot, so one approach to removing apparent inconsistencies is to make game mechanics more abstract i.e. having values that are broader and represent more things, though this may also make 'translation' more difficult.

Interesting applications of game mechanics involve taking an abstraction and translating it back into world details in an unusual way. One Pathfinder adventure path for instance has PCs waking from being poisoned, with characters recovering in the order of highest-to-lowest Fortitude saving throw result (interpreting abstract values ordinally). Or, a player might use a hit location system to random-roll the placement of a tattoo or mole (re-applying a mechanic in a different but slightly related context).

Common translation issues between mechanics and details, IMHO, might include:

*die rolls generating too much randomness. In this case the GM has to strain their imagination to create a description of how something would've failed or critically failed (the moral being - don't roll a check unless its something a character could fail).
How much randomness is appropriate for a task can vary depending on how much detail about the events has been predetermined, but game mechanics are not usually so flexible. (Having a seduction check on a random guard fail because they're gay may be fine since its preferences are previously unspecified, but the same result might not be reasonable against a named NPC). (As an  aside: tables of modifiers can give a clue to how the designer thought when they designed a game. In one game you might have a modifier to your roll before the check for something that would be generated as a description for a bad roll in another).
*Assuming random factors are largely "external", causing inconsistencies if a task can be re-rolled, or attempted again by a different character (e.g. again if the Str 6 wizard busts down the door that the Str 15 fighter couldn't open earlier).
*assigning a failure point to something using player skill predominantly. A GM may have to veto a player saying something/doing something because of a bad roll ('you didn't/couldn't think of that').
*mechanics reporting incomplete success, when task seems complete according to the details (but with the GM going based on mechanics). The players no longer have a working description of the scenario and can't determine how to continue to solve the problem.
* 'Double Jeopardy', where a player is required to both describe IRL exactly how they're doing something and pass a check as a character. This gives twice as many failure points (you can fail either due to mispronouncing the NPCs name as a player or by rolling a "1" on Diplomacy), or can lead to checks that fail for no discernable reason, like failing a Search roll on a pocket.
*"simulation of process" failure - the mechanics modelling the underlying factors do so poorly - see Gleichman's description.
http://www.therpgsite.com/showthread.php?t=26166
*untranslateable quantities (e.g. abstract bookkeeping mechanics). Some abstract mechanics are black holes that can't be looked inside without breaking the abstraction - e.g. an 'ammo' value in Dungeon World replaces detailed tracking, and then can't be collapsed back into an exact number of arrows while still working with the Volley move as-is. Not necessarily bad maybe, but limited in that respect. Abstract bookkeeping can also suffer from repeated rounding errors e.g. Wealth systems [D20 Modern] that result in characters being able to buy infinite numbers of low-cost items.
*players working primarily at a mechanical level, without needing to engage with world details. (cf. Vincent Baker's comments in podcast on "Trait Invocation" games coming out after DITV, where "the fiction" isn't modified by a trait being involved: http://theoryfromthecloset.com/2010/08/19/show059-interview-with-vincent-baker/

Over time, games' combat systems can be pushed to evolve from fairly abstract into detailed - perhaps D&D, or Shadowrun across editions - with rules that as a result of getting more complex character options (e.g. incorporating detailed martial arts moves) start modelling exact exchanges/strikes and so getting weird.


EDIT NOTE
Placing this here for the time being because its a high-level consideration which doesn't seem to fit anywhere else:

Quote from: John Morrow;196300But the bottom line is that I think that given a choice, a game designer should work with representational models over non-representation models unless there is a significant reason to prefer the latter, and I think it's no mistake that the biggest points of complaint about D&D -- armor class, hit points, and levels -- are the areas where the abstraction is least representational.

Another interesting note praising higher 'abstraction level' saving throws - with the theory a player can make something up to describe what's going on if what's going on isn't clearly specified in the reductive way.
http://hackslashmaster.blogspot.com.au/2010/12/on-abstraction.html

EDIT NOTE 2:
Another interesting systematic use of abstraction is in the 'tech noir' RPG (e.g. see the player's guide here:
Quoting Justin Alexander here-

QuoteTechnoir found a completely new (and truly compelling) method for action resolution by using Verbs (which describe what a character is good at) to push Adjectives (which describe the outcome of their action) onto a target. Once this is coupled with the mechanics which determine how the improvised adjectives affect the target, the result is incredibly compelling. (Technoir also features awesomely innovative plot-mapping mechanics.)

IMHO 'verbs' (Fight, Shoot, etc.) appear not too different from normal 'skills' or the like, though the 'adjectives' were interesting. In a sense perhaps Tech Noir's system is reminiscent of how 'effect based' powers systems work, but generalized to a more fundamental level.

