Tuesday, September 4, 2018

Contacts (Insurgency)

In Delta Green, bonds are meant to be your character's emotional support network, NPCs you can rely on (at least to a point) to help you deal with the massive psychological trauma inherent in the job.

These bonds could be with your spouse, childhood best friend, Very Good Dog, whatever. The point is that they don't directly represent an in-game resource as written, though they do help agents recover some of their sanity between sessions.

That doesn't model an insurgency campaign very well. For one thing, your closest bonds as a dedicated freedom fighter, after a little while fighting alongside your comrades, are probably going to be with them, not anyone you left behind. For another thing, if you have every reason to assume you're being watched by people who want you dead, you don't go visit your family between missions.

This frees bonds up to do something more appropriate to the context.

Basically what I want is: someone at the table says "hey we need fake IDs to try and get on base and none of us have the forgery skill" and then someone else says "fine then I have a cousin Smirnoff who makes those for a living" and they write it on their character sheet and then the player cell has a new resource and the occupiers may or may not find a new way to get close to the PCs.

Rule Changes:
>Call it "contacts" instead of "bonds".
>Each contact has 85% in one professional specialty.
>You don't have to decide who your contacts are or what they do until it comes up in-game.
>Your contact score = CHA x 4.
  >>Roll under with d% to call in a favor.
  >>If the roll fails, they still have a 50% chance of deciding to help you ("but I swear this is the last time do you hear me!?"). Either way, their contact score goes down by 1d20 points.
  >>If your contact score ever reaches 0%, they refuse to help or even see you ever again.
  >>You can potentially increase your contact score by doing favors for your contact.
>There are conditions that add bonuses or penalties to these rolls (maximum total of 99%, minimum of 1%).
  >>Absolutely no chance of being found out: +20% bonus.
  >>Favor will likely earn money or status for the contact: +20% bonus.
  >>Satisfies a preexisting grudge: +40% bonus.
  >>Secrecy* score is 60% or less: -20% penalty.
  >>Secrecy* score is 40% or less: -40% penalty.
  >>Public Support* score is -10% or less: -20% penalty.
  >>Direct physical danger to self or family: -40% penalty.

You could, if you wanted to, game this system to give you a contact you could use to reduce your PC's mental strain, just like a bond in the rules-as-written. You'd just need to use a contact slot on a mental health professional who could then use their skill to help keep you from losing your shit.

*These are two of the scales I've been using to track how the PCs' actions affect the overall campaign.

Tuesday, July 10, 2018

Faux Silent Fever

Time flies when you’re busy as hell. This seems like a good moment to step back and take stock.

First order of business: a project I worked on won an Ennie! I did the graphic design for Fever Swamp by Luke Gearing, which it was announced last week won Reece Carter’s Judge’s Spotlight award.

Daniel Sell of Melsonian Arts Council dumped a big messy pile of really great ideas into Jarrett Crader (editor) and my laps. It was our job to make those ideas as usable at the table as humanly possible (and make sure the physical object looks, you know, elegant and shit). Both the print and pdf version have gotten great reviews and every review I’ve seen has at least mentioned the part we played. So that whole project feels like a pretty big win for me.

This has all been a validation of my belief that modules like Maze of the Blue Medusa and Blood in the Chocolate have made a pretty big chunk of the scene realize they were starving for intelligent, well-designed supplements, that they can actually use to make running great games easier. The medium of role-playing games is at a similar moment to one that comics went though, when Karen Berger was in charge of DC’s Vertigo imprint. RPGs are officially for grownups now.

Speaking of Maze, I’m also doing layout on a pretty extensive adventure called Silent Titans, penned by Patrick Stuart and being illustrated by the absurdly talented Dirk Detweiler Leichty who seriously needs to be doing this stuff full time. It’s been pretty slow going because we've been reinventing a few wheels, but holy shit will all this back-and-forth experimentation and refinement be worth it.

And one thing that’s juuuuuuuust about ready to release: Faux Pas, written by Beloch Shrike of Blogs on Tape fame and illustrated by Anxy (plus a couple maps by me,  and it's edited by Jarret Crader). It’s the first issue of a new zine called Hocus. It’d work for a one-shot, but really it’s the sort of module you’d want to slot into a hex on your campaign map, for your players to stumble into when they least expect it.

