## Thursday, February 12, 2015

### I think I'd like to see a boom bust survival mechanic in games.

It's not economics, it's population growth.

When you model population you're generally not going for perfection over the long term.  So you make comparatively simple assumptions that you know will break down.

The simplest assumption would be linear growth.  Get two points, make a line, congratulations you have a useless prediction.

Unfettered growth tends to depend upon how much you have.  If you have a lot, they breed a lot, and thus there is a lot of growth.  This leads to an exponential curve and is known as the Malthusian model.

Growth, however, tends to be fettered.  Any finite environment can only hold so much.  Maybe it's a question of living space, or waste disposal, or resources, but at some point the growth starts to slacken.

What we tend to find is a carrying capacity.  If you start below it the population rises to reach it, if you start below it the population falls to meet it.  Assuming you start below the carrying capacity, growth when the population is low looks fairly exponential, but then it hits and inflection point and the population slowly begins to level off.

This is called the logistic model.

But this model would imply that if you had one person eventually the population would grow to reach the carrying capacity.  Really?  And this person is going to produce new people how?  And there's not going to be detrimental effects due to a very shallow gene pool why?

Thus we come to the idea of a minimum viable population.  It turns out that changing a single sign in the equation of the logistic model will yield a pretty good short term minimum viable population model.

In the long term successful populations will go off exponential like to infinity which isn't realistic, as discussed.

This model is the boom or bust model.  If you start below the minimum viable population you decay away to nothing, above it and you shoot off to infinity.

So where does gaming come in?

Let me go of on another tangent.

-

The movie The Core is about a six person team sent to save the world.

Two of them make it out alive.

The first person to die was taken out by a random (unpredictable thus unavoidable) accident from an in universe perspective and because PLOT and CHARACTER DEVELOPMENT from the perspective of the succession of writers who made the movie, all of them doing their best to turn the piece of crap into something fun in spite of ridiculous restraints.

The second person died for no good reason whatsoever.  It was supposed to be a character defining moment where the person thrust into the command chair was forced to make the heartbreaking call to save everyone at the expense of one character, but it was so ineptly handled that it turned out to be killing that character basically for shits and giggles.  (In isolation the scene could have worked, but in context it utterly failed because context demonstrated two or three times over that everyone could in fact live if the character in the cockpit would just push the fucking button.)

This is where it gets interesting.

At this point there are four people and they come up with a plan.  One person needs to pilot the ship.  Two people need to move heavy objects, one person needs to do something that's going to be certain death if done by a single person.

Because there are only four people left, and three of them need to be alive and unharmed to complete the plan, the fourth person needs to die.  They can't, for example, break the task into stages so that it won't be deadly but instead just harmful.  It has to be suicide because there are only four people.  If there were five there's a good chance they could have made it through without anyone dying here.

Now there are three.

Once again: one person to pilot the ship, two people to move heavy things.

One of the heavy things ends up on top of the leg of one of the two people.  The pilot can't help, the thing is too heavy for one person to lift.  Thus that person has to be abandoned and therefore dies,

Since there were only three people the person died.  If there were four, that person would have lived.

So think about what that means.

If the second person hadn't died then possibly they could have saved the third and definitely they could have saved the fourth.

Saving that one character would change the survival rate from 33% to either 66% or 83%.

Save one person early on and the number of survivors at the end either doubles or more than doubles.  Hell, if you start counting after the first death, when they're suddenly a group without a leader, which is something games like to have a lot, saving that one person could potentially change things from, "60% died," to "Everybody Lives!"

-

So now games.  For a long time its been common in games to have a character that you have to escort.  Keeping them alive was necessary and, quite frankly, annoying.  It's becoming more common to have allied characters who can actually, you know, be helpful.  They can watch your back, bring their firepower, whatever.

But what I don't think I've seen is a mechanic based on the fact that if there are two of you trying to save someone, you're probably more likely to do it.

I've played a lot of games set in a city that's falling apart at the seams (zombie outbreak, Pandora's literal box has been opened, so forth) or in a situation where there are groups of survivors (Half-Life comes to mind as a foundational example) and yet the games always seem to run away from the idea that there's safety in numbers, that teamwork might work, or so forth.

Technical limitations are a big reason for this.  "Why does your leader always go off alone?" because it's hard to make realistic small group tactics.

But things have improved, a lot, and also even when you the player are off alone, the plot that's moving when you're not looking might want to take into account that if you left a group of several survivors who were mostly uninjured, they'll probably be better able to fight off [evil enemy of this game] than if you left one or two in a state of barely alive.

-

If you picture a FPS then you can see how with a population of one things tend to go toward extinction.  Everything is trying to kill you.  If it succeeds, bust.

If you save one person, unless that person can keep up with you then it's probably going to be hard to keep them alive.  Bring them into combat and they might die, leave them behind and you leave them unguarded.  You're still tending toward extinction.

If you can have four people that can break into two teams of two then no one ever needs to be left to fend for themselves and you can have the weaker party stay behind where it's slightly safer while the stronger party scouts ahead.  Maybe that's the break point.  Maybe not.

But you can see where the thinking leads, the more people who are able to survive long enough to join your group, the more people your group is likely able to save and absorb and support.  The fewer people in the group, the more tendency it'll have for members to be picked off one by one until you're on your own again.

