I'm back from vaction (and plans moving forward)

jasonrohrer · 2013-10-08 12:49:32

I was gone on a family vacation for the entire month of September, and then at IndieCade after that. I just got back. Today was my first real day of working in more than a month.

I didn't make an announcement about this before I left. After all, I am making a game about breaking into houses while people are away...

So, a great deal of the game "dying" over the past month had to do with the fact that I wasn't here to keep improving it and moving it forward.

I've read all the posts about why people have stopped playing, and about the good old days, and so on. A few thoughts:

--So far, 4600 people have bought the game.

--A huge chunk of those people bought the game in the first month after release. The number of new players coming in over the past few months has greatly diminished.

--As players learn how to play any competitive online game, play variety community-wide tends to diminish over time (as optimal strategies are discovered and refined).

House variety really has very little to do with the variety inherent in the house tile set. There was a time, way back in vX, where there were nothing but combo locks with commit gates in front of the button chamber. Now we have a lot of "scary" entrance-ways with a commit gate. The point is, there's always some way to optimally spend your resources---at least in a game that is about out-resourcing each other.

My main goal going forward is still to crack this problem. It may be an impossible problem to crack, but all of my efforts over the past 6 months have really been about trying to crack this problem.

Whatever aspects of the game that I "ruin" along the way are not important: the game will never be interesting, over the long term, if this problem isn't cracked. Walking into a page full of 10 houses that are all nearly identical is not interesting.

In general, it seems like greater constraints lead to greater creativity.

Unfortunately, all of my efforts so far have been toward directly thwarting this or that uninteresting house construction. I can keep going on like this forever, it seems, as the next uninteresting house construction pops up.

Shall I now try to thwart one-way, pit bull triggered commit gates in the entrance corridor?

I'm now stepping back and thinking about the bigger picture.

--Why does building the exact same house as your neighbor work? Obviously, it doesn't, so you have to build an almost-exact same house as your neighbor, with minor changes (like a different combination).

--What's the optimal way to climb the ladder toward that expensive, rather-secure house? Can you do it by grinding? Can you do it by building the exact same house as your low-level neighbor? If so, the game is really uninteresting.

--How does money flow into the game?

And finally, the long-term roadmap is this: the game has a Steam deal now. Assuming that I can solve this potentially-unsolvable design problems, I will release the game on Steam, which will likely bring in thousands of new players in a single day. That will give us a huge boost in critical mass. However, even that mass will eventually dwindle back to where we are now if I fail to solve these problems. Thus, I must solve them first.

Please bear with me as I continue to try to figure this out.

You have all been instrumental in helping me to test and improve the game so far. It has obviously been a long and bumpy process, and I really appreciate your continued enthusiasm for this crazy experiment.

ukuko · 2013-10-09 00:05:30

Welcome back! Exciting news. Esp the Steam thing.

dalleck · 2013-10-09 02:17:09

Welcome back Jason, and glad to hear you are serious about steering the game back on track.

In considering the problem of the game, it seems finding an 'ultimate' house is an inevitable outcome of the game design, which is not necessarily a bad thing.

So I thought that this aspect could be nurtured rather than spurned by the game. Now hear me out, this might sound a bit of a leap, but...

...Random conditions.

As in, an procedurally generated variables. From the basics of prices for tools and building, to the starting cash, to the income rate, to the wife (shooter or not?).

Now, this may mean that trapdoors end up being only $50 each, but electric floors are $1000. Then guns are $50 but crowbars are $500. People now need to work around this set of variables to come up with a house design and also work on the method they can use to rob. Plus, they may have more or less income at their disposal. It becomes a race to find optimal designs within the set of limitations. Eventually a set will be found and those houses will dominate. Then to reset the world a wealthy player can buy a 'Nuke', which resets the world when used.

Then that resets the server and a brand new set of conditions are generated. Players will now need to find their way in this new balance. Constant novelty.

