Post-Con Debrief - The Survivalist

So here's a story about core gameplay, and why it's important to identify it early in the development process.

Just this weekend past I had the exceptional pleasure to attend Arcanacon in Melbourne.  I debuted my new game, The Survivalist, which took 34 players over seven sessions.  It was my seventh major convention game, and was a serious mechanics-driven game about survival on an alien planet.

The game was solid, and well-enjoyed, but it wasn't a showstopper, and it's clear to me that it wasn't the best kind of game I can run.  So why not?  What happened?

What follows is a debrief.  I'm going to talk about how the game operated in practice, what my intentions were, what worked, and what didn't.  And my ultimate conclusion - which is that I misidentified the core gameplay.

So: a warning.  This post, by necessity, contains massive spoilers for The Survivalist, including all the key secrets of all the characters.  If you haven't played the game yet, you're going to have trouble ever playing it in future if you read this post.  That's probably okay - this one's going in the closet for 12 months or so, so I'm not running it anytime soon - but it seemed only fair to flag it now.

How does the game work?

The Survivalist was born from an old teaching exercise that I got to play multiple times over the course of my education.  You're stuck on a deserted island, but you've salvaged ten things from the ship before it sank.  What ten items did you take, and why?

That's where it begins, but ultimately there's a lot more to it than that.  If you've played the game before, you can skip over the run-down below, but you might want to read it anyway to see what remains consistent across every session.
The meat of the game is this: five characters are stranded on an uninhabited planet.  They are low on food, low on water, and they have to survive a three day trek to an abandoned survey outpost to call for help.

* The characters

Halifax is the captain of the the Lady of Titan, a poorly-maintained starship whose crash precipitates the game events.  He is, secretly, a drug addict and a fraud, who knows nothing about piloting and who crashed the ship while high.  He is strong, fit, and (depending on the player) possibly quite charming, but he will also need to consume drugs or alcohol each day of game-time or freak out.  His goals are to conceal his role in the crash (including both finding and then disposing of the ship's flight recorder) and to convince the other characters that he is the strong, manly hero that he makes himself out to be.  His unique ability is that he is "brash" - his actions cannot be interrupted until complete.

Angus is a good and decent man whose wife was murdered 10 years ago.  Amongst the wreckage of the ship is a unique antique swordcane identical to the one wielded by the killer, so he feels certain one of the survivors is the murderer he has been looking for.  His goals are to find the killer, bring them to justice, and get the remaining survivors off the planet alive.  His unique ability is "teamwork" - while he has the support of other characters, he can easily accomplish difficult tasks and wins all conflicts.

Jocelyn is a thief and a con-artist who masquerades as a woman of high society.  She is insecure and difficult, and predisposed to manipulate and steal from the other characters.  Her goals require her to escape the planet with as many valuable goods as possible, while simultaneously convincing the other characters she is useful and necessary.  She refuses to do anything she is commanded to do, but her ability is "gutsy" - she can always succeed at anything she is told she can't do.

Paul-Henri is the killer that Angus is looking for.  He is an elderly Frenchman with a long string of unsolved murders to his name.  He was under police escort but his guard was killed in the crash and he has broken his own wrist to escape from his handcuffs.  Physically crippled by his age and broken wrist, and requiring painkillers each in-game day, Paul-Henri nevertheless starts with a very good read on the other characters and a penchant for being polite and likeable.  His ability is "killer" - once per character or animal, while in melee range, he may declare he is killing them and instantly do so.

Soo-Jin is a loner, and is possibly the "survivalist" of the game's title.  She is a young Korean woman working for an interplanetary drug cartel.  She doesn't work well with others and can't stand to be physically touched, but realises she needs the others to escape the planet.  Her goal is to survive without incurring debts to others, and to escape the planet with a soft toy rabbit which secretly holds the instructions for a valuable new drug preparation.  Her ability is "hunter" - twice per day she can near-magically prepare a meal or a drink, representing her abiilty to hunt and forage.

* The mechanics

Most of the conflict resolution in the game is by GM fiat, taking into account the abilities, competencies, equipment and relative health of the characters.

