IT’S PAYDAY FELLAS

Lately, I’ve been thinking about ways you could spice up the general playstyle of PF2E. Y’know how some feats like the entirety of the Medic archetype are godlike, because you almost always have 10 minutes to patch up, while things like Armour Assist fall to the wayside, because there is just no scenario where you have exactly 1-5 minutes to prepare, and not more nor fewer? How there’s almost no point in nonlethal attacks because every combat is to the death unless it makes sense storywise, or how having high carrying capacity doesn’t really matter because most people just handwave it, especially on Foundry where they can just slap loot into the hammerspace of the part sheet?

That’s when I thought: hey, I could do something PAYDAY-esque in PF2E. You know, with waves of enemies, between small breaks that the players could lengthen by trading in hostages. Hell, make the players attack a place, then defend their position while a drill gets through a vault, that’d be cool!

Funnily enough, this idea popped into my head originally unrelated to any of the above: I was trying to simplify looting, and I joked that the loot should be “whatever you can get out of the vault”. The idea was built on that more later, as I was pondering what would cause PCs to get slowed down or discouraged from taking all the loot out.

So, after a bit of a delay of about 5 months, I finally put together a playtest game where I would test all of this, and more.

Parameters

I announced the game in the Team+ Discord, explaining most of the overall details of the game. I allowed the use of SF2 content (I did plan on making this a modern-day bankrobbing game, so having guns around was just helpful), and I additionally allowed almost any 3pp content, but only after reading through it.

Ultimately, it didn’t matter too much. I got four players, and they brought eight characters to the game:

• Clutch was an Interrogation Investigator Catfolk, who was the primary technician of the crew. He specialised in Crafting, and was the person to get the crew through locked doors, electronic security measures, and the like. He also had the Vehicle Mechanic dedication to spruce up the escape vehicle.
• Mongrel was a Demon Summoner Dromaar Elf, with his Eidolon “Long silence pierced by flashes of rapturous bloodshed” (not a joke), who proved surprisingly useful in the social stealth aspect of the game.
• Nail was a Splintered Manifestation Oni Summoner Nephilim Human, with the Eidolons Hammer, Tooth, and Want. Nail and co’s ability to hold down multiple fronts made them a versatile glass cannon.
• Vanbury was a Bombard Soldier Catfolk, who was really good at decimating groups of enemies. She was also the crew’s designated driver during the car chase scene.

The party was level 3, and I allowed them the Basic Pack from Feats+ (i.e., one extra level 1 class feat, one extra level 2 class feat, and one extra level 1 general feat).

The Preplanning

After they created their characters, I had the players enter something called preplanning. Anyone familiar with PAYDAY 2 knows what I’m talking about—you get a floorplan / blueprint view of the bank, information about various notable rooms or areas, and heist asset opportunities.

For this game, I chose not to use favours—PAYDAY 2’s secondary limit on how many assets you can bring to a heist—and instead gave players an extra 750 credits (silver) beyond what they got from starting at level 3 to spend as they seek fit. The players elected to get a more secure and less exposed entry to the bank in the VIP parking lot, and planned to use the maintenance areas to move about the bank. They also took along both the thermite rig (free, but slow to open the vault) and a drill (expensive, but gets through the vault much quicker), and got themselves some extra cable ties, body bags, and other equipment.

Preplanning also informed players on what enemies they would be facing, but I believe this would only prove useful after the players had more experience in the game, so it wasn’t exploited all that much this game.

The players also had the opportunity to dead drop a few items into predetermined locations in the bank—for some credits, up to 3 Bulk of items would be hidden in there for them to be picked up. This did not get used, but I think this is fine: the players simply chose the more expensive, but more hidden entry, and thus managed to avoid major checkpoints that would necessitate them to hide their equipment in some way.

In addition, I also gave the players one week of downtime. They could use this for any downtime activities, or to do preparation activities for the infiltration. The party sent in a poisoned cake to the guards at the vault, which (unknown to them) they critically failed, which had repercussions later. The players had the opportunity to extend the preparation duration, with each additional week decreasing the loot by 10% (down to 9 extensions for 9 extra week but only 10% of the overall loot), but they didn’t take this opportunity.

The players were made aware that the bank held 10.000 credits worth of valuables in the vault.

Stealth / The Infiltration

For the bank, I handmade a quite large (70×80 squares, or 350×400 feet) map with a bank with 3 floors. The bank itself was about 34×38 squares of “playable” space each floor, or 170×190 feet.

Stealth progressed mostly as normal for the Infiltration subsystem, though I took the Red Mantis Assassination activity from Mark of the Mantis (renamed to Silent Takedown), allowing players to kill most NPCs that could cause trouble for some potential Awareness Point increases.