I also thought of another way to implement it rather than game-wide, and that is through having a list of cities which you can have a house in, and you would be free to have a house in each city. Let us say you have spread your seed wide and far. Chicago may be electric floor town, and Illinois may be dog central. Who knows? A death in any one city would destroy ALL you houses too. Death would feel oh so cruel. As before, a player could buy an expensive item to reset , but this time it would just reset that city.

Limitless novelty.

jere · 2013-10-09 08:45:47

I didn't make an announcement about this before I left. After all, I am making a game about breaking into houses while people are away...

LOL

Congrats on the steam deal. Especially in light of the last thread about steam (where it was mentioned that SiD was once rejected), that's great news.

jasonrohrer · 2013-10-09 09:39:10

Dalleck, this is an excellent idea.

Have you ever played James Lantz's game Mercury?

http://www.decisionproblem.com/mercury/

A roguelike where the top player in each round gets to add a new object to the game in the next round (or remove one, I think?).

What you're suggesting is like a "random re-roll" version of that. The fact that one successful player would be in control of when it happened is what reminds me of Mercury.

The problem here is how the hell to make it work thematically... A nuke?

ANYWAY.... before I move in this direction, I'm still trying one last thing that might turn it into a psychological game, where "trick people" is the only successful strategy, and you simply can't use the same tricks as everyone else if that's the case.

joshwithguitar · 2013-10-09 18:07:23

One issue I see with the random pricing idea is that if a certain condition set was broken in such a way that it is far too easy to rob and too difficult to defend then this would mean that no one will ever get enough money for the nuke and it will stay in this broken state indefinitely.

Thematically the only things I can think that might legitimately take out a neighbourhood are big pieces of infrastructure like a new highway or an airport. After the top player buys the "nuke" you could have a message come up when people restart: "Your neighbourhood has recently been rezoned to make way for the construction of a new 14 lane highway. Your house will be demolished in x days."

dalleck · 2013-10-10 01:12:04

jasonrohrer wrote:

Dalleck, this is an excellent idea.
Have you ever played James Lantz's game Mercury?
http://www.decisionproblem.com/mercury/
A roguelike where the top player in each round gets to add a new object to the game in the next round (or remove one, I think?).

That looks really interesting!

jasonrohrer wrote:

ANYWAY.... before I move in this direction, I'm still trying one last thing that might turn it into a psychological game, where "trick people" is the only successful strategy, and you simply can't use the same tricks as everyone else if that's the case.

When you say you wanted it more 'psychological' than 'tricky' (or 'puzzle-y' for that matter) I was thinking about playing with the lighting, allowing the home builder to control what the robber can see. The current robber sight would have to be dramatically dropped to perhaps a few squares away. Then there would be placeable light objects such as a lamp or an overhead light. To compliment this there can be a new 'Torch' object, which when turned on, will increase the vision cone in front of the robber. Each step the robber takes will lower the battery in increments until it finally is exhausted.

This will now allow the home builder to choose what the robber sees, unless the robber has torches. This could also allow for completely dark houses, which is kind of funny. The thought of robbing your own house in the pitch black, only having your hand-scrawled guide to refer to is hilarious.

dalleck · 2013-10-10 01:21:38

joshwithguitar wrote:

One issue I see with the random pricing idea is that if a certain condition set was broken in such a way that it is far too easy to rob and too difficult to defend then this would mean that no one will ever get enough money for the nuke and it will stay in this broken state indefinitely.

I can understand what you are saying here. However what you must realise is procedural generation can be the precise control of random elements. Just take a look at Minecraft to see this. So I don't think it is an insermountable obstacle. One way of many that Jason could tackle this is by tailoring the random generation to create various, let's call them 'biomes'. One such Biome may lean towards cheap dogs and expensive meat while another could be cheap electric doors and expensive crowbars.

jasonrohrer · 2013-10-10 10:33:06

Well, what I meant is this:

With the current salary that builds up over time, one valid strategy is: build a scary looking house with the starting money, then wait/grind to get your salary and add onto it (until you actually build up a scary house).

