The Castle Doctrine Forums

Yeah, a lottery or a robbing contest is a great idea to add money to the game.

For the lottery, to avoid sock puppets, there would have to be an unlimited number of entries per account, with a unique, human-only puzzle each time.

There has to be SOME human-only puzzle that could be invented here.  Obviously a visual one.

I think the hard part with CAPTCHA is that we want something machine-generated that is unsolvable by a machine.

Well, for respawned players, there was never a punishment for robbery failure, aside from chill timers, because they have nothing to lose.

Well, anything that assumes some real players (rich ones!) pay other players (poor, jobbing painters) fails in the face of inductive reasoning.  How did the rich players get rich to start with?

There has to be a source of money outside of getting it from other players.

In TCD, that source was the starting budgets given to respawning players.  But, if you control two accounts, you get 2x as much starting money.

I'm not convinced that you need a non-player source of money that ALL respawning players can take advantage of.  You just need one that SOME can take advantage of to get the thing off the ground, and keep it off the ground later.

If robbing costs money, then yeah, every respawning player will need access to some "grind money" to get started.

But the "ante" that you've now attached to robbing is supposed to prevent endless robber-spamming, given that death has no other cost.  Chill timers and such can be bypassed with multiple accounts.

What if there was a system-wide chill timer?  After someone dies in a house, it goes "off the market" for X minutes for EVERYONE but the owner.  Multiple accounts have no advantage, now.

Maybe there is a way you can "buy" your way into one of these off-market houses.  A gas mask that allows you to stand the stench of the dead body.  And also, the owner can instantly take the chill-timer off by revisiting the house.

Thus, the chill timer only affects the poorest players, but prevents them as a group from robber-spamming a house 1000 times in one day.

SO, if you really want to "work on" a house by rob-dying over and over in it, you'll need to buy a lot of gas masks to stand the stench of your own body.  Morbid, I know.  There's probably other fiction (police line that you have to pay to cross, etc.)

So, that frees up poor players to rob EVERYONE, and then there's no reason to given them a house budget at all.  Well, maybe a small house budge AFTER they have some money to protect.  So, we don't have to worry about poor players "grinding" money on every respawn, because they can steal it from others with no investment.

Still, we need a way for SOME money to get into the game in the beginning and from time to time later, as needed.

I hadn't thought about the need for it to be robot-proof.  I agree that a map algorithm that isn't amenable to bot-robbers sounds like a hard problem.  Though maybe it's not as hard as we think it is... I wonder if there's some class of maps for which algorithms would have trouble, but human players would easily see the solution.  Probably not, because maps are necessarily discrete.  Captchas are always leveraging problems from more continuous-seeming domains.  I mean, we could also tell the difference between a human and a robot by throwing a ball to the subject and seeing who could catch it.  I can't imagine any house designs that are simple for humans but not amenable to brute-force search by a computer.

Other player's discarded houses are easily solved by the past owners of those houses.  For a similar reason, we can't let any players be involved in designing the seed houses, because then they themselves have an advantage.

I wonder about paying players to do useful, valuable work for the game.  The dummy-robbers is a version of this, because we're paying players that design decent houses.  Of course, the problem is that this can be automated by finding one good house design and reposting it over and over on different accounts.

We also can't reward them via money creation for creating a house that kills real players, because multiple accounts again will abuse it.

You were essentially rewarding "real" players with gifts based on the activity you observed from them.  Oh, you built a nice house!  Here's some money for you.

The only idea I can come up with is if we let other players decide as a group who should get money.  Like a "beautiful foyer" contest where the whole neigborhood gets to vote on who wins.  If you make the numberde of required votes high enough, you prevent multiple accounts from being leveraged.

You can also give more weight to the votes coming from houses that got more votes themselves.  A pyramid scheme of influence, tracking who the legitimate players are and rewarding them, while reducing the impact of votes from dummy accounts.

Anyway, this would let other players do the job of figuring out who was "doing valuable work for the game."  Though it does suffer from tragedy of the commons (nobody votes) and also leader bashing (revenge voting).  So, a player who is clearly doing great work for the game might not get rewarded according to the whim of the masses.