Oh and of course Beloch is recording an audio version of Faux Pas, just to see what the reception looks like. I’m pretty sure the usage case he’s got in mind is you listen to it while you’re driving or painting or gardening or masturbating or whatever and then when you actually want to run the game, just a glance at the description should remind you of what’s going on. Also maybe it helps out the vision impaired? That would be pretty cool, I guess we’ll see.

There’s more stuff lined up but it’s way too early to talk about most of it so I’ll just stop now.

Tuesday, January 16, 2018

Kovakistan Liberation Party

This is not about a political faction.

So… we’re at a point where everyone reading this has tried out Jeff Rients’ carousing rules right?

If you aren’t familiar with it, the idea is PCs can get a bunch of XP if they blow hundreds or thousands of monies throwing a giant party. If they do that, they have to make a saving throw vs. poison or get shitfaced and have to roll on the table to see what they did while they were blacked out.

The other evening, I was running Kovakistan and an unusual situation came up, where I actually wanted to play the party out. The PCs weren’t looking for bonus XP at all, they wanted to throw a party and invite everyone in the same high-rise project as their safe house to build support for the revolution.

I didn’t tell them this but the risk they were taking was that the party might get too loud (with celebratory AK-47 fire and patriotic shouting and whatnot) and be heard by a passing USIncorporated patrol, possibly triggering a massacre that would be the PCs’ fault.

There wasn’t anything stopping me from still using the carousing rules but in a different way, making the PCs roll saving throws to avoid blacking out whenever they get more fucked up. But some of these players were doing a good enough job of keeping up with their characters that there was no need to artificially handicap their judgement.

If all goes well and all encounters are dealt with successfully, public support for the revolution goes up by 1% for every hour the party goes on for (this is for Delta Green, which uses a resource roll instead of keeping track of the players’ money. if i was using this for D&D, it’d just be 1% per 100 monies spent on the party or something). If your characters are open about their identities with the party guests, this also affects public support for their cell and/or faction specifically, but there’s a risk of someone giving away your location to USInc. forces.

Every hour (or more often if you feel like it, just adjust the public support reward to compensate for the added risk) roll 1d8 and 1d10. If the d8 results in a 1, two Kovakistani Police officers show up. They’ll call for backup and try to shut the party down if they aren’t dealt with somehow. KPs are usually open to large enough bribes, but they'll know they can come back to shake the PCs down in the future. Kidnapping them and using them for an execution video may be an effective propaganda tool, but would harsh the partygoers’ buzz enough for them to leave.

(and if the KPs had recognized any of the PCs from wanted posters the jig would be up, but they had all disguised themselves earlier)

The d10 is to see what happens at the party that the PCs now have to deal with. Things can happen more than once; ref's discretion whether they involve the same NPCs as the first time or not.
1. Someone has gone and grabbed their AK-47 and starts shooting into the air, while bellowing the Kovakistani national anthem. If this keeps up long enough for the Americans to notice, they’ll send a reaction force to “disperse” this illegal gathering.
2. A skeezy looking dude with a coke ring and his hair slicked back just pulled up in an American muscle car. He’s got a duffel bag in the trunk full of drugs that he’s willing to share, but it won’t be long before he starts creeping out the women. If they start leaving, everyone else will, too.
3. Someone is stealing peoples’ cell phones (1d20+6 so far) and people are starting to notice they’re missing. The thief will try to sneak away after realizing the jig is up. Public support goes down by 1% for every phone lost, so catching the thief is advisable.
4. A couple twenty-something dudes with a few years of krav maga training have lost their shirts somehow and are fighting for fun in the parking lot. People are laying down bets (yes of course the PCs can get in on this). Everything is fun and games unless someone loses an eye. If a PC challenges the winner and beats him, public support goes up by 5% in addition to the normal boost just for throwing the party. If the PC injures their opponent badly enough to hospitalize, maim, or kill him, public support goes down by 10% instead.
5. Someone catches their significant other at the party with the S.O.’s side piece. First words then blows are exchanged and a pistol will soon be drawn.
6. Basically the same as #5 except it’s either a rap battle or a break dance battle (referee’s discretion).
7. Someone who is too drunk to remember why is bleeding everywhere. Like a big-ass stream of blood spurts out with every heartbeat. If this person dies at or after leaving the party, no public support points will be awarded for the night.
8. A group of partygoers leaves for a while and returns with a goat and a chainsaw, with which they sacrifice the goat. Everyone yells HAIL SATAN and the public support boost for throwing this party is doubled.
9. The booze and/or drugs are starting to run low. The PCs have 2d6*10 minutes to replenish the supply somehow or the partiers will start to leave.
10. Someone decided to light a bunch of shots of dangerously strong basement hooch and the resulting kitchen fire surprised everyone enough that someone else dropped the bottle in the middle of it and now the safe house is on fire.

