Do I look like someone who is a professional editor in real life?

From that comment, I’d have to guess… yes?

D’Oh!

http://www.silverlinewindow.com

Probably someone has already mentioned it.

My preferred scale is to see what chances does an average unskilled character have, what chances a professional has, then a true expert, than a master or legendary figure, if all of those are relevant in the game.

A table that essentially says stuff to the effect of: Easy task is difficulty Ne and unskilled character succeeds around half the time at them, standard difficulty is Ns (> Ne) and professionals succeed at it half the time. Etc. Something to that effect.

It helps a great at at adjudicating difficulties of rolls in play, which is of utmost importance if the game should work without excessive prep.

I actually guesstimate a D&D’s player’s skill at the level he is at before setting DCs of various tasks (I usually aim at 50% success before buffs) but that’s because I play with the Oblivion “Scale as you level up philosophy”.

That’s my favored style, not THE best one…
As for the table, I’d say look at the PHB’s table for listen checks.

I’d show a physical and a Mental example for each

For physical feats it could be:

Easy: Punch through a glass pane
Average: Open a stuck pot of Jam
Complex: Drive a car against Traffic
Hard: Break 4 bricks with ypour fist
Very Hard: Pilot an airplane in a canyon while dodging missiles..

A team brainstorm effort could payoff here.

Here is our work:
http://www.silvervinegames.com/uploads/Svg/dice pool probabilities.xls

That spreadshseet is up on the site if you really want to get into that much math at the Gaming table, and some people do.
http://www.silvervinegames.com/index.php/Svg/Downloads

I think it is a matter of confusion that Game Master’s would need to take into account every variable. Let me ask this: As a DM/GM do you ask what a persons total bonus to a d20 roll is? and do you raise the DC based on that?

Any multiple dice system does bring into play multiple variables, and that will get unwieldy if you try to incorporate it all to set a difficulty for the roll. I guess I’m asking is it necessary to do that in any game system, or do you go “This should be this hard, and let me find that on the scale of difficulty for the system” D&D’s scale of difficulty is 10 to 45 as the target number. Ours is 1 to 10 on the required number of successes.

I am going to rewrite that table, Chatty. I’ve actually rewritten most of that page already. Stuff like this is why we are looking for so much feedback in the editing process. We want to make sure it comes across to more than just our core audience. What do you think would make a good example for that table?

So even if the distribution of a bell curve of the 3D6 is gurps is more complicate to get, once determined it does not change. You that to have 10 or less is 50% and to hit a 18 is less then 1% probability. As a GM, I’m not adding dice to it I’m just adding an offset to a well known curve. The math of which I don’t care anymore.

So I would try to limit the amount of variable. Either play a fixed amount of dice or have a fix amount to determine success (use a variable number d6 dice for exemple with 6 being a success).

I strongly suggest that the #R table provides more empirical examples of what representative difficulties are at each level.

Then Setting difficulty, while still not my favored approach (I’m with Yan in that one)becomes easier to gauge for the DM and use.