What other work could players do that would help the game AND be worth giving them money for?  The problem is that we need a way to judge the quality of the work done to prevent people employing bots to do it.  Or discover some work that bots simply can't do.

Well, the fact that you need a "temporary hack" like gifting "to get it up and running" is telling, and why I'm suggesting that thinking inductively is important.

The financial system in the original TCD didn't work to prevent sockpuppets, but it did "work" as a functioning economy for any sized player pool, all the way down to N=1 (you can build a worth-robbing house with your starting money and then wait for another player).  There was no hand-wavy "well, more players will somehow fix it" argument.

The reason why what you're describing doesn't work for N=2 is because by "gifting" or "mercy robbing" the other player, Zed isn't really playing the game.  Zed has no rational reason to rob the other player, because Zed knows for sure they have nothing worth robbing.

Once it "gets going," with many players, that problem will persist.  When the N+1 player joins, people will have no rational reason to rob them, so that player will be frozen out of the economy.  And, assuming it ever did "get going" through some hack or trick, over time, the wealth would tend to snowball into fewer and fewer accounts, gradually enlarging the pool of players who aren't worth robbing, and thus the pool of players who are frozen out of the economy grows.

The players with money will get to know which other players have money pretty quickly, just like Zed knows this in the case of N=2.

Like, with N=4, how do I know right now that I have all the money?  Because I stole it all!  I'm not going to try robbing anyone now.  Any new house that pops up has no money in it.

Well, I think you should be thinking inductively here.  If it doesn't work for N=2, it's not going to work for larger N either.

I get why you need to risk something to rob, to avoid robber spamming.  And chill timers can be bypassed with sockpuppet accounts.

But if you're "gifting" people money in the game, then sockpuppet accounts become valuable again!  Someone with extra accounts gets more gifts than someone with only one account.

What about if you only need a backpack if you're carrying stuff in, or carrying items out?  So you can take the money from the vault, but not the stuff, if you have no backpack?

You still want to prevent robber spamming, of course.

So you have to spend a valuable resource to rob, and there can't be a way for you to create that resource directly, or else you can create an infinite amount of it through dummy accounts.  There also can be no automated (or manual!) way of doling out this scarce resource, because then people with multiple accounts will be more likely to get more of it.

I still feel like this leads to the conclusion that:  "The only way to get 'new' money is to steal it from someone else," where we have to guarantee that "someone else" isn't actually you, yourself.  And the only way we can guarantee that is if "someone else" is Zed.  Then only Zed can cheat, but we trust him not to.

Or if Santa Zed dies while trying to rob your house.  But that feels much less satisfying, because it involves sitting and waiting.  And Santa Zed is more likely to pick your house to die in if you have multiple accounts.

What if you only need a backpack to rob player houses, and not Zed's house?  So robber spam is permitted Chez Zed?

What is another way for money to flow into the economy that doesn't give a benefit to multiple accounts?

A cash machine that gives out a certain amount per hour, and you just walk up to it and get it?  That obviously rewards refresh-spamming and is no fun, but multiple accounts won't help.

Remind me why players need to buy a backpack to go robbing?

Why can't they rob a house "bareback" with nothing at all?

What does the $200 backpack requirement have to do with the conservation of value or sockpuppets?

You also have created a system in which the vast majority of players create $2000 houses with nothing to steal inside of them!  But of course there's no way to tell this from the outside.  So even after you get enough money to start robbing, most of your robberies produce $0 on success.

And if you place the burden on the admin to "gift" people starting money to kick the whole thing off, then the admin needs to detect the difference between real players and sock puppets.  Otherwise, double account players get twice the gift money as single account players.  Putting you back to square one.

Let's start from first principles again:

For the game to make sense, robbing a listed house has to be worthwhile and produce some benefit.

If robbing a house produces some benefit, than any account that figures out how to rob the house will reap that benefit.

An account that is colluding with the owner-account of the house in question has an advantage in that they can ask the owner for advice on how to rob the house.  Therefore, they can unfairly reap the benefit of robbing the house before non-colluding players figure it out and reap the benefit.

Your way of fixing this problem is to attach a benefit ONLY to one house in the game, Zed's house, so that you have control over the information about that house and can ensure that no other player has an unfair advantage in reaping that benefit.

