D&D 5E Advice on how I should multi-class

[ClaytonCross]

Yeah right. That's not DPR, unless you somehow always hit.

It's not even a proxy for DPR, because SS reduces to-hit.
Rogue 15 SA will do 8d6 if any attack hits (~ 64% of the time).
Fighter 20 will have 2 extra attacks for 1d8+15 or so each (~ 40% of the time)
Or something like that, but you get the picture.


So your analysis is comparing apples to oranges, making it invalid.

You can not measure hit % on an unknown variable (AC). I made a calculation on DPR as 100% hit but then mentioned that you need to consider your hit % based on what type of enemies your GM sets against you. If your fighting large numbers of low AC NPCs it going to be closer to 100% and if you have single opponent with high AC it will be dramatically lower. The Statistics I am providing with the awareness that it changes per your enemy is more accurate a statement than giving a miss ratio on a made up or moving AC which is not relevant to individual fights. So you lay out the average DPR then give either a low and high AC comparison for the player to determine what is most relevant to them. Your 64% - 40% is not relevant to the players roles or GMs play style so it is twisting DPR and making it harder to see the picture clearly. However if you want to break down hit and miss percentages vs staged high and low AC with and with/without advantage/disadvantage so that they have a clearer picture, go right ahead. Other than that your not explaining your comparison to let the player know when and if your numbers are relevant to them. Not saying your math is wrong. Just that you need to state the situation like I did in my original post or you invalidate your data by using an undefined variable in a static representation such as a DPR.

 

[bid]

The Statistics I am providing with the awareness that it changes per your enemy is more accurate a statement than giving a miss ratio on a made up or moving AC which is not relevant to individual fights.

Were you aware of the impact, you would realize the 30-60% error on your method make the whole analysis pointless.
- 8d6 looks like it's doing 28 damage
- 2d8+30 looks like it's doing 39 damage
** Wow! 39 / 28 looks like 40% more damage!!
- 8d6 * .64 ~ 17.92
- 2d8+30 * .4 ~ 15.6
** Wow! It's actually 13% less damage!!

You can try against AC13: a Dex20 level 9 character with archery and +1 from AA needs 6+ to hit with SS. Very few creature will have an AC low enough to make the auto-hit error go away.


Using 100% hit is "twisting DPR" worse than a stopped clock. You rarely get 2 encounters per day where all your attacks hit.

 

[ClaytonCross]

Were you aware of the impact, you would realize the 30-60% error on your method make the whole analysis pointless.
- 8d6 looks like it's doing 28 damage
- 2d8+30 looks like it's doing 39 damage
** Wow! 39 / 28 looks like 40% more damage!!
- 8d6 * .64 ~ 17.92
- 2d8+30 * .4 ~ 15.6
** Wow! It's actually 13% less damage!!

You can try against AC13: a Dex20 level 9 character with archery and +1 from AA needs 6+ to hit with SS. Very few creature will have an AC low enough to make the auto-hit error go away.


Using 100% hit is "twisting DPR" worse than a stopped clock. You rarely get 2 encounters per day where all your attacks hit.

100 * .64 = 64
200 * .4 = 80

Yep numbers. Just as meaningless as yours without an explanation. Mine is less twisted because I explained that your hit % is going to change depending on the DM the enemy makes you fight. Again I am not saying your math is wrong only that you need to explain why your multiplying by .64 and .4 to your rolls or the OP (and myself) doesn't have a reasonable expectation of the context in which your "DPR" is relevant. Where I on the other stated up front that my 100% hit DPR is not reliable to determine damage alone. You seem to be trying to demonstrate additional calculations to account for accuracy but your not stating the limitations of calculation your simple stating it as fact. This extra step is bound to be right (if it is correct, which I am going to assume it is) but only with in the limit of the definition your using to make them. So my point is, my DPR is a better guideline as is, and certainly better than none, but yours could be better math that does not show its relevance because you did not list your reasoning and limits. Do that and you very well may have a better answer. Please do! I am interested to see it.

Kind of like the meaning of life is 42. Sure maybe it is, but until you tell me how you got that or under what context that is useful it is meaningless. Not saying it is wrong, just that it need elaboration.

