Is "Unearthed Arcana: Players Make All Rolls" Correct?

Question

In this Unearthed Arcana, they give rules so that players can make all the rolls, rather than the DM sometimes rolling for enemies. In one section, there are specifically rules for converting saving throw bonuses into an equivalent DC, which the players roll against, rather than the DM. However, after doing some math, I think they've made a mistake.

They say that you convert a Saving Throw into a Saving Throw Check by adding 11 to the defender's saving throw modifiers, and using that as the DC for the check. The player then rolls against this DC, adding their spellcasing ability modifier and proficiency bonus. However, according to the following math, this conversion does not produce the same results.

In this math, I compare the chances of a player succeeding on a saving throw check, using the rules in the UA article, and of a monster failing its saving throw using the standard PHB rules. If the math is correct, both of these results should have the same chance.

These are the formulas I used.

Chance to succeed Saving Throw Check = (20 - DC + 1 + proficiency + casting mod) / 20

Chance to FAIL a Saving Throw = (DC - 1 - save proficiency - save mod) / 20

Assuming a proficiency bonus of +2, +0 for all ability mods, and no save proficiency:

Monster's Saving throw Check DC = 11 + save mod = 11

Chance to succeed a saving throw check DC11: (20 - 11 + 1 + proficiency + casting mod) / 20 = 12/20 = 60%

Player's Saving throw DC = 8 + proficiency + spellcasting ability mod = 10

Chance for monster to fail saving throw DC10: (10 - 1 - save proficiency - save mod) / 20 = 9/20 = 45%

If this were a valid conversion, both formulas would result in the same chance of success for the player, and failure for the monster respectively, however, they are off by 15%. This suggests that the unearthed Arcana's rules are not a valid conversion.

Are the Unearthed Arcana rules wrong, or is my math wrong?

Answer 1

UA is wrong: a simple example shows it so.

Wizwiz, a level 5 wizard with INT 18 casts Charm Person on Barbar, a barbarian with WIS 8.

PHB Rules

Wizwiz's spell save DC is (8 + proficiency bonus + INT mod) = 15. Barbar's WIS save modifier is-1. If Barbar rolls a 16-20 he resists the charm. Barbar resists 25% of the time.

UA Rules

Wizwiz's save check modifier is (DC - 8) = +7. Barbar's resistance DC is (11 + WIS save mod) = 10. Wizwiz's spell takes full effect on a roll of 3-20; Barbar resists on a roll of 1 or 2. Barbar resists 10% of the time.

Therefore there exists a case where the results are not identical, and UA rules do not faithfully reproduce PHB rules.

Answer 2

UA is wrong

It doesn't reproduce faithfully the results of spells adjudicated by PHB rules.

I get the same error-by-three as you. (So does reddit, for what it's worth.)

Conventions:

"castmod": caster's proficiency bonus+casting ability modifier. (Yes, 8+castmod is just the spell save DC, but we need to break it out for algebra)

"savemod": target's proficiency bonus if proficient+targeted ability modifier

\$\text{d}20^+\$: result of the roll of a d20 on which a 20 would be the most beneficial result to the caster

\$\text{d}20^-\$: result of the roll of a d20 on which a 1 would be the most beneficial result to the caster

under normal rules a caster has full effect when $$\text{d}20^- + \text{savemod} < 8 + \text{castmod}$$ or, flipping sides, $$8 + \text{castmod} > \text{d}20^- + \text{savemod}$$

under UA rules a caster has full effect when $$\text{d}20^+ + \text{castmod} >= 11 + \text{savemod}$$ (see the '>=' vs '>'? It has to do with who wins ties.)
Note that being \$>= n\$ is the same as being \$> n-1\$, in the world of integers. UA rule becomes $$\text{d}20^+ + \text{castmod} > 10 + \text{savemod}$$ Here's the fun part: the result of a \$\text{d}20^+\$ roll is the same as 21 minus the result of a \$\text{d}20^-\$ roll. Algebraically, I'm saying \$\text{d}20^+ = 21 - \text{d}20^-\$. Substituting, $$(21 - \text{d}20^-) + \text{castmod} > 10 + \text{savemod}$$ add that \$\text{d}20^-\$ roll to both sides

$$21 + \text{castmod} > 10 + \text{d}20^- + \text{savemod}$$ and subtract 10 from both sides $$11 + \text{castmod} > \text{d}20^- + \text{savemod}$$

So I'm saying that the caster has full effect when:

\$8 + \text{castmod} > \text{d}20^- + \text{savemod}\$ under the normal rules, but when
\$11 + \text{castmod} > \text{d}20^- + \text{savemod}\$ under the UA variant.

Which is a difference of 3-out-of-20, the same as your 15%.

Answer 3

We've been using this mechanic for three years and are still using them. But here are the numbers we use:

Player character's "defense roll" → armor class minus 12.
Monster's "attack value" → 10 + to hit bonus.

Motivation: Players only have to do that subtraction when their AC change, which is rare. The DM has to calculate attack value all the time, so having an easy calculation (+5 → 15) is great.

For spells, here is what we use:

Player character spell attack → prof + mod.
Monster's "dex defense" (or "wis defense") or w/e → 14 + save.

Motivation: It's easier for players to only have to learn one number for spell attacks. It's worth the pain for DM's to add 14 instead of 10.

We did this from the start because we were aware of this type of mistake from the 3e version of this UA.

A side curiosity

Since the required number to make the math work out, 14, seems a bit high compared to AC, I at once point wondered if there was a similar mistake for spell saves when they constructed 5e; that, just as with this UA, the person holding the dice wins the ties (you have to roll that number or higher), a fact which has demonstrably lead to "off-by-2" errors both in this 5e UA and in the 3e UA, maybe they did an off-by-two omission here too, and it should be 10+prof+mod, or, worse, they did figure off-by-2, but in the wrong direction, and it should be 12+prof+mod. I decided to investigate by going through all the monsters in the SRD. If 8+prof+mod were to be correct, the average difference between the monster's saving throw and their AC should be around 4. (In other words, monsters would've been assumed to wear, on average, at least scale mail.)

It's not quite there. That number, for all the six saves averaged together, is 2.57 (which would make 9+prof+mod or 10+prof+mod correct). Sure, when only looking at int saves that number is actually higher, it's 5 (which would make 7+prof+mod correct). But for str it's 1.5, dex it's 2.13, con it's 0.75, wis it's 2.2 and for cha it's 3.6. Cha is also pretty close, but since many attacks use the physical saves, that means that "save" type spells have around a 10% lower chance of hitting than "spell attack" type spells. So maybe that number should've been 10+prof+mod overall, which would also have been way easier than a sudden&arbitrary 8, and with the this "Players Make All Rolls" UA, it should've been d20+prof+mod vs 12+save.

In our home campaign we still use 14+save though (i.e. the same probs as 8+prof+mod). These extra 10% are more of a curiosity. (Oh, and you can't crit on saving throw spells either, right?)