Originally Posted by
Dmnqwk
Hypothesis: Crits matter and this "nerf" is terrible...
Experiment:
We'll begin with a good old fireball, for nostalgia's sake. We're using it at level 10 vs some annoying bandits with 100hp each, 5 of the rotters.
Now, classic fireballs deal 10d6 damage at 10th level... but at some point they modified them to deal 10d3+30 instead
So, using an example:
Assume 20% crit (slightly high but it's nice and round!) for triple damage.
Assume 10d6 vs 10d3+30 with a 100% multiplier (again, round numbers)
10d6 averages 35 damage, 70 with multiplier.
Min 20, Max 120, average 70.... so assume 5 hits of 20, 45, 70, 95, 120
The mobs have 100 hit points (Fighters with no con 10hp/level) and never make a saving throw!
If a crit is rolled... 4 mobs die, 1 mob live
If a non-crit is rolled... 1 mob dies, 4 mobs live.
3 mobs can be killed by critting... so if we use a 20% chance in this scenario there's a 60% chance the crit is not wasted.
So in 5 runs of the scenario (seeing the crit hit a different person each time) we'll find 2 crits wasted, 3 crits kill people.
This means we'll see 5 people killed regardless of crits and we'll see 3 people killed because of crits so a total of 8/25 die to the fire!
Now... if we look crits being rolled per spell not per infliction we'll find 5 people are killed regardless.... and then on the crit fireball the 20 damage person still lives... so 3 additional people are poofed (the 45,70,95 as the 120 dies anyway) so a total of 8/25 die to the fire!
But... if we change the parameters for the damage buff to weighted dice...
10d3+3 now averages 50, doubled by potency and enhancements as before to 100 damage min 80 max 120 still.
The 5 damaging numbers are now 80, 90, 100, 110, 120.
On a non-crit 3 mobs die, 2 mobs live
On a crit EVERYBODY dies.
With individual crits... 40% chance it's not wasted (as 60% die regardless)
5 runs of the scenario sees 15 people killed automatically and then 2 extra's killed so 17/25 die because of the dice buffs.
In a crit rolled for spell not individual the numbers dont alter (because we're running this 5 times you may have noticed the numbers dont change hahaha!) so 17/25 killed.
Regarding spellpoint conservation:
Scenario a) 20 45 70 95 120 are the options for the damage... with 2 fireballs of the 25 combinations... 19 will result in smooshy deaths (20+20, 20+45, 20+70, 45+20, 45+45, 70+20 are the 6 that wont)
Scenario b) 80 90 100 110 120... 2 fireballs, everything dies!
Of the 25 combinations in scenario a... assuming 20% crit still (giving us 625 combinations due to crit possibilities on each fireball being 1 in 5, as I said simple maths) we find that only the 20+20 option requires a double crit to succeed. So 19*25/625 dont require crits 1/625 requires a double crit and the remainder need 1 of the 2 fireballs to crit.
So... 5 combinations become 125 combinations when we look at whether a or b crit... and if neither crits (chance neither is a crit is 80%*80% so 64%) we see 36% of the 125 combinations are boomsmooshes due to the crit... simplified 45/125. so 19*25(475) + 45 = 520/625... downgrading to the original 25 we find 20.8 deaths occur... so crits account for 1.8 deaths of 25 people... (7.2% of deaths caused by a 20% crit chance)
Kinda sucks people get worked up about crits when truthfully... they do less than you think.
Summary:
Crits play less of a role than you are aware, random dice see to that.
Weighted dice buff was a far more useful addition and plays a more significant role than crits.
Thanks for reading!
P.S. This is simplified maths and doesn't take into account saving throws, concentration checks forced by not auto critting etc... but it's a solid foundation to show why crits matter less.