D&D 5E Great Weapon Fighting

[occam]

What would it do to the math if you required that both sixers on the greatsword be either one or two for the reroll?

On 4/36 of the damage rolls, you'd replace the combined roll with another that averaged 7 points. On 16/36 of the rolls (1 or 2 on one die, 3-6 on the other), you'd have an average of 6 points (1.5+4.5). And on the remaining 16/36 of the rolls (3-6 on both dice), you'd have an average of 9 points (4.5+4.5). So:

(4*7 + 16*6 + 16*9)/36 = 7.44 points of damage on average

A boost of 0.44 points of damage. That makes for a pretty sucky ability.

 

[GrumpyGamer]

Damage on 2d6 will clump between 7-10, while distribution will be flat between 3-12 on a d12. Unless you really like rolling max damage, 2d6 is better in all ways with Great Weapon Fighting.

Modified output showing results for all 1,000,0000 random rolls of 2d6 and 1d12 (1,2,3,4,5,6,7,8,9,10,11,12):
>>> dicesim_py(1000000,2,6)
0
3111
6305
27873
49222
98358
148390
173044
197579
148119
98616
49383
(8.332742, 0.049383)
>>> dicesim_py(1000000,1,12)
13689
13891
96956
97404
96885
97548
97426
97222
97240
97148
97262
97329
(7.335896, 0.097329)

Edit: changed number of rolls from 100,000 to 1,000,000 to smooth out random distribution.

 

[GrumpyGamer]

You don't need to simulate; it's easy enough to calculate deterministic averages.

If only I had known how useful math was for gaming, I would have paid more attention in school.

Thanks for proofing my output!

 

[JRRNeiklot]

Why not just have two handed weapons do an extra die of damage and save the reroll? I.E, a greatsword could do 3d6 instead of 2d6.

 

[Iosue]

Why not just have two handed weapons do an extra die of damage and save the reroll? I.E, a greatsword could do 3d6 instead of 2d6.

Because that increases the range, which can be problematic in terms of balance, while what the design wanted to do was decrease the variability, which is generally less so.

 

[TwoSix]

Why not just have two handed weapons do an extra die of damage and save the reroll? I.E, a greatsword could do 3d6 instead of 2d6.

Solves the greataxe problem!

 

[JRRNeiklot]

Because that increases the range, which can be problematic in terms of balance, while what the design wanted to do was decrease the variability, which is generally less so.

It would only be 3-18 vs 6-12 as it stands now. And on a greater bell curve. And I actually like greater variability.

Solves the greataxe problem!

Yeah, 2-24 might be a bit much.

 

[variant]

I am just going to nerf the greatsword and maul to a d12.

 

[Kobold Stew]

Damage on 2d6 will clump between 7-10, while distribution will be flat between 3-12 on a d12.

Not quite -- you can still get a 1 or 2 on the great axe, it's just much less likely.

It really doesn't surprise me that some players are going to want the larger average range rather than the possibility of .5 points of average damage. Not every fighter is going to be a GWF (where the difference increases to 1 point difference on average [a bit more with crits, but not much more; not calculable since it depends on the AC of opponents], and that .5 difference will matter less than the tripled likelihood of getting max damage. (minimum damage is less important, since there will also be a bonus from strength -- even rolling a 1 still will mean a first-level fighter will typically do 4 points damage. I understand some players wanting that choice.

(my math agrees with Occam in post 5).

 

[redwizard99]

According to the text, if the melee damage die roll is 1 or 2, you can reroll the damage die.
For me, it means that unless the total melee damage die result is 1 or 2, you cannot do the reroll.
Any opinion?