Advantage & 4e Skills

I’ve written a fair bit about skills in 4e, and how I personally think the math behind skills is a little shakier than, well, anything else in the edition. You can find some of my earlier blogs on the WotC Community site: here and here.

It might seem a little late to be thinking about house rules for 4th Edition, but I don’t believe so. Not everyone will be making the switch from 4e to 5e, and even those that do might decide to play the occasional short adventure or micro-campaign in 4e. No matter of the final quality of 5e, there will be some adventures and stories that might be better served by 4e. And if the system can be made a little more playable that’s only a bonus.

The Problem

Here’s a quick summary of the problem with skills in 4e. Sorry for anyone who’s read this six times already. I’ll keep in brief.

TL;DR: The numbers are too large too fast and there’s a disparity between trained and untrained that only increases.

You add your ability bonus (which for 4e was at least 16 but often 20) and a training bonus of +5. This means from level 1 the difference between someone trained or skilled and someone untrained and unskilled as much as +5 and a trained and skilled PC could be +10. While, in theory, this doesn’t increase (unlike 3e) as you only add 1/2 your level to skills there are no shortage of other bonuses. Feats regularly give skill bonuses, often as a side perk. Races and backgrounds give skill bonuses. Magic items give bonuses equal to their plus – again often as a side bonus. And there are the stat boosts that bump primary attributes, further adding to the disparity.

As low as level 10 the difference between trained & skilled and untrained & unskilled could be as high as +20, which means a DC the trained character cannot fail even on a “1” the other character cannot succeed even on a “20”.

The Rule

Start by dumping the bonus from skill training. It goes away. Gone.

Instead, when a character makes a roll with a trained skill they roll two dice and take the higher of the two results, just like Advantage like in 5e.

This is easily done in the Character Builder but not choosing any trained skills and marking printed copies with a check but not changing the numbers.

The Reasoning

Just like in 5e, lowering the numbers reduces the math and allows everyone to participate with skills. The expert doesn’t need to be challenged by increasingly improbably problems that the rest of the party cannot contribute with, and neither does the expert need to sit out in situations that should be good as just because the DCs have to be low enough to challenge everyone else.

This also means that while training increases your chances for success, it does not increase your maximum beyond physical limits. A trained yet weak long jumper might not be able to jump as far as a strong yet untrained jumper but a trained jumper will have more average and consistent results. There’s still the potential for a lucky amateur and clumsy expert, but it’s reduced as the expert has odds of success with a bell curve. I like this because training is not the equivalent of natural talent. No matter how hard you practice on the parallel bars, if you’re not naturally agile you won’t be as good as someone with raw natural talent having a good day.

As was pointed out by the Online DM the math behind rolling two d20s is roughly equivalent to a +5 in the middle of the bell curve (8-12), so two d20s and keeping the highest is equivalent to a +5 for average DCs. This is important as it means Advantage training is equivalent to the +5 training bonus. So in practice it’s roughly average to the +5 without having to inflate the DCs to accommodate.

Other Possible Changes

A trickier change that would make the math easier would be halving the bonuses from feats and other sources. Instead of the standard +2 bonus granted by races, backgrounds, and many feats this becomes a smaller +1. It’s still a bonus but the numbers do not get as high as fast.

This is trickier, as it cannot be easily done in the Character Builder. So it works if most of the table makes their characters by hand or is comfortable writing in skill bonuses.

If your table prefers to use the Character Builder and doesn’t like writing in bonuses it’s probably best to keep the standard bonuses but just ban feats and magic items that only grant bonuses to a skills, and limit those that grant skill bonuses as a secondary option. Backgrounds might be limited to only providing skill training and not bonuses.

Skill Cap

One other option I quite like but might be a little complicated is the idea of a “skill cap”. Specifically, there is a limit to bonuses you can gain on a skill. This option would not ban feats, items, backgrounds and related options that grant skill bonuses, but would instead cap the bonuses and mostly limit those options to character that are not already optimized for that skill.

This is an idea pulled from World of Warcraft but that shows up in other games, typically ones with heavy gear optimization. In those games you can add and stack bonuses to a certain ability again and again up until a hard limit where additional bonuses either confer no additional benefit or you experience diminishing returns. Such as fire resistance. You can only have so much fire resistance before you either stop taking fire damage or only take minimum fire damage.

As the baseline example, we’ll say the skill cap on skills will be the maximum ability bonus of that level. So the person with magic items, racial bonuses, and the like can make-up for a sub-par ability score to become the equivalent of someone with natural talent but not surpass them, and someone with a high ability score would not be able to get that much better.

So the cap would be 5 + 1 at levels 8, 14, 20, and 28. Simpler but less accurate options might be 5 + 2/tier.

This gives players a little more with items and feats and allows players to build characters closer to what they would want, and allows character to still take options that might allow them to catch-up, which would otherwise be banned.

The problem with this method is it stretches verisimilitude a little. Magic items that make someone stronger of faster (and thus give a skill bonus) won’t function on someone already strong. Which is odd and a little silly. If this is an issue, a variant would a higher cap that would also allow a character with a high ability score to slip on a bonus from somewhere, but not more than one. But, for the reasons stated in an earlier section, it probably shouldn’t be too much higher.

The New DCs

Let’s start to look at what a table of DCs might look like.

Here’s a quick table of the skill numbers assuming an 18 in the requisite ability score. This is likely the average results.

Level

Talented

Talented + 1/2 level

Untalented

Untalented + 1/2 level

Disparity

1

4

4

0

0

4

4

4

6

0

2

4

8

5

9

0

4

5

10

5

10

0

5

5

14

6

13

0

7

6

18

6

15

0

9

6

20

7

17

1

11

6

24

7

19

1

13

6

28

8

22

1

15

7

30

9

23

1

16

7

Higher or lower ability scores change the disparity (a 20 in the talented and 8 in the untalented can skew the results and would be the biggest disparity). Likewise, race, feats, and the like can also increase the disparity.

For the chart I’ll start with the basic round numbers: 5, 10, 15, 20. The chances of success for someone with a +4 bonus in the skill would thus be 100%, 75%, 50%, and 25%, and for someone with no bonus would be 80%, 55%, 30%, and 5%. As the disparity still increases a little, I’ll have to increase the DC of Hard and Very Hard check, making those harder for non-trained. After about level 8, someone with a penalty and no other bonuses will no longer be able to make a Very Hard check, but that’s a catch of the system. But they can still succeed at Hard checks until the very end, and talented characters still have to make checks to succeed Moderate DCs at the end.

Level

Easy

Moderate

Hard

Very Hard

1

5

10

15

20

2-3

6

11

16

21

4-5

7

12

17

22

6-7

8

13

18

23

8-9

9

14

20

25

10-11

10

15

21

26

12-13

11

16

22

27

14-15

12

17

24

29

16-17

13

18

25

30

18-19

14

19

26

31

20

15

20

27

32

21

16

21

28

33

22-23

17

23

29

34

24-25

18

24

30

35

26-27

19

25

31

36

28-29

20

28

33

38

30

21

27

34

39

For precision I start with level 1 and have a new column every even level, when skills increase. The only anomaly is level 21, when characters hit the epic tier and everyone gets their second mass stat boost, which does shift the numbers a little.