The players chose to enter the bank from a side entrance, by buying access to the VIP parking, and moving through a maintenance entry from there. Clutch had managed to get the crew into the bank undetected through picking the maintenance door, and after Tooth silently took down a guard, Mongrel snuck up to the second floor in a guard disguise, took out the bank’s manager, and got Clutch in to start hacking her computer to get the managerial authentication to start the timelock.

The timelock was a door that separated the two parts of the bank, the “business” side and the vault area side. The timelock itself took 30 minutes to open, though as they managed to do this while still in Stealth, I merely gave them each an individual obstacle to keep a low profile with the timer ticked down.

A few unfortunate failures with the managerial authentication forced Mongrel to make his way down to the (very busy and very full of NPCs) offices, where he successfully bluffed and intimidated his way through, and managed to get an NPC to pass a two-factor authentication.

This was when things went to hell. The players forgot there were still guards in front of the timelock door, so when it started to open, they immediately took some Awareness Points from the guards becoming suspicions. The timelock door opening was also the trigger of their critically failed cake preparation, gaining them enough Awareness Points to cause combat to break out with a patrol of guards who went and checked out what was going on.

The guards at the vault already expecting their approach, the crew decided to go in guns blazing, finally entering the vault area.

Control and Assault

The crew dispatched the vault area guards relatively quickly, suffering one down from Nail and their splintered manifestations overextending, but otherwise getting through the fight with only some HP lost. As the HP losses were building up, they decided that it’s time to get the drill dropped in, and started collecting hostages.

They exchanged enough hostages to buy 10 minutes worth of time, enough to (fail to) patch themselves up with Treat Wounds, and move the hostages deeper into the bank. After some planning, they decided they’d funnel the incoming cops through the timelock door to the vault.

When it came time for the assault, the crew got really (un)lucky, rolling the same type of enemy three times almost in a row (this had about a 0.075% chance of happening), but they could luckily output enough area damage to handle them.

To represent the swarm of cops from PAYDAY 2, I unceremonously stole a troop rework idea Clutch’s player came up with, with some small reworks of my own—christened to “squadrons”, these enemies used the Troop Movement and Form Up rules, but instead of a mass of tokens representing inordinate amounts of enemies, each token represented a single individual of the squadron. The HP of the overall squadron was calculated using creature HP for that level per the building creatures rules, but was divvied up between each token evenly—for example, if the squadron’s overall HP should be 80, and it had 4 tokens, then that meant each token had 20 HP.
Additionally, tokens were affected by conditions and effects separately, unless said condition or effect affected their movement or attack modifier, in which case it was ignored for that purpose unless half or more of the squadron was affected. This meant that you could individually off-guard a member of the squadron, but they’d only be taking the speed and attack roll penalties from being suppressed if half or more was affected at the same time.
For attacking, they used a different rule than troops: they rolled once for an attack roll, then attacked everyone within reach or first range increment. However, the squadron must select a “primary target” when it does this, and this primary target is the only one they can critically hit—for everyone else, should they crit, their result will be downgraded to a normal hit.
Squadrons effectively acted as fodder for the PCs to nonlethally take down, and make hostages out of.

I mentioned that the crew “rolled for” an enemy during assault: this is to be understood literally.
For assaults, I devised a system. Every assault was made up of waves of enemies, starting at 4, then increasing by 2 every assault. The assault spawned enemies every 2 rounds during the assault. For the spawning algorithm, I decided upon an XP amount (equivalent to a moderate encounter for the party, 80 XP), then rolled from a table to determine what spawns, using the Creature XP and Role table to determine the “cost” of each enemy. I kept doing this until I spent as much XP as I could from this 80 XP. If I couldn’t spend it all, I would carry at most 20 XP over to the next wave.
This resulted in situations, where a more powerful enemy spawned alone, because I didn’t have enough lower level enemies to fill out the remaining XP, or couldn’t roll them.

I was pleased to see that while the party took quite a bit of damage, they didn’t immediately keel over from being overwhelmed. This is roughly what I was expecting, and what I wanted.

Cleanup

After a short discussion about the session with the assault, I decided to give every ranged, reload 1 or above weapon (and every weapon with the repeating trait) an additional die of damage. This die was akin to the one given by Vicious Swing; it does not count as weapon damage die, thus does not increase stuff like Gravity Weapon. I did not give this benefit to enemies—their damage is already set by the Creature Building rules.

This had an immediate positive effect. I was expecting this to tip the scales of balance towards ranged attacks over melee, but melee’s consistency with its guaranteed Strength-to-damage still saw use. The general feedback from players was that it made ranged weapons not suck ass to use, which I’ll take.

The crew gathered up enough hostages to able to skip any further assaults, and made it out with all the money in the bank. Though they chose an escape option that allowed them good chances to escape, we decided to skip rolling for that, and tried out the car chase.

