> My point, and I think Julian's point, is that it is unclear how to
> implement the narrative motif of
> slaying-a-dragon-to-marry-the-princess.
> ... this is a 'narrative' system ... It is therefore proper to ask
> how well the system allows various narrative motifs.
This is the source of many of the most lasting discussions on the list: "I want to tell a story in which foo happens -- how do I go about modeling that in terms of game actions?" For example, the archery thread: "It is important to my story that the opposition has ranged weapons while the heroes do not. What are my options for expressing this situation in game mechanics, and what are the ramifications of each?"
Now, one could say, as the rules do, that the choice between simple and extended contests ought to be made purely on the grounds of "how much screen time would it get," but unfortunately that doesn't actually work very well, because the outcomes are so different. For example, in message 9153 (Tue May 8, 2001 9:45 pm) Roderick Robertson wrote:
> Simple contests are for one- shot resolution, either because the
> action is over in an instant (like dodging a falling boulder), or
> aren't considered important enough to warrant extended rolling
> ("okay, you have 15w2 Close combat, the guards are 17 < quick roll>
> - you slaughter them in heaps on your way to the Prince's throne
> room, no-one manages to touch you").
Unfortunately, though this ought to be true, it isn't. He will slaughter them in heaps only in an *extended* contest. In a *simple* contest, what happens is:
Actor Skill: 55 Opponent Skill: 17
Complete victory: 4.75 %
Major victory: 9.75 %
Minor victory: 76.5 %
Marginal victory: 0 % Any victory: 91 %
Tie: 4.75 %
Marginal defeat: 4 % Any defeat: 4.25 %
Minor defeat: 0.25 %
Major defeat: 0 %
Complete defeat: 0 %
So less than 1 time in 20 are they killed; more than three times in four they will only be lightly scratched, and they will actually *defeat* the hero almost as likely as be killed by him. Also note this is one of those situations where gaining another mastery badly hurts the underdog: raise the guards to 1w, and while their chance of beating the hero does increase slightly, their chance of being seriously Injured instead of merely Hurt increases by a factor of seven!
Actor Skill: 55 Opponent Skill: 21 Complete victory: 3.75 % Major victory: 68.5 % Minor victory: 22 % Marginal victory: 0 % Any victory: 94.25 % Tie: 0 % Marginal defeat: 5.5 % Any defeat: 5.75 % Minor defeat: 0.25 % Major defeat: 0 % Complete defeat: 0 %
Furthermore, if the guards' skill is set at 20, they CANNOT be killed by a simple contest, no matter whether the hero's skill is 15 or 15w10! Similarly, no matter how high a shaman's Spirit Combat is, even if high enough for him to integrate a *god*, he can never integrate a spirit of might 20 if a simple contest is chosen as the resolution mechanism, because the might 20 spirit cannot fumble, no matter the shaman's power, but a might 10w10 spirit could, if you spend enough HP and do a big enough ritual to match its might.
This is why, as Frank Rafaelsen put it (in message 4356, on Wed Jul 26, 2000 9:55 am), "No sane hero would want his contests to be resolved as simple contests because of the very remote chance of getting a result better than minor victory." I wholeheartedly agree, adding that the chance of something going badly wrong in a simple contest is much greater than in an extended one. This is the opposite of what the rule is supposed to encourage: things that have no great dramatic importance ought to be resolved quickly. A worthy goal! Unfortunately, the implementation chosen creates the very strange result that things "of no importance" are *harder* to do than ones that are! How's that for anti-climax? It seems clear to me from this that simple contests, as written, need a thoroughgoing revision before they can sensibly be used for their intended purpose.
At a minimum, I think the changes need to include these principles:
Principle 1 implies to me that the most sensible way to resolve any contest involves not the two contestants' abilities separately, but rather the single *difference* between them. This has the added benefit that bonuses to your ability become the same as penalties to the resistance, and vice versa; the fact that these are not the same in the official rules, in fact that whether a modifier is a +5 to your ability or a -5 to the resistance is of paramount importance, I regard as clearly a flaw. That together with the fact that masteries cancel but target numbers don't leads to very weird logic that once a liturgist's Lead Prayer is high enough (probably by 10w2, definitely by 20w2), he should consult the various advanced magic tables to find a set of adverse/challenging conditions which generate a modifier between +7 and +19, because applying it will *increase* the expected benefit of the blessing even though the task is theoretically harder! The rules ought not to *encourage* silly metagaming, which in this fashion the current contest (especially simple contest) rules do. Making bump-*downs* a standard component of the game could do a lot to alleviate this problem, because much of it stems from the strange property that once you've got 3 masteries on someone, further increases in your ability are totally irrelevant: the only thing that matters is how close their target number (Note: *not* their ability) is to 20. However, I prefer the subtraction method because it smooths out all the mastery-level boundary effects, rather than just many.
Subtraction has been suggested before, in the context of variant augmentation rules; recently the idea has been attributed to David Cake, while the first time I recall it mentioned in the archives was by Andrew Graham in message 3428 -- likely several people have (and will continue to) come up with it independently. The standard way of putting this flavor of alternate augment is: roll an ability test of your ability minus the resistance.
Let's consider this from the viewpoint of making a new way to resolve all simple contests, not just augmentations. It is good because it cuts the number of die rolls in half, and prevents ties (IMO a tie in a simple contest other than an augmentation attempt often results in having to roll again). It also prevents both sides from losing, which is something I never liked either. But there is one change we need to make to this suggestion: when comparing abilities, equal skills should have equal chances of success. Since that happens on a test of an ability with value 10, the mechanic I propose to replace simple contests is: do an ability test of actor skill - resistance + 10.
This is even faster than before, because you roll only one die rather than two while cross-indexing their results, at the cost of doing a small amount of arithmetic. It is also in better agreement with the results of extended contests. If skills are equal, each side has a 5% chance of complete victory, 45% chance of minor victory, 45% chance of minor defeat, and 5% chance of complete defeat. The side with a 10-point advantage wins 95% of the time, but needs a 30-point advantage to win 100% of the time (a complete victory 95%).
In the case of Roderick's hero on the way to the throne room, do an ability test of 55 - 17 + 10 = 48; 8w2 criticals 95% of the time, so with this resolution method he really can slaughter the guards without breaking a sweat, which is what the simple contest rules are supposed to do but currently don't.
The only real problem with this that I see is that it never generates major or marginal results, only complete or minor ones -- but then, that's because it's taken directly from the ability test table, which has the same structure. The only ways around this that have so far occurred to me involve modifying the result based either on how close you rolled to the target number, or on the result of a second die roll -- both of which reintroduce things I was trying to avoid, namely weird breakpoint effects or rolling two dice and cross-indexing -- or trying to do something with how many uncancelled masteries didn't get used as bumps -- which seems confusing at best. Any suggestions?
Now, how you think the results of an augmentation contest ought to be interpreted is a separate question, but I will mention in passing that this method would produce slightly higher augmentations across the board, which are most helpful at the lowest skill levels, which I think could use the help. For example, one species of deer in Anaxial's Roster is supposed to try to augment with an ability of just *10*, which is either useless or foolish under standard rules.
Ryan Caveney
Powered by hypermail