Well, looking at the theoretical range of skill modifiers shows, where the problem lies:
Worst case: Paladin with Dex 8 trying to be stealthy while wearing plate and a heavy shield:
At level 1: -5; at level 30: +11
Best case: Halfling with Dex 20 and Skill Focus (Stealth) trying to be stealthy:
At level 1: +15; at level 30: +34
In other words skill modifiers at any given level vary by as much as 20 to 23! Imho, that makes it pretty obvious why it's not really possible choose balanced DCs for skill checks.
To arrive at a more realistic range, I decided to create what I'd consider to be an average party:
[sblock]
Code:
Race Class Str Con Dex Int Wis Cha Trained Skills Armor Penalty:
Dwarf Cleric 14 14 10 11 18 13 Diplomacy Heal Insight Religion - - Chain -1
Dragonborn Fighter 18 13 14 10 11 14 Athletics Endurance Intimidate - - - Hvy. Shield -1
Halfling Rogue 13 12 18 10 11 16 Acrobatics Bluff Intimidate Stealth Streetwise Thievery Leather
Eladrin Wizard 10 12 15 18 14 11 Arcana History Insight Perception Religion - Cloth
Elf Ranger 14 12 18 11 15 10 Acrobatics Dungeons Heal Nature Perception - Hide -1
[/sblock]Note that these aren't really optimized. The only thing I made sure was that every skill is trained for at least a single character.
For a skill challenge it is safe to assume that only those character that are best at a given skill will participate (the rest will probably use Aid Another).
So, the next table shows this party's best skill modifier over 30 levels:
[sblock]
Code:
Trained Level
Skills 2 5 8 11 14 17 20 23 26 29
Acrobatics 12 13 16 17 20 21 23 25 27 29
Arcana 12 13 16 17 20 21 23 25 27 29
Athletics 9 10 13 14 17 18 20 22 24 26
Bluff 9 10 13 14 17 18 20 22 24 26
Diplomacy 7 8 10 12 14 15 17 18 20 21
Dungeons 8 9 11 13 15 16 18 20 22 24
Endurance 6 7 9 11 13 14 16 17 19 20
Heal 10 11 14 15 18 19 21 23 25 27
History 12 13 16 17 20 21 23 25 27 29
Insight 10 11 14 15 18 19 21 23 25 27
Intimidate 10 11 13 15 17 18 20 22 24 26
Nature 10 11 13 15 17 18 20 21 23 24
Perception 10 11 13 14 16 17 19 21 23 25
Religion 10 11 14 15 18 19 21 23 25 27
Stealth 10 11 14 15 18 19 21 23 25 27
Streetwise 9 10 13 14 17 18 20 22 24 26
Thievery 12 13 16 17 20 21 23 25 27 29
Average 9,88 11,06 13,88 15,35 18,18 19,35 21,53 23,53 25,71 27,71
[/sblock]Note that these skill modifiers don't include bonuses from feats, powers or items.
Compared to the suggested skill DCs, you'll get the following probabilities to succeed in a single skill check:
[sblock]
Code:
% 2 5 8 11 14 17 20 23 26 29 Average
Easy 100,00% 100,00% 100,00% 100,00% 100,00% 100,00% 100,00% 100,00% 100,00% 100,00% 100,00%
Medium 100,00% 100,00% 100,00% 100,00% 100,00% 100,00% 100,00% 100,00% 100,00% 100,00% 100,00%
Hard 79,41% 75,29% 79,41% 76,76% 80,88% 76,76% 77,65% 77,65% 78,53% 78,53% 78,09%
[/sblock]
If all of the DCs are increased by 5, you'll get this:
[sblock]
Code:
% 2 5 8 11 14 17 20 23 26 29 Average
Easy 100,00% 100,00% 100,00% 100,00% 100,00% 100,00% 100,00% 100,00% 100,00% 100,00% 100,00%
Medium 79,41% 75,29% 79,41% 76,76% 80,88% 76,76% 77,65% 77,65% 78,53% 78,53% 78,09%
Hard 54,41% 50,29% 54,41% 51,76% 55,88% 51,76% 52,65% 52,65% 53,53% 53,53% 53,09%
[/sblock]
If all of the DCs are increased by 10 you'll get this:
[sblock]
Code:
% 2 5 8 11 14 17 20 23 26 29 Average
Easy 79,41% 75,29% 84,41% 81,76% 90,88% 86,76% 92,65% 92,65% 98,53% 98,53% 88,09%
Medium 54,41% 50,29% 54,41% 51,76% 55,88% 51,76% 52,65% 52,65% 53,53% 53,53% 53,09%
Hard 29,41% 25,29% 29,41% 26,76% 30,88% 26,76% 27,65% 27,65% 28,53% 28,53% 28,09%
[/sblock]It's now fairly easy to compare these results to the chances to succeed at a given skill challenge (taking Aid Another into account as appropriate).
To me, it seems the new DCs have been chosen to allow everyone to have a chance to succeed with every skill at every level:
- Easy DC:
Someone untrained will have an above average chance to succeed (even if handicapped).
Someone trained will succeed automatically (unless handicapped).
- Medium DC:
Someone untrained will have an average chance to succeed (lower if handicapped).
Someone trained will have a very good chance to succeed (above average, if handicapped).
- Hard DC:
Someone untrained will have a low chance to succeed (no chance if handicapped).
Someone trained will have an above average chance to succeed (even if handicapped).
If you agree with this intention the DCs are okay to use for single skill checks.
For skill challenges to be meaningful, however, it's not possible to use the same DCs:
- Even if every skill check is using the Hard DC and noone is using Aid Another, a Complexitiy 5 skill challenge will succeed with an average chance of 48,86%.
- If every skill check is assisted by one character, the chance goes up to 83,33%
- If every skill check is assisted by two characters, the chance goes up to 99,85%