The main mechanics centre around survival and inventory management.  All relevant items in the game are represented by cards.  All items are public knowledge and are placed where all players can see them.  All items have a "size" rating, representing their bulkiness and the difficulty of carrying them.  Characters each have a "carrying capacity", which differs from character to character.  Characters can only carry items whose total size is equal to or less than their carrying capacity.

Some items are meals or drinks.  Characters are told AFTER they have finished foraging the crash site that they will need one meal and three drinks every day of game time.  If they come up short, they will lose carrying capacity, representing weakness due to hunger and thirst.  If they come up short two days in a row, they will die.

"Death" is not always final.  Characters begin the game with a "bonus life".  The first time they die, the party assumes they are dead and moves on, or otherwise the character's corpse falls down a cliff or into some water or otherwise is taken away from the other players.  The character will then wake up, and can rejoin the party or follow at a safe distance as they please.  If they die again, they are out of the game.  The game is structured such that it is unlikely that anyone will be completely eliminated from the game prior to the endgame.

Each character has a liability (e.g. Halifax's need for drugs) and a failure to live up to the requirements of that liability results in them taking a panic card, representing something stupid and destructive they will do as a result of pressure, either immediately or in the near future.  Examples are a requirement to immediately attack another character, throw away a useful item, or disastrously fail their next important task.

There are some mechanics around "joint carrying" and created travois-like structures that are unimportant for the purposes of this write-up.

* Introduction

The game opens with the crash of the Lady of Titan.  The various characters wake from their cryogenic sleeping tubes (or Halifax, unmentioned to the players, from his drug-induced stupor) to find the ship plowing through the foliage of an alien jungle before eventually smashing into the surface. 

They start the game in various states.  Halifax is trapped in his captain's chair by its harness, and requiring help.  Angus is trapped in his cryo tube, and requiring help.  Soo-Jin is ensnared in the ship's internal wiring, and needs help.  Jocely and Paul-Henri are free and able to help, although Paul-Henri's wrist is broken.

This scene got added after earlier playtests, which basically started with the "foraging" scene I describe next.  It allows a good chance for everyone to describe themselves and how they react to a crisis.  Trapping Halifax and Angus initially balances out their natural tendencies to become the leader and dominate the early discussion.  It gives Paul-Henri and Jocelyn a chance to have people feel initially kindly towards them, to balance out their later inclination to look suspicious or downright threatening.  In theory this gives us a chance to establish Jocelyn's "she won't do what she's told" liability, because if anyone says "help me!" she'll be required to, in fact, not help them, but in practice this never seemed to come up, so clearly that didn't work.  It also means that Soo-Jin is likely to end up receiving help from other characters and therefore starting with an initial debt she has to repay, to balance off her tendency towards being an unhelpful loner.

* Foraging for items

What happens next is probably the defining scene of the game.  Wind clears away the smoke around the crash site, and it becomes clear that the cargo and other items of the ship are scattered everywhere around the crash site.  Halifax is worried the ship's engines will explode shortly, so the group has limited time to scavenge through the wreckage. 

A deck of about 80 item cards are placed in the centre of the table and scattered around.  These represent everything useful that has survived the crash of the ship.  It includes a couple of dead bodies, representing Paul-Henri's guard plus another unlucky individual.  Items range from the useful
(survival rations, hand axe, bottled water) to the amazingly useless (shampoo, a bible).  Players are given exactly six minutes of real time to scavenge through the items and take what they want before they are forced to run for cover before the ship explodes.

Before this task, they are given some strict rules: they are to maintain a strict separation between what they are carrying and the pile generally.  They cannot touch more than one card at a time.  They may take cards from other players, but not while they are being touched.  At all times the cards in front of them must be equal to or lower than their carrying capacity.  They may not deliberately hide cards from other players.  They may eat or drink food and water they find immediately, if they choose, by turning the card face down and putting it near their carried items.  They may use items at the crash site while still at the crash site, but the timer won't stop and long actions may involve them having to take their hands off the table for a while.