 

[FrogReaver]

100 * .64 = 64
200 * .4 = 80

Yep numbers. Just as meaningless as yours without an explanation. Mine is less twisted because I explained that your hit % is going to change depending on the DM the enemy makes you fight. Again I am not saying your math is wrong only that you need to explain why your multiplying by .64 and .4 to your rolls or the OP (and myself) doesn't have a reasonable expectation of the context in which your "DPR" is relevant. Where I on the other stated up front that my 100% hit DPR is not reliable to determine damage alone. You seem to be trying to demonstrate additional calculations to account for accuracy but your not stating the limitations of calculation your simple stating it as fact. This extra step is bound to be right (if it is correct, which I am going to assume it is) but only with in the limit of the definition your using to make them. So my point is, my DPR is a better guideline as is, and certainly better than none, but yours could be better math that does not show its relevance because you did not list your reasoning and limits. Do that and you very well may have a better answer. Please do! I am interested to see it.

Kind of like the meaning of life is 42. Sure maybe it is, but until you tell me how you got that or under what context that is useful it is meaningless. Not saying it is wrong, just that it need elaboration.

tldr.
Stop using 100% chance to hit. At least pick a target AC that will yield somewhere between 50% and 85% chance to hit or something like that. Do this anytime you look at sneak attack or sharpshooters -5/+10.

.64 is the chance a sneak attack lands when using 65% base chance to hit after -5/+10 sharpshooter and 2 attacks.
.4 is the chance to hit after the -5/+10 sharpshooter assuming 65% chance to hit.

As his analysis shows. Using 65% base chance to hit means the high sneak attack rogue version does more damage.

 

[FrogReaver]

Were you aware of the impact, you would realize the 30-60% error on your method make the whole analysis pointless.
- 8d6 looks like it's doing 28 damage
- 2d8+30 looks like it's doing 39 damage
** Wow! 39 / 28 looks like 40% more damage!!
- 8d6 * .64 ~ 17.92
- 2d8+30 * .4 ~ 15.6
** Wow! It's actually 13% less damage!!

You can try against AC13: a Dex20 level 9 character with archery and +1 from AA needs 6+ to hit with SS. Very few creature will have an AC low enough to make the auto-hit error go away.


Using 100% hit is "twisting DPR" worse than a stopped clock. You rarely get 2 encounters per day where all your attacks hit.

Not sure why you are calculating the percentages that way. There is a base component you are leaving out of the percentage you are calculating which doesn't make it very useful IMO.

 

[ClaytonCross]

tldr.
Stop using 100% chance to hit. At least pick a target AC that will yield somewhere between 50% and 85% chance to hit or something like that. Do this anytime you look at sneak attack or sharpshooters -5/+10.

.64 is the chance a sneak attack lands when using 65% base chance to hit after -5/+10 sharpshooter and 2 attacks.
.4 is the chance to hit after the -5/+10 sharpshooter assuming 65% chance to hit.

As his analysis shows. Using 65% base chance to hit means the high sneak attack rogue version does more damage.

Against what AC? your doing what your accusing me of! lol. I don't need a base AC because I am giving Damage per hit and then stating that hits will be variable on AC (Which I was wrong for saying DPR because that is inaccurate). And Your saying:

Fighter 5 / rogue 15, fighter 11 / rogue 9 and fighter 20 do about the same DPR over progression. IOW, at level 15-19 the rogue MC will do more damage than the pure fighter, but at level 20 fighter does better.

Curving shot is the main reason stopping at fighter 5 is a bad idea. Beyond that, I don't expect balance to be different as your arcane shots are more limited than BM's SD.

If you have SS (which you should), you are removing some attacks at 1d8+15 for 5d6 to 8d6 SA damage. For instance at level 15, you are exchanging your 3rd attack (roughly 10 DPR ~ 1d8+15 * .5) for SA damage (13 DPR ~ 5d6 * .75). Overall not a big gain/loss in damage.

If you MC, I wouldn't do it before fighter 5. I would MC to rogue 3, then fighter 7, then rogue 9 to finish at fighter 11 / rogue 9.


It all comes down to which build matches your RP concept and has the best utilities.

My point is you can't have one DPR for not set AC and since you didn't list one your answer is incomplete where mine is not though less mathematically accurate it provides the means for some assessment using total DPH.