Having your house scare people away is just as effective as having your house trick people into dying in there (ignoring the income potential of tools dropped by those tricked into death).

And "Scary house" has an optimum point given any house tile configuration.... essentially the scariest house. Currently, that seems to be a pit-bull-triggered commit gate right by the door. Everyone can build the same thing---it's always scary.

But what if "trick people into dying" was the name of the game? What if you couldn't climb the ladder by hunkering down in a scary house?

Then the optimal house would forever be a moving target. Clearly, a commit gate wouldn't be it, because everyone would recognize it and walk out. Maybe some configuration that lets a dog out behind you. But as soon as everyone starts using that design, it stops tricking people. So, you have to think of something else. Something that doesn't look dangerous. Something that invites exploration. And after you find it, and everyone copies it, that stops tricking people. Eventually, as people forget about the trick that everyone was using last month, that old trick can bubble back up as a viable strategy again.

All without me constantly tweaking prices.

The idea to make this work is to make the ONLY cash injected into the game come from killing a robber (the bounties discussed in the other thread). You can hunker down, but you will just remain there, hunkered down, going nowhere.

There was a "chicken and egg" problem discussed in that other thread, where everyone would spend down to $0, so there would be nothing to steal, and no robbers would die.

However, I think this ignores the metagame of climbing the ladder. If everyone has spent down to $0, and you make a tricky house with $500 left in it as bait, you will be the only one attracting and killing robbers, and the only one getting more money.

This also opens up strategies of spending down to $0 right before going to bed, or building a temporary scary entrance before going to bed. I'm not sure that I like these side-effects, but...

Regardless, a page full of houses with $2 in them is better than what we have now (a page with a grand total of 2 active houses).

ukuko · 2013-10-10 11:18:12

Would there be a base payout for every robber death, regardless of their bank balance or tools?

Otherwise I imagine folks will probably dump their cash via tools elsewhere before committing to a real robbery.

jere · 2013-10-10 13:25:20

Then the optimal house would forever be a moving target.... Something that doesn't look dangerous. Something that invites exploration.

The thing is it's very easy to divide the house objects into things that can't kill you and things that can kill you (of which there are only, I think, 5 objects). Therefore, I don't see how areas of a house could be dangerous, but not look dangerous simultaneously. You know walking over a powered trapdoor could be deadly if that trapdoor turns off. There's never a safe looking trapdoor.

The only exception is dogs, since they're mobile. So you already mentioned a dog that pops out behind you unexpectedly. But now we're talking about an optimal strategy again...

dalleck's idea is quite interesting. It would allow even more trickery if the conditions were different per person. For instance, you start a house and, within that house specifically, powered floors are super cheap, but dogs are expensive. Now someone robbing you doesn't know what your individual prices are, but they can deduce it from the content of your house. However, you could put a dog or two in the front to trick robbers into thinking it's a dog-cheap house. I think there's worthwhile deception and metagame possibilities with that... I'm not going to pretend there's a reasonable explanation though.

Last edited by jere (2013-10-10 13:27:03)

gyuri · 2013-10-10 16:52:19

jere wrote:

I'm not going to pretend there's a reasonable explanation though.

What if you could choose a profession like dog-breeder, architect or whatever when you start a new life, and this profession comes with a certain set of prices and conditions?

jere · 2013-10-10 18:52:29

Hmm... that's not bad. Pretty clever actually. With my idea, I'd be kind of worried if you could choose the profession because then, if an optimal profession does arise, everyone is just going to choose that one. It'd be slightly better if you were locked to a certain (random) profession/prices for like a day or so.

Not sure if this is the right way to go, but some sort of random element would help break up the optimal houses. Anyway, I'm really eager to see what Jason has in mind.