Wednesday, December 13, 2017

How To Make Pathfinder Good (in like 6 or 7 easy steps)

It has come to my attention that there are people who still play Pathfinder.

I get it, I guess. There are some things I genuinely like about the system. The way it facilitates gonzo parties made up of PCs out of the monster book is kinda cool. The threefold save system is a genuine innovation, even if the rules don't leverage it as much as they could (more on that in a minute).

The core issue with the game, for me, is the crunch. The sheer number of rules creates two problems for me: how long everything takes in comparison to most other D&D retroclones, during character generation, actual gameplay, and DM prep, and mechanics that reward system mastery more than player creativity. The first problem makes the game hard to run, and the second makes it not feel worth it to run.

So I started thinking about how to turn Pathfinder into a Good Game (as arbitrarily defined by me).

Not just like, playable, actually good, like what would it take to turn this into something I would choose to play?

Throw out the entire skill system as written and make the threefold saving throws do double duty. Fortitude still means fortitude but now it also means “stuff fighters are good at” like athletics or horse riding. Reflex maps to thief stuff, will maps to wizardpriest stuff. Now whenever you need to make a skill check, the DM tells you which ability score to add to which saving throw/skill score and you add that to a d20.
You can still keep class skills for the sake of niche protection by assigning advantage to those tasks as a class ability. So rangers, druids, and barbarians get advantage to checks involving nature stuff, bards get advantage to rocking out while everyone else does the work, etc. Rogues would get advantage to thief stuff *in addition to* having really high reflex saves and usually having good dexterity.

…by eliminating them.
This would run the risk of nerfing fighters. You can avoid this by letting them use ad-libbed mighty deeds like in DCC, just replace the funky DCC dice chain with 5e-style advantage like so:
fighter pc: i wanna charge the ogre and knock his ass off the cliff
dm: roll two d20s. if one hits, you do normal damage. if they both hit, it does normal damage and also knocks his ass off the cliff
fighter pc: fuck i rolled a 1 and a failure
dm: he sees what you’re doing and steps aside at the last second and your ass goes off the cliff
ex-fighter pc: dammit. oh well at least you adopted those house rules from that blog that make chargen take ten minutes even for drunk noobs. imma be a troll bard this time 

what the fuck would a troll bard write songs about

Ignore all of the generic magic items in the book and most of the unique ones. They all suck. +x longswords are boring and trying to hide that with extra description or history just makes it worse.
Instead, come up with unique items whenever possible. If you need a random table of semi-generic items like the rulebook provides, recycle the more interesting feats you threw out in Step 2. A scimitar that lets you do a whirlwind attack or a wand that lets you extend or empower spells would both be more fun than the equivalent lower tier magic items in the book. Remember, items that give players new options are always more interesting and fun than items that just make them 5% or 10% better at what they already do.

The first thing everyone everywhere should throw out of their game is alignment. It’s terrible as written and I would argue that it doesn’t add anything valuable enough to make it worth keeping.
cons (if run by the book): eliminates interesting gameplay that results from moral ambiguity, can lead paladin types to adopt genocidal attitudes toward creatures listed as evil in the monster manual, prescribes and/or judges pc actions based on an outmoded Abrahamic conception of morality, limits roleplaying opportunities for players that want certain powers (why the fuck can’t there just be a paladin of Tiamat or a drunken master that’s just a regular monk without the alignment restriction you don’t need a whole new class for that shit if you just don't put the restriction in), allows the party or even NPCs to skip situations where they would otherwise need to use roleplaying and player skill to determine if someone is evil or not like why even have investigations and trials at all you know what I mean, probably a lot more but I'm done with this list
pros: uh it makes smite and detect work

