Bonus: some extra cantrip analysis#
These are some things that are interesting, but not necessarily the focus of this article. Please read the previous article first if you haven’t already.
Emoji key#
We’ll represent the classes with emojis to keep the chart a bit more readable.
🤺 Swashbuckler
🩸 Sorcerer
#r "nuget: Plotly.NET.Interactive"
#r "nuget: FSharp.Json"
#load "../Helpers.fs"
#load "../Bestiary.fs"
#load "../Compare.fs"
#load "../DamageDistribution.fs"
#load "../Library.fs"
#load "../Transform.fs"
#load "../ChartTools.fs"
#load "../publish-backend/average-level.fs"
open Plotly.NET
open PublishBackend.AverageLevel
open PathfinderAnalysis.ChartTools
- FSharp.Json, 0.4.1
- Plotly.NET.Interactive, 5.0.0
Shortbow vs. arbalest#
Sometimes, it’ll be worth it to spend 3 actions to Strike, Reload, and Strike again. But implied in that are the many times you’ll only have the ability to Strike once.
We’ll let ABPs be the standard backup case. We’ll let the property runes be max level for curiosity’s sake. And then we’ll include shortbow at PL-4 because its average damage over the course of 20 levels is so comically close to the ABP shortbow, despite their extremely different scaling. (Please click on their names in the legends to hide them and look at what you want to look at.)
So, how do the d6 reload 0 shortbow and the d10 reload 1 arbalest compare against each other?
cantripArbalestVsShortbowCharts 0
cantripArbalestVsShortbowCharts 2
cantripArbalestVsShortbowCharts -2
Here we see that it’s pretty common for 2 shortbow Strikes in a round to be roughly halfway between 1 and 2 arbalest Strikes. That’s true whether they’re backup weapons or mainline weapons (excluding things like sneak attack, for example, where it’s clear that the ease of reload 0’s Strikes is very valuable).
So I guess a takeaway here is a rather intuitive result: as the weapon gains more property runes, reload 0 starts becoming more valuable. On the other hand, for classes that specialize in reloading, those classes will be able to keep pushing for 2 Strikes more often. Those two factors balance each other out, though the Reload compression actions are not always .
However, those classes are also not using arbalests as backup weapons! They’re using crossbows or guns as their main weapons. So your backup weapon is better off, on average, as a shortbow unless you need it for overcoming higher physical damage resistance.
Spout’s saves#
Middle, highest, and lowest#
Most casters are probably not going to continue to use Spout if they realize the enemy’s highest save is Reflex. Let’s take a look at a hypothetical where Spout always targets the lowest save, always targets the middle save, or always targets the highest save and compare it to its standard usage against the reflex save.
cantripSpoutSavesCharts [SaveSelector.Reflex; SaveSelector.Middle; SaveSelector.Lowest; SaveSelector.Highest] 0
You might’ve wondered why I used Spout vs. Reflex instead of Spout vs. Middle on the base cantrip-level analysis. Well, this answers that! They’re too similar to matter.
This shows some rather intuitive results:
Targeting the high save is an almost 20% drop on average relative to targeting the middle save.
Targeting the low save is an almost 20% bump on average relative to targeting the middle save.
Always targeting the reflex save is extremely similar on average to always targeting the middle save.
This suggests a relatively healthy spread, where reflex saves are balanced in how often they’re the lowest, moderate, and highest save. It’s technically possible that it’s almost always the middle save, or almost never the middle save and equally as often the highest or lowest, or some bizarre ratios that ultimately achieve the same results.
Reflex, will, and fortitude#
Let’s check on that by pretending that Spout can target other saves.
cantripSpoutSavesCharts [SaveSelector.Reflex; SaveSelector.Will; SaveSelector.Fortitude;] 0
Reflex and Will are very nearly identical in the long run, though there is the consideration that will save targeting often has the mental trait, which can be caught up in mental immunity even if will is the lowest save.
Always targeting Fortitude, however, is a little over 5% worse than always targeting Reflex. Very strange!
Perhaps fortitude saves run a little high, even when they’re the moderate save. Or perhaps they’re a little bit extra higher when they’re the highest save relative to when reflex or will is the highest save. We’ll have to dig into that another time.
Level gap of 2 vs. lowest and highest saves#
Let’s combine some of our insights from the level-based analysis earlier with our insights about how targeting different defenses impacts your expected damage.
For this, we’ll compare a Spout that always targets the lowest save against an on-level creature to a Spout that always targets the middle save against a creature 2 levels below the party. We’ll simultaneously make the same comparison for highest save on-level against middle save 2 levels above.
cantripSpoutSavesLevelsCharts [
SaveSelector.Lowest, 0;
SaveSelector.Middle, -2;
SaveSelector.Highest, 0;
SaveSelector.Middle, 2
]
Those two graphs are extremely close together!
This result is somewhat intuitive if you understand the way that bestiary creatures scale and the frequent gaps between higher and lower saves in the Building Creatures section of the rulebook.
A creature’s defensive scores almost always increase by 3 every 2 levels. And the gap between a high save and a moderate save is almost always 3. A moderate save and a low save is also almost always 3.
This suggests that, compared to targeting a middle save, targeting a low save often makes the creature feel like it’s 2 levels lower. And targeting a high save often makes it feel like it’s 2 levels higher. At least, for the purose of the action you’re currently doing!
Off-guard and weakest saves#
Let’s take a final look where we compare a martial’s backup shortbow to our hypothetical Spout cantrip that always targets the weakest save to see how they fair against each other.
cantripSpoutWeakestVsShortbowCharts 0
cantripSpoutWeakestVsShortbowCharts 2
cantripSpoutWeakestVsShortbowCharts 4
cantripSpoutWeakestVsShortbowCharts -2
Here we can see that off-guard represents a significant boost when creatures are of similar or lower level.
Off-Guard for the backup shortbow: about 30% higher
Due in part to benefiting from Off-Guard twice (once for each Strike)
A bit higher against PL+4 where it’s extremely difficult to hit regularly
A bit lower against PL-2 where hitting is extremely easy
Off-Guard for the Telekinetic Projectile: about 25% higher
Pretty standard for a single Strike against an Off-Guard
Much higher (like 40%) against PL+4 where it’s extremely difficult to hit
A lot of that hit chance is coming straight from the Off-Guard condition
Weakest save for our Spout: about 15% higher
Saves benefit more slowly from swings in accuracy because damage is usually halved on a successful save. So a lot more of the probability space is scaling up from .5 to 1, whereas with AC it’s scaling up from 0 to 1.
That said, if you’re slinging debuffs that are much more powerful than some extra damage, like Slow denying 1 action vs. up to 10 actions going from success->failure, or Fear giving more than twice the value on a failed save compared to a successful one, then that “15%” value number is much higher in practice. It’s just not as quantifiable as damage.
Conclusions#
Shortbows are often better backup weapons than arbalests, unless you’re pretty sure you’ll be able to Strike twice or need higher base damage to get past resistances.
The higher level the enemy, the more valuable debuffs to their defenses becomes.
Targeting the lowest save is like ignoring 2 of the enemy’s levels! Targeting the highest save is like making the enemy 2 levels more defensive!
Exclusively in the domain of strictly average damage, the relative -2 from Off-Guard is often more powerful than the relative -3 of targeting the weakest save.