Once the six minutes are up, the cards not being carried are removed.  The GM goes around the table quickly and highlights what items players have taken, and reads out the clarifying text on cards that are not fully explained by their names.  He then gives players the bad news about how much food and water they're going to need. 

I find this one of the most fun parts of the game.  It's hectic, it rewards skill, thought and teamwork in realistic ways, it produces meaningfully different games based on the combination of items taken, and it's unique and (to the players) unexpected.  It's definitely a successful part of The Survivalist.

Some of the other key items in the foraging include the antique swordcane that identifies Paul-Henri as the killer, the soft toy rabbit that Soo-Jin needs, and a pair of guns plus their associated ammo.

* Timing note (1 hour)

At this point in the game, I'm generally at the end of the first of the three hours.  All Arcanacon, the final player would routinely arrive 15 minutes late, and then there's introducing the game and its assumptions, character selection, explainaing the rules, the intro crash scene, characters introducing themselves, and the foraging.  It continually freaked me out that it took an hour to get to "the start of the game" but everything in that hour was both necessary and held strong interest from the players (except the lateness, obviously) so I'm disinclined to cut anything from that hour even if I could identify an easy target.

* Day One: The Jungle

One of the items in the crash site is a survey map, which describes a three-day march to an abandoned survey outpost from which the characters can presumably call for help.  The survey map is super-important, so I always highlight it and explain its importance during the foraging scene.  The map shows that the first day of this journey will be through jungle, the second across "a flat area of some kind", and the third will be a vertical climb up a cliff.

The jungle contains several threats.  The players will have a hard time progressing through the woods, and will benefit from having a hand axe or another solution for clearing undergrowth.  Later, they will be bothered by stinging insects, which can be solved in a variety of ways, including using the various insect-repelling items from the crash site, or discovering that the local foliage has insect-repellant sap and can be burned to create insect-repellant smoke.  Finally, they will be attacked by the cat-like predators that live in the jungle, an event which can reveal Paul-Henri's competence at killing. 

The animals can be driven away by any creative solution, including scaring them by firing one of the firearms, firing the flare gun, using the emergency klaxon, settting a large fire, or a reasonably decent defence that involves Angus' "teamwork" power.

In some games, I mentioned to players that the map showed a possible stream of water in the forest, half a day's walk out of their way.  When this was mentioned, players uniformly chose to visit it.  It was full of undrinkable alkali rather than water, although many groups found ways to distill, purify, or otherwise use the water anyway.  I did not do this in all games, and eventually came to the conclusion that it was a mistake to mention it.  Later games did not allow players to visit the stream.

* Night

Each night the players are required to offer up their consumed food and drinks.  They may then engage in a round of asking questions of the other characters.  Finally, they have a choice between camping the night and pressing on through the night.  Walking through the night is inefficient and incurs an exhaustion penalty to their carrying capacity, but does move them faster towards their goal.  If they camp the night, they may choose to keep watch in "shifts".  Shifts give each character the chance to mess with or steal from characters who are asleep.  I didn't have any challenge planned for if they didn't keep shifts - if they all just slept all night, we moved straight on to morning.

* Day Two: The Salt Flat

The "large flat area" turns out to be an alkali flat, barren and devoid of vegetation.  Halfway across the flat is a lake of caustic alkali.  Fish can be seen deep in the lake, never coming closer than five metres to the surface.  Large flat icebergs float in the lake, brown in colour, and numerous enough that a person could feasiblhy walk directly across the lake by jumping from iceberg to iceberg.  Detouring around the lake will add half a day to the journey.

By now, players who didn't manage to procure water in the jungle are probably critically short of resources and facing death.  Many groups try to procure resources here.  As with the stream, a sufficiently inventive group can distill water from the alkali.  The fish can be caught and eaten f players can think of a way to reach them.  In early games the icebergs yielded drinkable water.

In later games this situation was made less generous.  The water from the icebergs was made "poison water", which sated thirst (and prevented death) but permanently reduced carrying capacity by one for each litre consumed.  The fish became skittish and could only be caught in very limited quantities (typically only one, or two or three if abilities were used or clever solutions concocted). 