Last edited by jere (2013-10-10 18:52:57)

largestherb · 2013-10-10 19:16:32

dalleck wrote:

The thought of robbing your own house in the pitch black, only having your hand-scrawled guide to refer to is hilarious.

my memory is bloody awful, i was always robbing my own house in the dark! https://i.imgur.com/ACHsbdl.png https://i.imgur.com/fV1Mdb3.png

good thing i left a phone number in one of these. hmm

Last edited by largestherb (2013-10-10 19:17:19)

ukuko · 2013-10-11 01:03:32

Those notes are things of beauty.

Cent · 2013-10-11 01:48:19

Yesyesyesyes!

Thats what I wanted to hear. I haven't had enough time to get back to playing and wrapping my mind around the new meta, but I like it when a dev tries to know and analyze his game, and treats the alpha testing cycle as an alpha test. Get out an interesting, deep game now, and you can worry about audience retention after release. The only time for big revamps is now, so make the most of it.

dalleck · 2013-10-11 04:06:19

jasonrohrer wrote:

Currently, that seems to be a pit-bull-triggered commit gate right by the door. Everyone can build the same thing---it's always scary.

An easy way to stop these commitment traps working is by allowing power to still flow through a cut powered floor.

jasonrohrer wrote:

But what if "trick people into dying" was the name of the game? What if you couldn't climb the ladder by hunkering down in a scary house?
...
The idea to make this work is to make the ONLY cash injected into the game come from killing a robber (the bounties discussed in the other thread). You can hunker down, but you will just remain there, hunkered down, going nowhere.

I like the notion of this idea, but I think the only way this will work is if a home owner can make attempts at a house without the penalty of death. The concept would be that you hire a scout with an insurance fee which is returned to you if the scout gets out alive. If he doesn't, the owner gets the insurance money. If a scout get to the safe it is treated as broken and cannot be robbed.

This will mean players will have a constant supply of scouting attempts to die in their house, while allowing them to contribute to the economy as well by scouting. I think the chills mechanism will help out with any balance issues this might cause.

Last edited by dalleck (2013-10-11 04:45:54)

Roborob · 2013-10-11 06:02:23

*Ahem*
=P

But seriously, I think making an environment where homeowners want intruders to break in (and obviously die horribly) is a solution (or at least a step in the right direction) to the uninteresting house problem.

jasonrohrer · 2013-10-11 15:40:58

Yes, Roborob, this is not my idea! I last heard it from Josh, but clearly, you came up with it back in June.

In my defense (for not listening to you back in June), it takes a while for particular issues to become a design focus. Back in June, I was focusing on other stuff that had to be ruled out first.

The latest incarnation of the idea, to answer Ukuko, involves a bounty that is separate from the robber's wealth. It starts at 500 and goes up by 500 for each outing of theft and 1000 for each outing of murder. The robber has no access to this money or even awareness of it. The homeowner that kills the robber gets the money. The amount for each death is shown on the corresponding security tape.

Now, a big question arises here:

If the metagame changes to favor tricky houses instead of impenetrable fortresses, should the other house-weakening changes be rolled back as unnecessary? Should we go back to an 8-slot backpack? Should we go back to animals moving behind walls? What about wires through walls?

I think that the answer to this question is NO, because the new metagame will only affect the lower rungs of the economy. After attracting robbers and building up a fortune, I think the tendency will be to make a strong, scary fortress to protect that fortune, since no additional money would be needed at that point. Also, lower-rung players might convert their house into scary/strong mode before leaving the game for a while (going to sleep).

I don't want ANY player to be completely invulnerable to robbery.

Also, animals moving behind walls is just plain annoying. Wires through wooden walls might come back, though...

joshwithguitar · 2013-10-12 04:30:13

Interesting idea to have the bounty increase not depend on the size of the robbery. I guess this will ensure that even fairly new accounts dying will inject a decent amount of money into the economy if they managed to rob a few places before they die.

