Re: 4e Pros & Cons

Jasin Zujovic wrote:
Justin Alexander wrote:
This is actually one of my primary problems with skill challenges (the
completely fucked up probabilities inherent in the system no matter
how much you errata it being the other):

Eh? This sounds a bit like "the mechanic doesn't work even if they
change it so it's a completely different mechanic", and makes me
suspicious that you're have to much of a hate on for this to be objective.

While you could theoretically rewrite the rules so that the term
"skill challenge" applied to a completely different mechanic, that's
clearly not what I meant.

The probability problems are inherent in:

(a) Counting successes-vs-failures to determine overall success of a
task; and
(b) Asserting that all potential solutions to a given problem have the
same number of discrete steps.

The former is the fundamental basis of the mechanic. The latter could
theoretically be removed, but if you did so it would make the entire
mechanic completely pointless.

(1) In a worst case scenario they're illogical railroads.
(2) In a best case scenario they're illogical wastes of time.
(3) And the probabilities don't work.

The probabilities actually seem to work pretty well if you use errata
DCs + 5.

That's only one of the many different probability problems the system
suffers from.

No matter what you errata the DCs to, for example, the fundamental
nature of "count successes" as a mechanic creates a steep probability
curve -- small variations in the modifier for any single skill check
will result in large changes in the odds of success. This problem has
been dampened somewhat by the errata (because they no longer vary the
number of failures), but it cannot be eliminated. It's inherent in the
mechanic.

(This is why dice pool games have increasingly moved towards models in
which "the target number is always the same" in order to fix the type
of probability issues that, for example, the original Storyteller
system had.)

This steep probability curve can be useful in modeling specific types
of tasks -- generally complex tasks in which there are multiple steps
to success. Defusing a complex bomb, for example. Or winning a game of
chess. Which is why lots of games use a similar mechanic for complex
skill checks. (And I've house-ruled similar mechanics into D&D.)

But they're notably problematic for the stated design goals of skill
challenges, in which many different skills can be used by many
different characters. Why? Because skill challenges are supposed to
get everybody at the table involved in the task. But they don't do
that. In fact, they do exactly the opposite: They heavily reward and
encourage the character with the best applicable modifier to make all
the checks.

The other problem with the system is the probability effects of
asserting that all solutions to the problem have the same number of
discrete steps (i.e., required successes on a skill check). The only
way to actually achieve that is to vary the amount of detail handled
by a single check depending on not only the specific skill challenge,
but the specific solution to the skill challenge being attempted. This
causes huge probabilistic swings in the likelihood of the individual
steps being handled successfully depending on what skill challenge
you're attempting while making the checks.

In my games, the players tell
the DM what they're going to do, not the other way around.

When they casts spells, don't they select from a limited menu of
options: fireball or lightning bolt or haste? When they fight with
swords, don't they do the same: Combat Expertise or Power Attack?

Those are class abilities. Which, notably, the players choose for
themselves.

What's
wrong with handling negotiations or wilderness survival or street chases
in a similar manner?

And even if they were pregenerated characters, there's nothing
particularly similar between the two. There is a fundamental
difference between:

(1) I have resources X, Y, and Z... how can I use these to overcome
the obstacle in front of me?
(2) You have an obstacle in front of you... do you want to do A, B, or
C to overcome it?

The former is an open-ended, player-driven scenario. The latter is a
DM-driven railroad -- basically a Choose Your Own Adventure book with
some dice rolling tacked on.

--
Justin Alexander
http://www.thealexandrian.net
.