But you then allow players to create a bunch of other houses, freely, for which there can be no inherent benefit to robbing.  On the house list, these houses are indistinguishable from your house, causing people to try robbing them pointlessly anyway.  But they can't afford to do any robbing unless they have successfully robbed your house to get some money.  And they can't afford to rob your house in the first place!

SO... for this to really work, the admin, or his trusted, non-player team, needs to produce a constant stream of houses with god-given money inside of them.  No one but the trusted non-player team should have access to information about how to beat these houses.

To avoid a bunch of waste-of-time player-created houses, players should start with $0 when they start the game.  They should immediately go out robbing, and rob repeatedly until they finally get some money, and then be motivated to build security to protect it.  Their family is safe until they have money, because who would bother robbing them?

Conservation of value isn't necessary.  Money CAN be created all the time, but only from sources that the players have no control over or information about.

As players rob your "seed" houses, more and more player-created houses will be built and be worth robbing.  However, sockpuppets still have no use, because of where the money came from originally.  Sock puppets can't create money.  They can only steal it for real from either a real seed house OR from a player who stole it for real from a seed house or from some other real player.  Yes, you can have multiple accounts, and freely transfer fairly-stolen real money between them, but this gives you no unfair advantage (aside from reducing your risk of a total wipe-out through diversification).

This is essentially what you have now, but there was only one seed house, and a fixed sum of money.

And you can't re-use seed houses.  I already robbed your main house, so if you re-fill it with money, I'll have a huge advantage when I go to rob it again.  I robbed you twice, even after you made changes, don't forget.

So, this makes maintaining the game a ton of work for the admins!

But you really don't want people wasting their time robbing $0 starter houses.  And you've gotta get rid of the $200 backpack thing, because starting players can't afford it anyway, so it's really not going to help to put seed money in the $0 houses through robber deaths.

I wonder if an algorithm for randomly generating the seed houses could work...  Then the thing could run itself.  It might not even matter if the seed houses are solvable without tools.  The harder ones would naturally grow in value as players die with more and more tools in them.  And don't forget, all of those tools were "earned" for real with legit money stolen from other seed houses.  Some seed houses can be very easy, and they will be robbed right away for a small amount of money.

Seed houses don't need families in them....  and they wouldn't look hand-designed.  They could even have a different name on them, like STATE HOUSE #257.

And then you can go back to listing the actual robbable value on each house in the list (because there's no reason to trick someone into robbing a $0 house anymore).

And maybe seed houses would only be added when the list of non-$0 player houses shrinks below some level.  Thus, once things heat up, players can't keep grinding easy seed houses forever.  They eventually have to rob other players to keep growing.

Okay, Zed old sport!  I cleaned you out, along with everyone else in the game, and built a fortress to hold my wealth.

It seems that without the Mayor constantly infusing the game with "gift" money, this will happen eventually.  No one can currently afford to even come into my house, as far as I can tell, and no houses are worth me going into.

How the heck did you guess the timing in my house?  CHEATER! 

Well, I do notice a slight lag when taking a step robbing.  But it makes it feel more "thinky," so that's okay.

Something weird happens with the shroud, though.  When the map pans to a new, unseen area with walls that you don't know about, you see empty floor for a second before the map is received and you see the walls there, causing the shroud to come back.

Also, you take a step and THEN see the dogs move afterward.

Have you thought about waiting for the response from the server BEFORE showing the player's move happen?  I know it would feel less responsive, but maybe you need to show an icon (arrow near the player) to confirm that the move is in progress.  Then show the move happen once the map changes have been received by the server.  Then everything would be visually consistent (just a slow-to-respond version of the original look and feel).

Okay... so the only way to get money in the game is to design a house with the initial 2000 that actually KILLS people, right?

Well, I should say, the only way to get backpacks is to design a house that kills people.  Those can be sold for money, then, or used to rob.

But people can't see that my house has 0 in it when they try to rob it?

How is value not destroyed when I use one-time tools while robbing?

And do we have the deflationary economy thing in place where you periodically (randomly) adjust everyone's balances to keep a total of $10K in the game?

Wow... I've never regretted making my stuff open source!  Amazing work here.

Thank you, Zed.  Will check it out this weekend (late to the party, I know...)