What has led you to lean this way as opposed to a bounty based on the percentage of total robbed wealth or a combination approach? (I'm just curious, I'm not sure which approach would work best.)

dalleck · 2013-10-12 19:15:39

jasonrohrer wrote:

The latest incarnation of the idea, to answer Ukuko, involves a bounty that is separate from the robber's wealth. It starts at 500 and goes up by 500 for each outing of theft and 1000 for each outing of murder. The robber has no access to this money or even awareness of it. The homeowner that kills the robber gets the money. The amount for each death is shown on the corresponding security tape.

...so you have to kill robbers to make any money... That sounds like a significant change. As you said, it will now not be about keeping players out but creating deadly traps with which to kill them with.

Thematically it will be a bit different though. Instead of being about a homeowner defending his wealth and family from intruders it will be a story about some crazy guy who invites robbers into his custom-made deathtrap house just so he can earn a living!

jasonrohrer wrote:

Wires through wooden walls might come back, though...

Pretty please.

Dantheman · 2013-10-13 00:46:21

Nice to hear you are back!

I played this game quite a bit eight months or so ago, and have been keeping tabs since then. Good to know it hasn't been abandoned.

*** (digression)
A really interesting concept. I especially loved setting up exactly those kinds of "trap" houses that make for interesting video tapes. My favorite was to make a giant box in the middle of the room that had doors on four sides. It was a big chihuahua powered combination lock that you had to circle around in exactly the right way for the correct door to open and lead you into the safe. Screwing up would open the doors to the layer of pit bulls, but not right away. People would circle the mystery box first (after doing the entrance dance). When the doors opened and the dogs came out, it was already too late. I liked it because it was not a perfect lock, and it was not a certain death machine. You could be clever and break through into the heart with the right tools, even if you didn't know the combination. Eventually somebody got lucky and figured it out, which was fun to watch.
***

Anyway....

A thought about possible future directions. But have you considered alternatives to the giant open work space everybody current starts out with? What if there were several different home layouts of varying sizes? You start a new game with a relatively small house, and can purchase a larger one with more money. This adds something to shoot for. Meanwhile, by "layout", I mean what if you started out with an actual house (walls, doors, furniture?) pre-built that you then can modify/destroy according to your own plans.

dalleck · 2013-10-13 05:01:04

Dantheman wrote:

Nice to hear you are back!
A thought about possible future directions. But have you considered alternatives to the giant open work space everybody current starts out with? What if there were several different home layouts of varying sizes? You start a new game with a relatively small house, and can purchase a larger one with more money. This adds something to shoot for.

I think something like that has definitely been suggested before, but is still interesting to consider. Would it change the game that much? I am not sure.

Dantheman wrote:

Meanwhile, by "layout", I mean what if you started out with an actual house (walls, doors, furniture?) pre-built that you then can modify/destroy according to your own plans.

Would it be random? If so, what would stop people just re-generating their house until they end up with something they like? I can see it being interesting to save some cash on an early house by not spending on walls and just using the regular setup. But ultimately it could just present an arbitrary limitation on players, leading them to knock down most of the walls.

But hey, it could be a nice leg-up when you are first building, especially if there was a power supply or two thrown in there and perhaps a dog along with a nice little layout of rooms.

Last edited by dalleck (2013-10-13 05:02:15)

bey bey · 2013-10-13 09:21:02

Wasn't the original tool-dump invented for this purpose?

With current design restrictions, a labyrinth of commitment traps would still occur imho.

I think what we are forgetting right now is that animals behind walls weren't all bad. They meant that there was stuff to figure out and I did manage to crack the odd dance and go for the booty. Simple dances aren't the problem, they can be cracked sooner or later by starter-robber-players, as long as they can't be an entire row of switches. In a way, they made houses less safe than what's going on now and they allowed for creative defenses.

