Reconsidering Level Bonus: Difficulty Classes

The change I am considering to the Level Bonus (make it equal to level rather than level/2) actually has less impact on skill check DCs than it might have seemed.

Evaluation by Consideration

This change results in a difference of +12 in the Level Bonus at 24th level.  Within a single level there is no difference whatsoever, and within a tier the difference is only a little larger (+4 vs. +1 or +5 vs. +2, rather than +2 vs. +0 or +2 vs. +1).  This may be safely ignored.

The difference between levels farther apart become greater — a Legendary character will have a Level Bonus +20 higher than a Basic character instead of +10 higher.  Given Echelon’s premise that higher-level characters are just generally better at everything, all else being equal than lower-level characters this is fine by me.  When they get this far apart in level they’re no longer in the same world anyway (even if they are on the same planet) and really shouldn’t be competing with each other, directly or indirectly (facing similar challenges).  The increased difference in Level Bonus is acceptable to me.

The eight levels needed to overcome one step in Difficulty Class is hinderingly awkward, especially since it doesn’t quite align with the tiers.  This will be more evident in the tables below.  If I shorten the Difficulty Class steps (+2 per step rather than +4) so they align better with the tiers, the impact of training and ability scores becomes too large.  I don’t mind someone gifted (Dex 9 rather than the average Dex 5) can do things one step more difficult, but two is a bit much (and if they’re trained as well and can get four steps up instead of two it’s worse yet).

Just through consideration I think there is reason to believe this is a good change to make.  Let’s look now at the actual numbers.

Existing Difficulty Classes

In the tables that follow, the character descriptions mean:

Code Description Meaning Expert
(8th level)
Legendary
(24th level)
D Deficient Untrained and ability score of 1; modifier = level/2 + 1 +5 +13
N Normal Untrained and ability score of 5, or trained and ability score of 1; modifier = level/2 + 5 +9 +17
T/G Trained or gifted Trained and ability score of 5, or untrained and ability score of 9; modifier = level/2 + 9 +13 +21
T+G Trained and gifted Trained and ability score of 9; modifier = level/2 + 13 +17 +25

This assumes that a normal mature human is fourth level and has an average ability score (Dex 5 for example).  A person that is trained in a particular skill gets a +4 bonus to checks with that skill, while a person who is gifted with the relevant ability score as a score of 9 (Dex 9 in this case); in both cases the person would have +9 to his checks before Level Bonus.  Someone who is ‘optimized’ is both trained and gifted, +13 to checks before Level Bonus.  Someone who is deficient has a score of only 1 (Dex 1 in this case), but may still be trained (+1 or +5 to checks, before Level Bonus).

Difficulty Target DC Normal 50% Percentage Chance of Success
Expert (8th Level)   Legendary (24th Level)
D (+5) N (+9) T/G (+13) T+G (+17) D (+13) N (+17) T/G (+21) T+G (+25)
Routine- 4 110
Routine 8 90 110
Routine+ 12 70 90 110 110
Easy- 16 50 70 90 110 90 110
Easy 20 Expert 30 50 70 90 70 90 110
Easy+ 24 Master 10 30 50 70 50 70 90 110
Hard- 28 Legendary -10 10 30 50 30 50 70 90
Hard 32 Epic+ -10 10 30 10 30 50 70
Hard+ 36 Epic+3 -10 10 -10 10 30 50
Very Hard- 40 Epic+5 -10 -10 10 30
Very Hard 44 Epic+7 -10 10
Very Hard+ 48 Epic+9 -10
Impossible- 52 Epic+11
Impossible 56 Epic+13
Impossible+ 60 Epic+15

Or, to put the frequencies into more human terms:

Target DC Descriptive Frequency of Success
Expert (8th Level)   Legendary (24th Level)
D (+5) N (+9) T/G (+13) T+G (+17) D (+13) N (+17) T/G (+21) T+G (+25)
4
8 always
12 routinely always
16 usually routinely always always
20 often usually routinely always routinely always
24 sometimes often usually routinely usually routinely always
28 rarely sometimes often usually often usually routinely always
32 never rarely sometimes often sometimes often usually routinely
36 never rarely sometimes rarely sometimes often usually
40 never rarely never rarely sometimes often
44 never never rarely sometimes
48 never rarely
52 never
56
60

I don’t like how large the overlap is between Basic and Legendary success rates, and I don’t really like the narrowness of the range of possible DCs.