Crossing the icebergs is in reality mostly harmless, and is an opportunity to congratulate players on their good decisions.  I highlight Jocelyn if she has managed to find good boots to replace the high heels she starts in (most do - there is a pair at the crash site).  I highlight if the group is working together to cross the icebergs.  I take note of uses of Angus' teamwork power, or Jocelyn's "gutsy" power.  The only time someone falls in is if they have recently consumed a lot of alcohol or the hallucinogenic drugs (this is usually Halifax) and they can be easily saved.  Or if they have a relevant panic card, they can fall in as a result of that as well.

At the end of day 2 is another night segment, where they again must eat and drink, and may ask questions of each other.

* Timing note (2 hours 15 minutes)

We're generally at about 2 hours and 15 minutes by the time they reach the base of the cliff.  45 minutes left to wrap things up.

* Endgame: The Climb

The last few times I ran the endgame at Arcanacon, I started it by informing everyone they no longer had a "second life".  I did this immediately before decisions about shifts et cetera but after they'd eaten and drunk.  With this change, the game finished neatly in 3 hours, but it was a bit unsatisfying to be the first person to be killed.  Without the change, I think the ending was more enjoyable for all players, but the game would routinely run 10 to 20 minutes long, which is unacceptable at a convention.

The mechanic for the climb is this: anyone who is at less than full health is unable to make the climb alone.  This includes Paul-Henri's with his broken wrist, plus anyone who has lost carrying capacity.  In later games I started telling people about this mechanic immediately after leaving the crash site, when they were surveying the map, which I think is probably better, but no-one seemed too cheated to only hear about it when they got to the cliff. 

If you cannot make the climb alone, you need a buddy.  Two wounded people can buddy for each other; that's fine.  A buddy group can be more than two people and it can in fact include all five characters.  During the climb, the climbers will have to trust and rely on each other.  Anyone in a buddy group will have the chance to kill other members of their group.  During the climb I count "one, two, three".  On three, anyone may raise their hand.  Anyone who raises their hand can kill one other member of their group.  If ONLY one person raises their hand, they may kill ALL of their group. 

If no one kills anyone, and I feel pacing requires it, I will then have the weakest (lowest carrying capacity) survivor fall from the rock, and give everyone a count-of-three chance to save them.  This usually provides new information, where some choose to help and some don't.  I then do a further count to give people the chance to kill someone.

After that, the remaining survivors make it to the top of the cliff, where the survey outpost has enough supplies to last everyone until they are rescued.  A last round is called in case people want to do something that means that the remaining survivors do not just escape the planet with the items they are currently carrying.

* Debrief

A final debrief is held, where players reveal their secrets, score their goals, and (if required by circumstances) provide their characters with an epilogue describing what happens after they leave the planet.

* Timing note (3 hours and finished)

If I've run the shorter endgame (no second lives in the endgame) I'm now at 2 hours and 55 minutes, right on schedule, and the game is over.

Intended gameplay

So that's The Survivalist.  The gameplay was built to try and bring out several key tensions:

* The individual versus the group - The best interests of the group are for the group's survival to be maximised by the group collectively bringing only items that are necessary for survival, and as many of them as possible.  However, for most characters their individual interests are best served by taking at least one useless or selfish item, and convincing the others to carry the necessary things.  Likewise, they will need at least some of the group to successfully navigate the final challenge (the cliff) but in most cases their personal interests are best served by as few other characters escaping as possible.

* The immediate versus the long term - People have a hard time accurately assessing their long term needs.  Players will likely take an insufficient amount of food and water at the crash site.  Some tests suggest that they do this even if they know how much food and water they are going to need.  They also have a hard time balancing their immediate and high-priority need to survive against their long-term wish to come out with as many points as possible.

* Strengths and liabilities - While players were perfectly happy to kill characters off for personal reasons, goals, or vendettas, groups were very reluctant to cut out people merely because they were a liability, even after they became obviously violent or unreliable.  I'm not sure what that says about people or my game, but it's clearly something that players thought about and engaged with while playing.  Even a clear warning at the beginning of the game that player-versus-player violence was playtested and encouraged still had people hauling full groups of five all the way to the cliff despite plainly not having enough food and water for everyone.