I would say again that the game would flourish if the best defense was a diversified, non-obvious house made up of different elements, not just a simple design for max. thickness. Restrictions need to apply to the size of bits of design but if they make 90% of previous design choices impossible, we can all see what happens... (In v15 that defense strategy was working brilliantly by the way...)

I still think that something similar to sight-activation for animals applied to electronics - effectively limiting the range of power distribution - would have brought about houses that would be open to trial and error starting robberies as well as to sophisticated cutting in the right places, but where the less obvious houses would have been a lot safer than the usual fortress.

dalleck · 2013-10-14 01:39:48

bey bey wrote:

I still think that something similar to sight-activation for animals applied to electronics - effectively limiting the range of power distribution - would have brought about houses that would be open to trial and error starting robberies as well as to sophisticated cutting in the right places, but where the less obvious houses would have been a lot safer than the usual fortress.

You know what would be REALLY interesting? Instead of power supplies being passive 'batteries', they could instead be switch which when activated count down from a pre-determined number to zero, once per step taken by the player. If you set a limit of 20 or so steps, then you end up with power supplies that can only affectively offer power 20 spaces away (although clever house design could work around this). It would be great for the home owner to customise a few power supplies to different timing and design their traps around this, kind of like a race. They would also effectively work like the old pet timers, but what makes it different is they are completely accessible to new players and also far more versatile and compact to use.

Of course this would allow for incredibly simple one-way passages, but to compensate for that I thought the power-meter could make a comeback (yay!) and it could be used to tinker with the timings of the power supply or perhaps trigger it to be permanently on.

Last edited by dalleck (2013-10-14 06:06:24)

The Castle Doctrine Forums

#1 2013-10-08 12:49:32

I'm back from vaction (and plans moving forward)

#2 2013-10-09 00:05:30

Re: I'm back from vaction (and plans moving forward)

#3 2013-10-09 02:17:09

Re: I'm back from vaction (and plans moving forward)

#4 2013-10-09 08:45:47

Re: I'm back from vaction (and plans moving forward)

#5 2013-10-09 09:39:10

Re: I'm back from vaction (and plans moving forward)

#6 2013-10-09 18:07:23

Re: I'm back from vaction (and plans moving forward)

#7 2013-10-10 01:12:04

Re: I'm back from vaction (and plans moving forward)

#8 2013-10-10 01:21:38

Re: I'm back from vaction (and plans moving forward)

#9 2013-10-10 10:33:06

Re: I'm back from vaction (and plans moving forward)

#10 2013-10-10 11:18:12

Re: I'm back from vaction (and plans moving forward)

#11 2013-10-10 13:25:20

Re: I'm back from vaction (and plans moving forward)

#12 2013-10-10 16:52:19

Re: I'm back from vaction (and plans moving forward)

#13 2013-10-10 18:52:29

Re: I'm back from vaction (and plans moving forward)

#14 2013-10-10 19:16:32

Re: I'm back from vaction (and plans moving forward)

#15 2013-10-11 01:03:32

Re: I'm back from vaction (and plans moving forward)

#16 2013-10-11 01:48:19

Re: I'm back from vaction (and plans moving forward)

#17 2013-10-11 04:06:19

Re: I'm back from vaction (and plans moving forward)

#18 2013-10-11 06:02:23

Re: I'm back from vaction (and plans moving forward)

#19 2013-10-11 15:40:58

Re: I'm back from vaction (and plans moving forward)

#20 2013-10-12 04:30:13

Re: I'm back from vaction (and plans moving forward)

#21 2013-10-12 19:15:39

Re: I'm back from vaction (and plans moving forward)

#22 2013-10-13 00:46:21

Re: I'm back from vaction (and plans moving forward)

#23 2013-10-13 05:01:04

Re: I'm back from vaction (and plans moving forward)

#24 2013-10-13 09:21:02

Re: I'm back from vaction (and plans moving forward)

#25 2013-10-14 01:39:48

Re: I'm back from vaction (and plans moving forward)

Board footer