Community

Two things I have never liked about the d20 system are how variable the outcome of events are (due to the d20) and little of a role skill plays in the outcome of events (again due to the d20).  

Now 5e has a chance to change this.  What if all trained skills used 2d20 take the highest instead of 1d20+3 in Next?  This would even apply to weapons you are proficient with (instead of the +3 proficiency bonus).  For untrained skills and abilities a player would still only roll a single d20.

Now the math of the game would need to account for this change. Getting a 11+ is 50% on 1 die, 75% on 2.  Getting a 13+ on 1 die is 40%, 64% on 2.  65% accuracy seems like a good starting place for a trained individual with a decent ability score.  Skill focus should still exist as a feat that grants a flat +3 bonus.  (Note: I was thinking DCs should be 10 = easy, 15 = average, 20 = hard, 25 = very hard, 30 = impossible)

This change has a few beneficial effects.  First advantage would grant an extra die rolled while disadvantage would subtract a die rolled.  Going from 2d20 to 3d20 is about equal to a +2 bonus.  This makes advantage far less powerful which in turn allows for it to be more common and less drastic.  

Secondly it keeps in line with 5e bounded accuracy far better than +3 for trained/proficiency.  For example the playtest minotaur has only a +4 bonus to hit meaning it will be very easy for a player to make himself almost unhittable.  If AC is reduced across the board to the point where 15 or so is the average AC (to compensate for the proficiency bonus being removed) then the minotaur whill not encounter unhittable targets (Note: the minotaur would not be considered to have enough training to roll 2d20 for its attacks).  

Lastly it will make skill very important to the outcome of an event.  If a player normally has a 50% chance to succeed at an untrained task a +3 only bumps that up 15% to 65% while rolling an extra d20 brings the chance to 75%.  This is even more noticable for very easy tasks.  If a character has a +3 and 1d20 he succeeds only 70% of the time at a DC10 task while he succeeds 80% of the time with a +0 and 2d20. This method will also bring a probability curve into the outcome of events to reduce the swingyness of a single d20.

Jul 26, 2012 -- 3:45PM, Gnarl wrote:

The only problem with multiple die rolls to hit is combat length. Imagine a Fireball against 5 critters. That's already 10d20 rolled instead of 5 and you can't roll all 10 at once because they go by pair.

It would be interesting to get feedback on this though. Maybe it's not as bad as I think it is.

Two things, 1) fireball doesn't currently have an attack roll, and 2) I think AoE effects should be a single attack roll anyway.  (In your example caster uses fireball gets a 5 and an 11, has a +5 bonus so total DC is 16.  The 5 targets then make their 5 saving throws)

As I posted earlier I have already been doing it in 4e. 

But I am not changing the DC tables and here is why

A easy DC8 is still 65% with 1d20 for untrained

A moderate DC12 is the 70% with 2d20 for trained, moderate is designed for max ability +4 or an edge if trained instead

A moderate DC13 is the 65% with 2d20 for trained, which is same odds as the +5 modifer from trained or max racial ability

A hard DC19 is the 64% for 2d20 +6 , which is nearly same odds as +11 from trained along with max racial ability and a +1 skill thing or max ability and a +2 racial skill.

Any one +5 skill thing is the same as (temporarily) trained, except for leveling skills is modifiers to keep up with the 4e DC table.

I am taking it further with 3d20 for equivalent to training focus feat (+3).

A hard DC19 is then 65% with 3d20 +3 ability and training focus and training, equivalent again to the +11 allowing training focus to compensate less ability, without making it the same as ability.

A moderate DC12 the 3d20 may only be a +2 equivalent rather than +3 so you could also say a +2 racial skill, or +2 power skill could also use 3d20.  While I have done this it makes the rule for those +2 things complicated on if you have training you can jump to 3d20, if you do not have training you stick with the +2 modifer, and if you have training focus it makes a +1 equivalent to jump to 4d20 so you are better off with the simple rule of +2 modifier never stacks.

Bell curves from adding up die reduce the variance around average, but maxing the stacked die instead tilts the odds making high numbers more likely and low numbers less likely.  It has nearly equivalent odds for needed rolls as the modifier, with the tradeoff that you have low chance of failure instead of never fail too easy tasks, and for tasks too hard from lacking ability the odds actually get worse if not impossible.  I think that is a more realistic model for training vs. ability.  So those who say it tweaks the odds too far, it actually does not do that at all when it is replacing a +5+3 modifier stack.


I agree 5e advantage being best/worst of 2d20 is overpowered relative to 4e since it is essentially a +/-5 modifer when it used to be a +2/-2 modifier.   But I cannot figure out how to make it be consistently +2/-2 using die stacking.   in 4e the weapons proficiency of +3 is only on the best weapons, is +2 less accurate weapons, with basic weapons and implements being a +0 profiency, while weapons/implement experts feats is only +1.   Since you do not want to give up modeling weapon accuracy, proficiency needs to stay as a weapon attack modifer and not a die stack.  

So lets presume for 5e you agree weapons/implement experts deserve to have the same bump as training for skills in 4e, that an expert should see a 25% boost in attacks not just a 5% boost.   The problem is disadvantage is only a -2 equivalent for the advantaged expert but for the unadvantaged expert or the novice it goes back to being a -5 equivalent  disadvantage.   So it then becomes min-maxing to take the expert feats with turn slow downs seeking out how to keep an advantage.    

The other problem is how do you do this as a DM, disregarding the don't want to fondle more die argument because DMs can and will 'workaround' that for mobs.   The issue is monsters prefer to simply think in terms of +attack based on how often DM wants them to hit, rather than applying the same rules as PC to come up with an NPC, and it is not fair if one side stacks the die and the other shifts the die.  

Take advantage/disadvantage back to the +/-2 modifers of 4e, use the best of die mechanic for training instead,   Keep 4e concept that training can replace ability for moderate tasks and both are required for hard tasks without blowing out the DC20 roll, and it improves upon 4e that training does not replace ability for impossible tasks.

For monsters that are not humanoid just call training instinct, and also stack the instinct die so it is fair to PC.