You can make smite and detect work without alignment though.
1. Any divine powers like smite evil that are designed to encode crusader-like values into lawful good religions now affect enemies of the PC’s specific faith. What qualifies as enemies is pretty much up to the PC and DM to determine as they go, once the religion has been sort of outlined ahead of time.
2. Any arcane spells or abilities like detect evil are now just called “detect” (for example) and the PC would have to say what they want to detect when they cast it. It’s up to the DM to make on the spot rulings about what sorts of things are too specific or not specific enough to target with the spell, but any of the creature types should probably be fine. Like:
wizard pc: alright we don’t know where the wounded minotaur is hiding so i cast detect evil
dm: good and evil are societal constructs and have no place interacting with a form of magic that focuses on rearranging the fundamental structure of the universe
wizard pc: fine, nerd. i cast detect monstrous humanoid
dm: you sense that the minotaur has stuffed himself into a barrel that kinda seems a little too small for it

or even just:
wizard pc: wait i think this npc is a demon in disguise. detect outsider

The challenge rating system that by-the-book XP is based on is too much damn work. It can make DM prep take forever just to plan out an encounter that, in more rules-light versions of D&D, you’d be able to randomly generate on the fly. It also tends to make it difficult to assign XP rewards for any traps or puzzles that are interesting enough to not fit into the strict guidelines presented in the rulebook.
This is one of those problems that has annoyed so many people so much that there are easily hundreds of totally valid takes on how to fix XP on blogs and social media. I won’t pretend that the way I’ve started handling it is the best for everyone, but this is how I run it in most of my D&D-like games:
-Your PC levels up every time they achieve a set number of challenges. I’ve been going with 13 for good luck. If I ever decide that I want progression to move faster or slower, that’s the only number I’ll need to adjust.
-Valid “challenges” are obstacles that pose a legitimate threat to either the party’s safety or their goals, and require some level of player skill to negotiate. So solving any puzzle door would qualify, and so would figuring out and negotiating a trap, as well as winning any genuinely threatening combat encounter (defined as beating anything with more hit dice than you have levels). Slaughtering a village of goblins won’t give a level 7 paladin any XP unless there’s like a surprise dragon or something.
-I also count spending 100 monies on just partying in between adventures as a “challenge” for leveling purposes to encourage situations in which players must roll to see what their PCs did while blackout drunk.

I just realized that getting rid of skill points and feats kinda fucks humans over since that’s all they normally get at chargen. I say just start ‘em off at level two and be done with it. If nothing else, that should encourage players who don’t care what race their PC is to just pick human and move on.

Almost all the races in the game (pretty sure this applies to every non-human, right?) get a +x bonus to one or more ability scores and a -x penalty to others. Which I guess is alright, but it’s kinda more fun to interpret those bonuses as extra dice. So like an elf would get to roll 4d6 for dexterity but only get 2d6 for constitution.
I’ll admit that this one is more optional really (all of these are optional dummy). I like it because the whole point of allowing players to role-play as non-human species is to add variety, so any change that accentuates that variety by making differences more extreme is worth making.

Okay now I’m getting into aesthetic preferences instead of just ways to make the game run more smoothly so I’ll just quit while I still feel like I’m ahead.

Friday, November 17, 2017

Kovakistani Mountains Die Drop Table

In a guerrilla warfare campaign with a foreign occupying force, the local terrain is going to become the resistance’s greatest weapon. The earliest historical anecdotes I’ve read examples of that statement in were rebellions against Roman Imperial rule, but I’m sure it goes back further than that.

At the end of the first session of my KLA campaign, the PC cell fled the capital city to head into the mountains. Which meant that my prep for episode two basically consisted of making another one of these:

sized to fit in the bottom of a 12 pack of ramen (as all die drop tables should be)

This one’s a little bit different from the two previous ones I posted, partly because of the more realistic genre, but also to account for less extreme weather conditions, variable alertness levels for local units of the occupying force, and human-made infrastructure. The other key difference is that (unless they REALLY fuck up) the PCs will be sticking to the roads, which means that smaller wilderness features won't generally even be noticed, at least not as readily as if the party cell was hiking through the woods. 
The random alertness (bottom edge) in particular makes traveling more of a hassle by increasing the number of encounters per journey from the standard 1 in 6. It's also helpful when figuring out how any soldiers encountered should behave by default.

Here’s a more legible version:
click here to download pdf

I'd like to say I'll clean this up at some point but that's what I said about the jungle one like a year ago so let's assume I care more about pure functionality with these.