However ....
20 sided dice = 5% chance to roll any number
+ hit modifier (Dex 20 so 5 + proficiency levels 5 to 8 are +3 - 5 for Sharp shooter +1 Magic arrow, +2 Archery Fighting style) = So +6 for levels 5-8 or +7 for level 9-11
- (AC -1) to remove the missing roles, the -1 is because if you meet the AC it is a hit.
Then multiply that number by the 5% value of each number you can roll on a 20 sided dye to determine the chance to hit as decimal or multiply by 100 for the percent.

level 5-8 vs AC10 = .85 or 85%
level 5-8 vs AC15 = .6 or 60%
level 5-8 vs AC20 = .35 or 35%

level 9-11 vs AC10 = .9 or 90%
level 9-11 vs AC15 = .65 or 65%
level 9-11 vs AC20 = .4 or 40%

Now Sneak attack is different because you have to figure out the chance it happening if at least once with 2 attempts. My understanding it of that formula is:
1st attack = % Example: .85 or 85%
2nd attack = Inverse of 1st % X % + Last total % Example: (.25 X .85) + .85 = .9775 or 97%
3rd Attack = Inverse of 2st % X % + Last total % Example: (.0225 x .85) + .9775 = .996625 or 99%

Corrected for the revised arcane archer with percent to hit for proper DPR.

level 6: Possible Ability Score Improvement or Feat
VS AC10
Basic Attack (1d8+15) = 24 X .85 = 20.4 per basic attack
Sneak attack for 2 attacks 1d6 = 3.5 X .9775 = 3.42125
Rogue 1 / Arcane Archer 5: (20.4) + (20.4) + (~3 SA) = ~44 DPR
Arcane Archer 6: (20.4) + (20.4) = ~41 DPR + Crossbow expert?

VS AC15 <-- 14.25 is the average for all the monsters in the book, while that does not reflect what your DM will make you fight its not a bad place to start.
Basic Attack (1d8+15) = 24 X .6 = 14.4 per basic attack
Sneak attack for 2 attacks 1d6 = 3.5 X .84 = 2.94
Rogue 1 / Arcane Archer 5: (14.4) + (14.4) + (~3 SA) = ~32 DPR
Arcane Archer 6: (14.4) + (14.4) = ~29 DPR + Crossbow expert?

VS AC20
Basic Attack (1d8+15) = 24 X .35 = 8.4 per basic attack
Sneak attack for 2 attacks 1d6 = 3.5 X .4375 = 1.53125
Rogue 1 / Arcane Archer 5: (8.4) + (8.4) + (~2) = ~19 DPR
Arcane Archer 6: (8.4) + (8.4) = ~17 DPR + Crossbow expert?

--- This is were it got late, I got tired and went to bed, lol. If no one sees has a correction for me I will try to remember and finish tomarrow and add the Revised ranger since the OP added this:

A reminder to a few of you that I am playing the revised arcane archer. Also, I am now heavily considering taking a 2-3 dip into the revised ranger. So maybe go fighter 7, then go ranger 3 then back to fighter.

level 8: Possible Ability Score Improvement or Feat
Rogue 3 / Arcane Archer 5: (1d8+15) + (1d8+15) + (2d6 SA) = 48.5 DPR
Rogue 2 / Arcane Archer 6: (1d8+15) + (1d8+15) + (1d6 SA) = 45 DPR + Crossbow expert?
Arcane Archer 8: (1d8+15) + (1d8+15) + (2d6 hex) = 48.5 DPR + Crossbow expert? + Magic Initiate for Warlock Hex?

Level 11: Possible Extra Attack
Rogue 6 / Arcane Archer 5: (1d8+15) + (1d8+15) + (3d6 SA) = 52 DPR + Uncanny Dodge + Expertise
Rogue 5 / Arcane Archer 6: (1d8+15) + (1d8+15) + (3d6 SA) = 52 DPR + Crossbow expert? + Uncanny Dodge
Rogue 3 / Arcane Archer 8: (1d8+15) + (1d8+15) + (2d6 SA) = 48.5 DPR + Crossbow expert? + Magic Initiate?
Arcane Archer 11: (1d8+15) + (1d8+15) + (1d8+15) + (3d6 hex) = 71.5 DPR + Crossbow expert? + Magic Initiate? + Indomitable

 

[bid]

Not sure why you are calculating the percentages that way. There is a base component you are leaving out of the percentage you are calculating which doesn't make it very useful IMO.

Do you mean it should be 4 attacks vs 2 attacks + SA?


The important point is that any time you compare DPR of 2 attacks that have different chances to hit (GWM, SS, SA with extra attack), ignoring to-hit adds an error that easily reaches 50%.


Anyway, the point has been raised. Buyers beware!