Car Chase

The car chase rules were the vehicle chase rules lifted straight from Starfinder 1. I mapped most pilot actions as 1 action activities, but Engage and Speed up, I made 2 actions. I removed Double Manoeuvre—that was just given by the 3 action system of PF2. I instead gave every pilot action except Ram and Slow Down a unique “piloting manoeuvre” trait that applied a MAP-esque penalty to multiple piloting actions in the same turn. This came up twice altogether, and then I promptly forgot about actually applying it. Going back though, it wouldn’t have changed any of the results.

What we found was that none of the Starfinder 2E weapons actually had enough range to comfortably hit enemies or vehicles that they weren’t directly engaged with. With the exception of a comically lucky Strike-Strike-Strike turn from Mongrel who completely erased a cop car from the face of this good Earth with his Laser Rifle, most PC turns were spent either aiding or getting skipped, while most enemy turns were spent on the Piloting Actions trying to get close to the PCs.

Vanbury’s player (correctly) foresaw their 15 ft. Scattergun being of less-than-zero use in a car chase, so they took to driving. The bonus from Clutch’s Vehicle Mechanic signature vehicle came in, well, clutch at several points, being the deciding factor in several Piloting checks being successful.

Normally, creatures in a vehicle chase benefit from a bonus to their Piloting checks and attack rolls against vehicles and creatures that they’re ahead of. I created an effect to represent this in Foundry (mapping it to a +2 circumstance bonus), but thanks to the general scuffedness of the weapon ranges, I also completely forgot about this.

Conclusions

• Giving players info ahead of time about the “dungeon”, and options on how to enter and leave, should be done more. I had the players plan their heist for a good half an hour to an hour during Session 0 after character building. It was honestly very fun to see them plan their moves out. Additionally, I originally wanted to not reveal how long it takes to open a vault with any of the vault cracking options, but I decided to do so on a whim—after seeing player reactions, I do believe it’s best to keep this info in the open.
  Additionally, giving about 2-3 “big” decisions, like how to enter and how to leave, seems to be the perfect amount of choices to not cause a choice paralysis. In both cases, one of the decisions already implied going loud, further limiting the players’ choice for their own benefit.
• Slash SF2 ammo prices in 1/10. Holy fucking shit, the economy’s in shambles. I expected my players to be considering ammo, sure, but not this much. PF2 ammo genuinely has a ratio of 1 cp to 1 ammo, usually, but SF2 went with 1 sp to 1 ammo. At low levels, this is terrible and makes PCs barely be able to afford anything fun.
• Make it more abundantly clear that the Stealth entry is gamified. Surprisingly, despite both Pathfinder 2E and PAYDAY 2 being actually quite lenient with NPC detection, every player thought about the infiltration as if a single failure could blow their cover. I used Mark of the Mantis as a base, and that adventure has you doing this kind of infiltration, down to sneaking, fighting, etc., in 3×3 rooms.
  Though I mentioned that the locations and the things happening in Stealth are not exact, they’re more abstracted, the players were still less brazen than I expected them to be. This resulted in half the party sitting in a safer part of the bank, while the other half handled the entry to the vault.
• Fill the map with guards with the assumption that the players will somehow trigger ALL encounters at the same time. This is honestly a skill issue on my part, but I definitely added too many guards to the bank. There were some 16 guards in the bank—half of that would have been more than enough.
• Ranged meta woes. I was admittedly quite surprised at how… nonexistent the “ranged meta” of Starfinder weaponry is. We found both the range and the damage of the weapons too low for our liking. I mentioned how we solved the damage for the game, but I’ve entertained the thought of doubling the range of all non-thrown weapons after we looked through some of the SF2 weapons—to wit, an IRL gyrojet pistol’s rocket only speeds up to max speed at about 30 ft., while the in-game version has a range increment of 20 ft., and the autotarget rifle’s range was genuinely halved between editions.
• Hostage negotiation be its on phase, allow for trading hostages with resources. This was more of a suggestion by the players, but I ended up liking it. For now, I just had players be able to trade for hostages any time during Control. But by codifying it in its own phase, I can make it more clear what’s going on, and force players to think ahead some. The “trade hostages for resources” was another suggestion that I really liked, as I do want to bleed players dry of their hostages (as it’s extra time), and if they wanna spend them on ammo / heist assets, so be it.
• AAAAAAAAAAAAAAAA!! I NEEEEEEEEEEEEED A TWO-ACTION HEAL!! Though I got what I wanted with the game in regards to how wounded the PCs got, the players felt it mandatory to bring (and cast) Heal. I don’t really want anyone to healbot unless that’s what they want to do, so if I next run a game like this, I’ll probably use the Stamina variant.
• Waves should be slightly less random. The general “randomly generated waves” idea is good, but it can lead to assaults being frontloaded, sputtering out at the end, or both. Current ideas are either:
  • Make first and last wave of the assault severe instead of moderate,
  • Generate waves, instead of individual enemies,
  • Use a “deck” of waves, shuffling them at assault start, then drawing from them during the assault (a similar logic to what Tetris uses with its tetromino generation).

  I’m currently leaning towards the last one.