Despite the issues, abstractions can sometimes be much easier to work with that 'real' things. For instance, the original 'kender pockets table' tells exactly what a kender pulls out - from string to feathers to anything really. The intent of the ability is really to see if the kender can produce 'something' relevant to the situation, so this could be replaced by just having a roll vs. a base % chance (adjusted by the GM for how specific what they require is), avoiding the need for a d100 table of results.

A lot of this depends on how the abilities are defined. Not just "Broad" and "Narrow", but edge or center-defined, or whether they overlap. (Ref.)

-clash

Thanks FM. I love the definitions there.
Bonus points if you can tie up how 'centre-based' or 'overlapping' relate to abstract-ness? (I think you have something there, but its a bit fuzzy to me as yet).

Never forget, the level of abstraction can also aid the campaign setting specificity.  Rules are often representations of setting specific events or differences.  One of the best ways to buttress the ways your setting is unusual is to build the system to support and showcase it.

Center-defined abilities are inherently abstract - by their nature their definitions can be adjusted to suit the circumstances and your taste. You need to interpret them to use them.

When abilities overlap, the abstractness is more subtle - there are several ways to approach the situation - the GM says "You need to decide where to place the ambush. Make a Tactics check", and the player says "I'm not real good with Tactics... can I use Evaluate instead? Maybe it can help me decide which course is worth more?" The abstractness is in the nature of an ability. What information the ability returns may not be the same, but in essence, the check called for by the GM is just a placeholder, an example of the kind of thing that will work, rather than a concrete thing-in-itself. The important thing is not the ability, but the result - what will give the player the information needed to resolve the question.

Quote from: flyingmice;753632Center-defined abilities are inherently abstract - by their nature their definitions can be adjusted to suit the circumstances and your taste. You need to interpret them to use them.

When abilities overlap, the abstractness is more subtle - there are several ways to approach the situation - the GM says "You need to decide where to place the ambush. Make a Tactics check", and the player says "I'm not real good with Tactics... can I use Evaluate instead? Maybe it can help me decide which course is worth more?" The abstractness is in the nature of an ability. What information the ability returns may not be the same, but in essence, the check called for by the GM is just a placeholder, an example of the kind of thing that will work, rather than a concrete thing-in-itself. The important thing is not the ability, but the result - what will give the player the information needed to resolve the question.

In terms of where my thinking's at currently I think either sort of roll - Tactics or Evaluate - is a game mechanic and hence an abstraction to some extent. Evaluate may be further from reality, I guess. To an extent you could use either skill for Tactics in different ways depending on the level of abstraction you're working at - as in, whether the GM has laid out the out the lay of the land in detail and has to interpret a Tactics success by working out themselves where the enemy will attack and telling the PCs, vs. just telling the PCs the attack is going ahead and giving the PCs a to-hit bonus based off their Tactics roll.

In general though, I would agree that yeah, a certain level of detail would probably generate uses for niche skills that aren't immediately apparent (pointless to determine if all the relevant situations are covered by the same skill anyway - overlapping), or conversely detailed or edge-defined skills will usually force the GM to make decisions about details to determine if the skill is applicable at all.

Quote from: flyingmice;753632Center-defined abilities are inherently abstract - by their nature their definitions can be adjusted to suit the circumstances and your taste. You need to interpret them to use them.

When abilities overlap, the abstractness is more subtle - there are several ways to approach the situation - the GM says "You need to decide where to place the ambush. Make a Tactics check", and the player says "I'm not real good with Tactics... can I use Evaluate instead? Maybe it can help me decide which course is worth more?" The abstractness is in the nature of an ability. What information the ability returns may not be the same, but in essence, the check called for by the GM is just a placeholder, an example of the kind of thing that will work, rather than a concrete thing-in-itself. The important thing is not the ability, but the result - what will give the player the information needed to resolve the question.

Interesting way to see it.
instead of a simple system and resolution end goal, there are many ways to reach a system goal.
But I've always encouraged exactly that.  I even set up rules for adjudicating additive skill use

Quote from: LordVreeg;753828Interesting way to see it.
instead of a simple system and resolution end goal, there are many ways to reach a system goal.
But I've always encouraged exactly that.  I even set up rules for adjudicating additive skill use

These things are stylistic differences that I gave names to, not things I invented. They go back to the very early days of the hobby, but no-body ever named them or talked about them so far as I know.

Quote from: flyingmice;753849These things are stylistic differences that I gave names to, not things I invented. They go back to the very early days of the hobby, but no-body ever named them or talked about them so far as I know.

no...don't think so.  Good job.  I use this all the time, so I appreciate it.

Yep centers & edges is a good topic.