I had intended much of the gameplay to be discussions about who can be trusted, who is useful, and what items should be carried and by who.  That's more or less what I got.  I also got a lot of discussion about survival methods and techniques for extracting water from things, which to some extent was expected (given the setting) but turned out to consistently be more exhaustive and less fun than I had hoped.

Evolution of the game

The first playtest suffered from some issues but was fun enough that I booked myself to run The Survivalist at Arcanacon based on its strength.  Players seemed to mostly have fun with what, at that stage, were some very rough mechanics.  Most of the issues came from (a) players trying to bend the ruleset in ways I hadn't expected to maximise their survival chances and (b) players finding some of the mechanics and statistics so unrealistic as to break their suspension of disbelief.  As a result of the first playtest, I significantly overhauled the core rules, added new items, removed others, changed the sizes on a range of items, and removed a lot of the less plausible environmental challenges from the game.

The second playtest was fairly disastrous.  Play became bogged down in logistics, with far too much time spent handing cards back and forth.  At this point I was asking for food and meals to be handed over three separate times each day of game time, and there ended up being a terrifying amount of book-keeping.  Players were also confused by Angus' ability, which was quite different at that stage and allowed him to reduce the size of items jointly carried between two people.  It was an abstraction that was often hard to explain, was subject to gross exploitation, and was administratively difficult to manage.  They were also frustrated by Jocelyn's ability, which at that stage enabled her to steal the items of other characters without those characters noticing.  The players felt there was far too much book-keeping and not enough roleplaying.

As a result of the second playtest, I changed the abilities of Angus and Jocelyn to their current form.  I added a range of new "created item" cards for items found or created after the foraging scene, most notably including a "travois", which would be handed out if players tried to use items (for example the tarpaulin) as a sledge or sled.  The travois has size zero as long as it is being joint-carried by two people, and itself has a carrying capacity of 8.  It represents a response to the scenario where people try to build a sled, rewarding them for their idea without breaking the game.

The subject of roleplaying was trickier, and I still haven't solved it today.  The point of the game was to be a mechanical exercise; if I was to discard the inventory mechanics, I may as well just run a different game.  But, being mechanics, and particularly inventory mechanics, they eat time, and they emphasise puzzle solving and object-oriented gameplay over characterisation.

I reached a compromise.  I condensed the resource management down to only once per game-day, thereby cutting a lot of logistical time and picking up the pace.  The majority of that time was reinvested in a the new "introduction" scene where the players experience the crash and then help each other out of it.  This turned out to be a huge improvement to the game.  Unfortunately the rest of the game still didn't have the time budget to allow casual banter or detailed discussion of every obstacle.  I added a new "questioning" scene to each in-game night, to at least ensure that players who needed to learn things about other characters had a vehicle to do so.  As it is, the game's still very light on characterisation and dialogue compared to basically anything else that I run.  You could fix that by adding about 30 to 40 minutes to the game time, I'm reasonably sure, and then both encouraging and promoting in-character discussions and characterisation, but that's not a helpful answer in view of a three-hour time limit at conventions.

The third playtest was very successful, from the perspective that the game ran very close to time and the players had a lot of fun with it.  The logistical issues were certainly much more under control and the players were much more inclined to directly engage with the themes and core gameplay. 

During the third playtest I added the time limit to the foraging scene.  Previously players had had unlimited time to loot the crash site, which led to a perhaps more reasoned and emotive discussion about what to take, but also went on too long and put too much pressure on players to only take useful items.  At the third playtest I tried a ten-minute limit on searching.  The players felt that even 10 minutes was too long, so I've subsequently run it as only six minutes, and that seems about right.

At Arcanacon, I ran seven sessions of The Survivalist and made two significant changes over the course of the convention.  The first was to remove the stream from the game.  In early games, players had been able to detour to reach a stream of alkaline liquid in the jungle.  This took up time, it reduced the resource pressure, and it was ultimately a puzzle rather than something that directly addressed the themes of the game.  In the last two or three sessions, I stopped mentioning a stream, and didn't allow players to look for it.

