Starfinder: Soldier’s Punishing Salvo

A Soldier with a Stellar Cannon shoots with Area Fire at two targets (a primary and a secondary), followed by a Punishing Salvo against the same primary target. How’s the damage distribution? What are the chances that the targets will be suppressed?

# Install in JupyterLite
%pip install -q pathfinder2e-stats

import matplotlib as mpl  # noqa: F401  # Needed by JupyterLite
import numpy as np
import xarray

import pathfinder2e_stats as pf2

Note: you may need to restart the kernel to use updated packages.

level = 5
atk = (
    pf2.tables.SIMPLE_PC.weapon_attack_bonus.soldier.sum("component")
    .sel(level=level)
    .item()
)
area_fire_DC = (
    pf2.tables.SIMPLE_PC.area_fire_DC.soldier.sum("component").sel(level=level).item()
)
print(f"{atk=}, {area_fire_DC=}")

atk=14, area_fire_DC=22

weapon_dice = pf2.tables.PC.weapon_dice.improvement.sel(level=level).item()
weapon_specialization = pf2.tables.PC.weapon_specialization.soldier.sel(
    level=level
).item()
stellar_cannon = pf2.armory.starfinder.ranged.stellar_cannon(
    weapon_dice, weapon_specialization
) + pf2.armory.upgrades.auto(level=level)
stellar_cannon

Damage 2d10 piercing

enemy = pf2.tables.SIMPLE_NPC.sel(level=level, drop=True)[
    ["AC", "saving_throws", "HP"]
] * xarray.DataArray(
    [1, 1], dims=["target"], coords={"target": ["primary", "secondary"]}
)
enemy.isel(target=0, drop=True).display()

variable	AC	saving_throws	HP
challenge
Weak	16	6	31
Matched	21	12	75
Boss	25	18	148

# Primary and secondary targets use the same damage roll, but save independently.
# 'challenge' is a what-if analysis - let's compare the same dice rolls against
# progressively harder-to-hit enemies.
pf2.set_config(
    check_dependent_dims=("challenge",),
    check_independent_dims=("target",),
    damage_dependent_dims=("challenge", "target"),
)

primary_target = pf2.check(atk, DC=enemy.AC)
primary_target["outcome"] = xarray.where(
    primary_target.target == "secondary",
    pf2.DoS.no_roll,
    primary_target.outcome,
)
primary_target = pf2.damage(primary_target, stellar_cannon)

area_fire = pf2.damage(
    pf2.check(enemy.saving_throws, DC=area_fire_DC, primary_target=primary_target),
    stellar_cannon.area_fire(),
    dependent_dims=["target"],
)

# Note: Primary Target does not increase MAP, but Area Fire does
punishing_salvo = pf2.check(atk - 5, DC=enemy.AC)
punishing_salvo["outcome"] = xarray.where(
    punishing_salvo.target == "secondary",
    pf2.DoS.no_roll,
    punishing_salvo.outcome,
)
punishing_salvo = pf2.damage(punishing_salvo, stellar_cannon)

full_round = xarray.concat(
    [
        primary_target,
        area_fire,
        punishing_salvo,
    ],
    dim="action",
)
full_round["action"] = ["primary_target", "area_fire", "punishing_salvo"]

Suppressing Fire

What is the probability of giving the targets the Suppressed condition?

All targets are suppressed when rolling a failure or worse against Area Fire. With Bombard, they are suppressed on a success (but not a critical success). With Action Hero, a target is suppressed when hit by a Strike.

suppressed_area_fire_lt_outcome = xarray.DataArray(
    [pf2.DoS.success, pf2.DoS.critical_success, pf2.DoS.success],
    dims=["subclass"],
    coords={"subclass": ["Action Hero", "Bombard", "others"]},
)
# Note: strikes against secondary targets have been masked with DoS.no_roll
suppressed_strike_gt_outcome = xarray.DataArray(
    [pf2.DoS.failure, pf2.DoS.critical_success, pf2.DoS.critical_success],
    dims=["subclass"],
)
suppressed = (area_fire.outcome < suppressed_area_fire_lt_outcome) | (
    np.maximum(primary_target.outcome, punishing_salvo.outcome)
    > suppressed_strike_gt_outcome
)

suppressed.mean("roll").stack(col=["target", "subclass"]).to_pandas()

target	primary			secondary
subclass	Action Hero	Bombard	others	Action Hero	Bombard	others
challenge
Weak	0.99616	0.94924	0.93882	0.74969	0.94954	0.74969
Matched	0.91010	0.94924	0.80043	0.45024	0.94954	0.45024
Boss	0.68163	0.65114	0.40156	0.15012	0.64953	0.15012

Mean damage

total_damage = full_round.total_damage.mean("roll")
total_damage = xarray.concat(
    [total_damage, total_damage.sum("action").expand_dims(action=["TOTAL"])],
    dim="action",
)
total_damage = total_damage.stack(col=["challenge", "target"]).to_pandas()
total_damage

challenge	Weak		Matched		Boss
target	primary	secondary	primary	secondary	primary	secondary
action
primary_target	15.40943	0.00000	9.92385	0.00000	6.06280	0.00000
area_fire	13.68238	12.61777	10.15106	8.14128	6.28547	4.83232
punishing_salvo	9.91375	0.00000	5.52224	0.00000	3.31635	0.00000
TOTAL	39.00556	12.61777	25.59715	8.14128	15.66462	4.83232

Damage distribution

bins = full_round.total_damage.max().item() + 1
_ = (
    full_round.total_damage.stack(col=["challenge", "target"])
    .sum("action")
    .to_pandas()
    .hist(bins=bins, sharex=True, figsize=(10, 10))
)

Let’s break down the damage distribution to the primary target:

_ = (
    full_round.total_damage.sel(target="primary")
    .stack(col=["action", "challenge"])
    .to_pandas()
    .hist(bins=bins, sharex=True, figsize=(12, 10))
)