Also, the two tiers needed to keep step with the target DCs is annoying.  Now, I don’t mind needing to be high level to do impossible things (after all, they’re impossible, it shouldn’t be possible for ‘regular’ people to do them at all!), this looks like a bit much.  Let’s extend the tier table a bit.

Tier Level D&D Level Normal Character   Optimized Character
50% DC 50% Difficulty 50% DC 50% Difficulty
Basic 1-4 0 18 26
Expert 5-8 1-4 20 Easy 28 Hard-
Heroic 9-12 5-8 22 30
Master 13-16 9-12 24 Easy+ 32 Hard
Champion 17-20 13-16 26 34
Legendary 21-24 17-20 28 Hard- 36 Hard+
Epic 25-28 21-24 30 38
Epic+ 29-32 25-28 32 Hard 40 Very Hard-
Epic++ 33-36 29-32 34 42
Epic+3 37-40 33-36 36 Hard+ 44 Very Hard
Epic+4 41-44 37-40 38 46
Epic+5 45-48 41-44 40 Very Hard- 48 Very Hard+
Epic+6 49-52 45-48 42 50
Epic+7 53-56 49-52 44 Very Hard 52 Impossible-
Epic+8 57-60 53+56 46 54
Epic+9 61-64 57-60 48 Very Hard+ 56 Impossible
Epic+10 65-68 61-64 50 58
Epic+11 69-72 65-68 52 Very Hard 60 Impossible+
Epic+12 73-76 69-72 54 62
Epic+13 77-80 73-76 56 Very Hard+ 64
Epic+14 81-84 77-80 58 66

Now, it’s possible that the defined labels are bogus and should be adjusted as well (which has issues I’ve mentioned above).  However, if the DCs are too close together they collide on the RNG.  I’m not prepared to do that.

Where D&D 3.x suggested that an optimized (and since it was in the Player’s Handbook on publication in 2000, before years of splatbooks and a system revision, this isn’t particularly optimized) 20th-level character could expect to hit DC 43 (23 ranks in a skill and +9 from the related ability score gets you to 50% success, and that is trivial optimization; once you get magic items and/or masterwork tools, assistants, and circumstance bonuses you can tack another +10 on that pretty easily).

Here, the same DC would require (from an optimized character, and optimization is somewhat limited in Echelon) and Epic+3 character (level 38-39 for the needed bonus).  In D&D 3.5 terms, this is a level 34-35 character.

Now, my view of ‘Very Hard’ might be a little skewed (I worked out the DC definitions for the new model before doing this part of the analysis), but even so the level needed seems awfully high compared to D&D 3.x.

Proposed Difficulty Classes

In the tables that follow, the character descriptions mean:

Code Description Meaning Expert
(8th level)
Legendary
(24th level)
D Deficient Untrained and ability score of 1; modifier = level + 1 +9 +25
N Normal Untrained and ability score of 5, or trained and ability score of 1; modifier = level + 5 +13 +29
T/G Trained or gifted Trained and ability score of 5, or untrained and ability score of 9; modifier = level + 9 +17 +33
T+G Trained and gifted Trained and ability score of 9; modifier = level + 13 +21 +37

This assumes that a normal mature human is fourth level and has an average ability score (Dex 5 for example).  A person that is trained in a particular skill gets a +4 bonus to checks with that skill, while a person who is gifted with the relevant ability score as a score of 9 (Dex 9 in this case); in both cases the person would have +9 to his checks before Level Bonus.  Someone who is ‘optimized’ is both trained and gifted, +13 to checks before Level Bonus.  Someone who is deficient has a score of only 1 (Dex 1 in this case), but may still be trained (+1 or +5 to checks, before Level Bonus).