The other change was to the endgame.  The game had been running long at Arc by about 10 to 20 minutes, which was not okay in a convention context, and I was desperate to get it to run within its allotted timeslot.  So for the back half of the convention I started taking away the "second life" mechanic once people reached the cliff. 

It immediately fixed the time problem - in the original version, the endgame turned into a long procession of jump scares as character after character presumed dead returned to wreak havoc - but I could tell that being the first character eliminated really sucked.  It was unsatisfying and usually came without real warning.  I don't like the new ending much, but a game that doesn't run reliably within its timeslot is not an acceptable game.  If I brought it back to a convention again, I'd like to find another way to cut the time and preserve that original endgame.

What worked

I don't want to understate this: The Survivalist was a complete game, with a beginning, middle and end.  It had clear, achievable goals, it had relatable characters, and the players always knew what they were doing and where they were going.  That's sadly not as common at conventions as it should be, and so there's good reason to be satisfied in clearing even that very low bar.

Secondly, I think players had fun.  For various reasons - mostly psychological health on a very tiring weekend - I didn't invite criticism directly after sessions, but my feeling was that while some players were frustrated by parts of the game, overall few if any regretted playing it, and most had a good time with the material despite the rough spots.  That's really all a convention game needs to accomplish to be worthwhile, so I'm okay with that.

So: let's talk more specifically.

* I think players enjoyed the introduction and set up.  It's polished, it's confident, it's cinematic, and even though it's mostly linear it's still enjoyable and dramatic content.

* The foraging scene at the crash site was a sucess too.  Players' eyes tended to light up as I explained what was going to happen, and the allotted time and rules appeared to be both fair and exciting.  This was really the core idea that the game was built around, so I'm happy that this worked out okay.

* Players liked the character sheets.  I tended to get a lot of chuckles from players reading them, which is a good sign, and only a few players looked stumped or stressed by what they'd just read.  The sheets were two pages, but even the late night/early morning groups didn't seem to have trouble absorbing the content as they read, so I guess that worked out.  (There were signs that key things had been overlooked, though - more on that later.)  Halifax's sheet appeared to be a particular highlight; everyone loved it and took a lot of relish from summarising it at the debrief.

* The original endgame (with second lives still intact) seemed to also be a success.  It was a bit silly with everyone endlessly popping back up for their second chances, which a lot of players openly recognised, but I think despite that everyone really enjoyed that aspect of the game.  I was very sorry to cut that for time, and am leaning more and more in the direction of thinking I've made the wrong choice about that.

* The trust mechanic for the climbing seemed to work.  It's simple and reductive but everyone seemed appreciative when they heard about it, and it appeared to produce real tension. 

* Lastly, the suspension of disbelief held up.  I'm not totally clear on what I did to influence this, but I didn't get too many complaints of things being unrealistic, or groans about cliches, and people appeared to buy what I was selling all the way through to the end.  It may be that it's because of the way I emphasise the degree of abstraction in the rules up front, or it may be that I put in just enough believability that they trust me for the rest, but either way this could have been a real issue with this game and it wasn't.

What didn't work

The main issue with the game is days one and two - the jungle and the salt flat.  There's a lot of sub-issues with them - pacing and resource management being the two biggest - but the key one is that they ultimately just don't engage with the thematic core of the game.  They're solvable puzzles when they should be meaningful decisions.  I'm only realising it now but what I should have done was thrown out all of this nonsense and replaced it with a series of straight up ethical challenges.

For instance, there are animals you can eat but they take forever to die and they scream like a baby in obviously terrifying levels of pain.  There's a fruit you can get but it's next to a hive of insects and whoever retrieves it is going to get badly stung.  The alkali flat rips bare skin to flesh but you only have four sets of boots between five people. 