It may just be that my players were really on it, but this worked really well in practice. The random terrain and constantly shifting fog added an element of tactical variety that gave the PCs room to deal with pretty repetitive encounters differently in different situations.

Tuesday, November 7, 2017

After Action Report

Decided to give this freedom fighter campaign a go starting last evening. I’m still sorting this out in my head so I guess you get to watch me do it in real-time.

I’m using Delta Green/Call of Cthulhu rules, with the house rule that players can come up with their Bonds (NPC relationships) on the fly as needed. So far this has been used twice, to establish a base of operations each time (they kinda fucked the first one up already which is a little impressive in a way).

I’m not currently planning on adding any of the supernatural weirdness that game was made for, but the equipment and skill lists are pretty much perfect for this and there’s PTSD rules baked in.

Pretty much every mechanic in DG is based on a roll-under percentile mechanic, so that’s how I’ll be approaching Secrecy, War Weariness, and Public Support. Secrecy starts at 100% and goes down as the PCs make contacts and build up a reputation. War Weariness starts at 0%. Public Support also starts at 0%, but goes from -50% to 50%. This is because I don’t think I’ll ever need to roll under it, I’ll just be using it as a bonus or penalty to other rolls.

The setting is Northern Kovakistan, a fictional country somewhere in eastern Europe, which has been invaded by America Inc sometime in the near future. The PC cell’s major goals at this early stage are to get some cash and connect with a larger terrorist network to gain access to intel and equipment.

The party has already done a few things that affect these three scores. In order of occurrence:

1. They robbed a store for bomb-making supplies.

2. They killed the store owner.

(Even if 1 & 2 aren’t linked back to the PCs, it hurts Public Support for the revolution in general when civilians are targeted. Secrecy refers to the PC cell, but Public Support is for the whole movement.)

3. They did manage to not only blow up an American convoy without any direct civilian casualties, but also to get the explosion on film.

4. They betrayed the only bond which had been established at that point, Pizza Josef, who was the employer, landlord, and possibly illegitimate father of one of the PCs, and got him arrested in connection with the convoy attack.

5. They staged the attack literally right outside their base of operations (although i think the original plan was for the bomb to go off inside the restaurant which… well fuck Josef I guess).

6. They completely abandoned that base of operations.

5 and 6 would normally both be pretty extreme adjustments but in this case they mostly just cancel each other out. Although they would probably have left some clues behind… I’ll have to get them to make a Forensics check at the top of next session to see how well they cleaned up. If the check fails, Secrecy goes down by the same amount the check fails by.

4 is easy. The party set Josef up to take the fall, he knows it was them, and is gonna tell the Americans everything he knows even before they start torturing him (which they will do anyway). The strength of that bond was 13%, so Secrecy is gonna take a 13% hit. This represents the value of whatever intel Josef was able to give the interrogators. If the bond had been stronger, he would have known more about the PC, so the Secrecy rating would suffer more of a penalty.

1 & 2 are definitely going to directly affect Public Support, which will indirectly affect Secrecy. For now I’m just going to say that each fuckup (including multiple fuckups in the same event) results in a 5% penalty. So this puts Public Support at -10%, which means a 10% penalty to Secrecy.

3 will increase War Weariness. Once the PCs release the video, it will also increase Public Support. Five soldiers died in the attack, so War Weariness is now at 5%.

(This is a little unreasonable, in reality tens of thousands of soldiers need to die in vain for the civilian population to start caring, but I’m only counting deaths directly caused by the PCs and you can always assume there’s a lot more going on in the setting than just what they’re up to and besides it would take forever even to get as high as 50% otherwise.)

So now I need to make two out-of-game rolls to see what’s going on in the world.

The PC cell’s Secrecy rating is currently 77%. I rolled 03, so the authorities are not yet onto them. Every time this fails, the occupying force will get one step closer to the PC cell by connecting some dots about their activities or interrogating one of their contacts or something similar.

The War Weariness of America Inc is at 5%, so a 93 tells me that nationalist jingoism is doing quite well thank you very much. If I ever start rolling under this, dissent will start growing in the occupiers’ homeland, germinating in academia before spreading to the voters and eventually the politicians and business leaders.