Difficulty Target DC Normal 50% Percentage Chance of Success
Expert (8th Level)   Legendary (24th Level)
D (+9) N (+13) T/G (+17) T+G (+21) D (+25) N (+29) T/G (+33) T+G (+37)
Routine- 4
Routine 8 110
Routine+ 12 90 110
Easy- 16 70 90 110
Easy 20 Basic 50 70 90 110
Easy+ 24 Expert 30 50 70 90 110
Hard- 28 Heroic 10 30 50 70 90 110
Hard 32 Master -10 10 30 50 70 90 110
Hard+ 36 Champion -10 10 30 50 70 90 110
Very Hard- 40 Legendary -10 10 30 50 70 90
Very Hard 44 Epic -10 10 30 50 70
Very Hard+ 48 Epic+ -10 10 30 50
Impossible- 52  Epic++ -10 10 30
Impossible 56 Epic++3 -10 10
Impossible+ 60  Epic+4 -10

Or, to put the frequencies into more human terms:

Target DC Descriptive Frequency of Success
Expert (8th Level)   Legendary (24th Level)
D (+9) N (+13) T/G (+17) T+G (+21) D (+25) N (+29) T/G (+33) T+G (+37)
4
8 always
12 routinely always
16 usually routinely always
20 often usually routinely always
24 sometimes often usually routinely always
28 rarely sometimes often usually routinely always
32 never rarely sometimes often usually routinely always
36 never rarely sometimes often usually routinely always
40 never rarely sometimes often usually routinely
44 never rarely sometimes often usually
48 never rarely sometimes often
52 never rarely sometimes
56 never rarely
60 never

Ironically, I had to remove the textual labels (not-number indicator of difficulty) so this table of descriptive frequencies would fit on the page.

The +12 difference between the Deficient and the Trained and Gifted characters is still fairly solidly on the RNG.  Basic and Legendary characters are still close to being the RNG (there are overlaps, but not by much — the Trained and Gifted Basic character and the Normal Legendary character are +12 apart, in the Legendary character’s favor).

Extending the tier table, as shown above (but a bit shorter because we don’t need so many tiers) gives us

Tier Level D&D Level Normal Character   Optimized Character
50% DC 50% Difficulty 50% DC 50% Difficulty
Basic 1-4 0 20 Easy 28 Hard-
Expert 5-8 1-4 24 Easy+ 32 Hard
Heroic 9-12 5-8 28 Hard- 36 Hard+
Master 13-16 9-12 32 Hard 40 Very Hard-
Champion 17-20 13-16 36 Hard+ 44 Very Hard
Legendary 21-24 17-20 40 Very Hard- 48 Very Hard+
Epic 25-28 21-24 44 Very Hard 52 Impossible-
Epic+ 29-32 25-28 48 Very Hard+ 56 Impossible
Epic++ 33-36 29-32 52 Impossible- 60 Impossible+
Epic+3 37-40 33-36 56 Impossible
Epic+4 41-44 37-40 60 Impossible+

All things considered, this looks simpler, more regular, cleaner, and much easier to use.  Because the major determinants of ability (ability score, training, and tier) now all have 4-point spreads they align much better with 4-point differences in Difficulty Classes.  I can predict that if something will routinely done by an optimized Heroic character, trained or gifted Heroic characters and optimized Expert characters will usually succeed and normal Heroic characters and optimized Basic characters will often succeed.  I started examining this potential change to the rules today (okay, yesterday now) and it’s already fairly internalized.

Closing Comments

I think I’m about done with this topic.  Even superficial examination of the two sets of tables makes it clear ot me that th eproposed model, using straight level as the Level Bonus instead of using half level will be simpler and easier for me, without costing me anything I care about.  Again, this looks like a good change.

As long as hit points work out (and I’m confident they will, I have run some numbers but haven’t presented them here) and I can come up with a workable spell casting system, I think I will make this change.

Points to Ponder SelectShow

2 Comments

  1. Pingback: Reconsidering Level Bonus | Echelon d20 - An RPG framework based on the d20 system.

  2. Pingback: Links of the Week: October 17, 2011 | KJD-IMC - KJDavies "In My Campaign" Articles

Leave a Reply

Your email address will not be published. Required fields are marked *