I wish I'd realised it sooner; that game - the dilemma game - is a much better game.  I want to run that game now.  I think these kinds of decisions were my original intention but I placed far too much trust in the rules and mechanics to generate them organically, when I should have directly interposed to create them.  In other words, I misidentified the core content, and thought the fun was in the process when it was actually in the underlying concept.

A subsidiary issue was that the resource management was too easy.  Players were too able to find the food and drink they were missing.  Not only did this defeat the bleak choices I wanted them to make, but the process of them solving the puzzle was time-consuming and dull for everyone involved.  It wasn't fun gameplay AND it made the other gameplay worse.  In future - and I did this in that last two sessions, to some extent - I'd remove the possibilities of finding more food and water.  The characters SHOULD deyhdrate to death.  They SHOULD realise they don't have enough food and water for five people, but they do for, say, three.

Characterisation and roleplaying was an issue.  This was a mechanical game, so I didn't intend it was going to be deeply moving and filled with anecdote and metaphor.  But I would have liked more than there was.  Sadly, moving players through the basic beats of the plot wholly occupied the time.  I think cutting all those solvable puzzles and replacing them with the dilemmas mentioned above would go a long way to fixing this.

The revised endgame definitely was not as good.  I need to find something else to cut so I can go back to the original ending.

Players occasionally missed things on their character sheets, which suggests they need a rewrite.  Soo-Jin's backstory appears to be too complicated as players had trouble summarising it at the debrief sometimes.  Paul-Henri regularly missed the note telling him not to immediately claim his sword-cane (which makes it too easy and fast for Angus to tell he's the killer). 

The abilities and liabilities appeared problematic.  Everyone got a lot of use out of their ability, except Jocelyn.  Angus' needs a rewrite to further clarify it.  Jocelyn often found both her ability and liability going to waste.  Honestly, nobody ever flat out told Jocelyn to do things, even when the Jocelyn player went out of their way to invite it.  I thought it would be overpowered and overused but it ended up being dull because nobody would ever trigger it.  I can probably fix it by changing the description of Jocelyn on, say, Halifax and Soo-Jin's character sheets to add something like, "She clearly needs someone to authoritatively tell her what to do, and that someone should be you."   Soo-Jin's liability (not wanting to be touched) also seemed underwhelming in action.  It might be the dilemma-style gameplay would fix this but possibly she just needs something else entirely.

Final thoughts

The basic problem with The Survivalist is this: I incorrectly identified the core gameplay.

This is absolutely critical to any game.  Any game.  If you don't understand what the heart of your game is, which bit of it gets underneath people's skin, then your game is incomplete.  D&D is a game about power creep and loot collection.  World of Darkness is a game about self-expression through differentiation.  Diablo is a game about poker machines.

The core gameplay in The Survivalist is "difficult decisions, under pressure".  There should be no right answers.  There should be no successes and failures.  There should only be decisions, and consequences. 

The genuinely fun initial scrummage for items misled me into thinking that this was a game about resource management, with the hard decisions rising organically from the game mechanics.  But that's wrong.  The fun in the foraging at the crash is not in WHAT you chose, it's in what YOU chose.  It's self-expression by way of survivalism.

And that's where the rest of the game should have gone.  I made the challenges solvable puzzles when instead they should have been mirrors held up to the players (and their characters).  They should have been zero-sum games, where you always lose something and maybe gain something, and really what mattered is what your decision said about you.

The ending worked largely by coming back to these core themes, and once again talking explicitly about trust and dependency.  But if I was to take the time to improve The Survivalist again - and there's a good chance I won't, as I have new games to write that take precedence over fixing an old one - I'd completely rip out the jungle and the salt flat, and write new content throughout, less mechanically driven and more centred around dilemmas of character.

What do you think?

So that's what I'm thinking.  I'd love to hear your thoughts, whether you've played the game or just read the write-up above (but let me know which category you fall into!). 

And finally, thank you so much to all the wonderful players I had at Arcanacon, who engaged with my game enthusiastically, genuinely, and with an open mind.  Every one of you was a privilege to run for, and I hope you'll see me again at Conquest or Phenomenon this year, where I'll be bring a very different game to this....