Oh, if you’re curious, these are the pregens I threw together. We ended up with an engineer, an ex-secret police inspector for the deposed Kovakistani regime, a burglar, a journalist, and a smuggler for this session. No one went for the shock troop or the paramedic.

Thursday, October 26, 2017

bella ciao bella ciao bella ciao ciao ciao

Pretty much every OSR setting is horrible for the people who live in it.

Which obviously isn’t for everyone. I’m told there are people who prefer optimistic settings they can use as a pure escape from the everyday horrors of real life. Some of them look at an RPG setting that features rampant slavery (for example) and see a product that explicitly supports slavery. Whereas I mainly see an opportunity for the PCs to roleplay as freedom fighters.

I’ve been thinking about how to codify the revolutionary story arc. Not as an adventure path or anything scripted like that, but with a set of mechanics that shapes the direction and pace of the campaign based on PC activities and a bit of randomness. Ideally this would work with any setting where the status quo sucks and pretty much any ruleset, just a short, slimmed down set of mechanics designed to completely redirect the flow of whatever campaign you’d want to bolt it onto.

TSR Marvel’s Nightmares of Future Past gets part of the way there with its Search Flow Chart, which keeps track of how well hidden the PCs are from their sentinel oppressors. The secrecy of their hideout (in this case abstracted into more or less powerful jamming devices) is the only factor that comes into play though, and it doesn’t do much to build tension unless the players can actually see it. It kind of works in this high-tech setting, and you could easily replace the sentinel’s scanning devices with divination spells in a high-magic campaign. This doesn’t cover low-magic or modern though. I think it’d be pretty fun to play as members of the French Resistance, or Viet Cong, or any number of Roman-occupied territories, or any of the European colonies from the 15th to 20th centuries. Or fuck it, the US after Trump’s re-election.

For those campaigns, it would make the game more interesting to keep track of more than just the secrecy level of the PC cell (assuming good graphic design that actually allows the referee to do that without thinking any harder obviously).

Popularity is clearly pretty important too. It would go up when the occupiers cause collateral damage and down if the PCs or one of their parallel cells do. When it’s high, secrecy goes up, cause ain’t no one trynna snitch. 

It seems like a pretty safe assumption that the PCs will be making contacts during gameplay, including business owners, officials, and members of other revolutionary cells. Anyone who can substantially help the PCs or hurt the occupying force, really. As the party builds up trust with these NPCs, they become more helpful when called upon. I’ll be referring to the sum of the strength of these contacts as resources. When this score goes up, secrecy goes down, because trusting anyone is fucking dangerous. (i guess this kinda replaces the secret tracking of henchman loyalty, which i’m fine with because it always struck me as awkward anyway)

Right now I’m thinking the best way of handling that is letting the players themselves explicitly decide which NPCs they want to maintain as contacts, so they always have the option to sacrifice a little secrecy in exchange for whatever in-game advantage that NPC can provide (equipment, intel, access to restricted areas, etc).

So it should be pretty difficult for the PCs to accomplish major goals without first seeking out resources, but doing so is dangerous. Stakes automatically increase as the game progresses.

Some of those contacts may also be public figures, but the venn diagram doesn’t completely overlap. Public figures could be scholars, religious leaders, politicians, generals, whatever. They can be opposed to or in favor of the occupation, and wouldn’t necessarily be native to the occupied lands. They would have influence on the revolution’s popularity, and it will be in the PCs’ best interest to keep those friendly to them alive, free, and vocal. 

Public figures are the key NPCs of the campaign, and should be written up and/or randomly generated ahead of time.

The war weariness of civilians in the occupier’s homeland increases as their casualty count gets higher or scandals come to light, and drops dramatically anytime the PCs or a rival cell attacks their homeland or otherwise provides what they see as a justification for the war. Assassinating military public figures on the occupying side should increase war weariness, but targeting civilian public figures may have the opposite effect. If this score maxes out, the occupiers will have no choice but to pull out, like the US in Vietnam. It’s possible that this score shouldn’t exist in ancient historical settings, but it would also be pretty easy to ignore.

I haven’t decided how exactly to handle random events yet, but they should (in no particular order) be general enough that they work in any setting, be affected by the various scores I laid out, motivate the PCs to do something if they aren’t already up to much, and automatically become more intense as the campaign progresses. I think the way I organize them is going to depend on the list I end up with.