Resident Evil 2 is what got me thinking about this.  There's a scene of death by cutscene in it early on.  I hate death by cutscene.  I don't mean that I hate cutscenes where people die, though I'm no great fan of them, I mean I hate it when someone dies in a cutscene because it is a cutscene.

In this particular scene your character gets to a spot where someone is trying to wave down a rescue helicopter.  Your character stops and does nothing.  Zombies approach the person from behind.  Your character continues to do nothing.  Zombies attack, your character does nothing.  The attack causes the victim to accidentally fire his weapon wildly as he is dying thus hitting the helicopter.  Your character does nothing.  The helicopter crashes.  Your character does nothing.

The cutscene ends, now your character can do stuff.

(I wouldn't have nearly as much of a problem if the doomed character died in a cutscene where you were running up the stairs and so there was nothing you could have done, but instead it's a scene where your character could have saved him, if not for authorial fiat.)

Resident Evil 2 is an old game, but death by cutscene exists to this day.  Tomb Raider (2013), a very modern game in a lot of ways, has a cutscene where you stand and do nothing in spite of being armed, having the element of surprise, and having clear shots at the unarmored parts of the murderers.  You just stand there while they advance on a hapless character and kill him.  You totally could have saved him if you'd had control over your character.

Anyway, Resident Evil 2.  Save that guy and you save the helicopter pilot too.  You also stop the helicopter crash which prevented you from accessing part of the station where someone is alive right now but will be dead by the time you finally manage to clear the wreckage and get to her.

In addition, if you had the person you saved on your side, you might be able to talk another person into coming with you instead of waiting in a place he erroneously thinks is safe and thus dying.

Save one person and you can in turn save at least two, maybe three, others.

The game as it is shows a downward spiral.  People die, their deaths set up even more people to die, and in the end there are only three known survivors (five if you count the two unknown survivors who were corporate spies.)  There could have been as many as 8 known ones + the two corporate spies and maybe you could even have managed to save one of the antagonists too (the other two antagonists were beyond all help for various reasons.)

There's nothing wrong with that.  Downward spiral is part of the boom or bust model I'm talking about.

But what about a corresponding upward spiral if you work your ass off to make things go well?

It should be fairly straightforward to adjust game mechanics so that extra people doesn't make things easy.  For one thing some of them would be the annoying as hell ones you have to escort and keep safe.  For another, in a game like Resident Evil getting any quantity of people in one place is probably like ringing the dinner bell for zombies, so even if you get more people with guns on your side, it could probably be leveled out by making that lead more enemies to you.

-

Short version: games have a long history of isolating the protagonist and showing a downward spiral toward (but, assuming you don't lose, stopping just short of) oblivion, but they don't seem to have the possibility of a corresponding upward tendency where, just as deaths lead to more deaths, keeping people alive leads to keeping more people alive.

I'd like it if they did.  Well, if at least some did.

Maybe you could measure success not by, "I made it out alive," but by how many people you managed to bring out (likewise alive, no need to drag dead bodies) with you.

1. I remember a dragon-themed vertical-scrolling shooter on the NES where, if you beat the first level, you would play the rest of the game in hard mode, and if you didn't, you'd play the rest of the game in easy mode. You could do a similar thing here - not hard or easy mode, but one set of challenges versus a completely different set of challenges. The boom-or-bust mechanic would make that happen very naturally very quickly, too:

If you fall on the bust end of the spectrum: You are trying to survive on your own, or with very few other survivors. No-one will be awake if a zombie breaks in in the middle of the night. Your only chance of survival is to figure out how to get away clean and not be detected before morning. On the bright side, that means that the zombie necromancer has no idea you exist, which means if you're smart and you're fast, you can take her down by surprise in an ambush and end this before it spreads.

If you fall on the boom end of the spectrum: You have dozens of survivors whose lives are dependent on you. It's not that they're helpless (although some of them are - and what are you going to do with the people who have been bit, anyway? It's not as if you know they'll turn, but everyone is scared), but when everything went to hell you were the badass at the center of it all that kept these people alive, and by panic logic that makes you boss. You need a base big enough to give everyone a place to collapse from exhaustion, you need a base small enough to guard from assault, you need enough food and water to keep everyone alive, and, preferably sooner rather than later, you need to figure out a plan to stop the necromancer before she manages to overwhelm you all.

If you start off on the bust end, you might run into other survivors, but there's not much you can do for them - it's not as if you have a safehouse they can stay at.

If you start off at the boom end, you can't afford to be hemorrhaging survivors - it's not as if the necromancer will forget you exist.

I would want to make the boom ending harder to get, but they'd both be hard, even if it is for completely different reasons.

2. There's a technical constraint, of course - even now games aren't very good at dealing with multiple people who are supposed to have human-like levels of mental complexity. The Lone Wolf gamebooks were notorious for this, because they told an ongoing story - after a few books you'd gained some reputation, so you'd get an escort of soldiers when going off to do something dangerous, but they'd all get killed or at the very least you'd be separated from them so that you could go back to single-protagonist adventure.

I like Packbat's suggestion and I'll offer a closely allied one: you can face the same overall "plot" problems, but you'll also have problems that change based on the size of your group (a big group has more chance of including someone with the key skill and can defend itself better, but it moves more slowly and needs more food and shelter). That wouldn't even need to be separate endings, but rather separate approaches to the same problem.