Discuss the massively-multiplayer home defense game.
You are not logged in.
At least one of the top houses, namely mine, is not broken - it's just hard.
(assuming it hasn't been broken into in the last few hours.)
It's the one with a vertical column of electric floors stretching north a bit in from the entrance, and trapdoors just in front of the entrance.
Hint for anyone who wants to try to solve it - try to understand what the electronics in the south is doing. This may involve some experimentation (or sourcediving) to figure out the details of how electricity works in this game. As a simple exercise, determine and understand the behaviour of this setup:
+-$
|+s
|+V++
+--v+I
$: power
s: switch
+,|,-: wiring
V: voltage inverted switch
v: voltage triggered switch
I: indicator floor
Ah. Hmm. That may make some sense thematically, but I fear it's only a matter of time before people figure out how to use it to start making bitlocks again - with the hidden information now being which (say) chihuahuas are fried. I can imagine setting something like that up without too much difficulty.
For now - apologies for having broadcast what should have been hidden information about your house, Matrix. Also - that was the trickiest house I've seen so far, the first to actually pose a real challenge. After failing to get my head around it on paper, I actually made a mock-up of the critical elements in my own house, and experimented with various techniques until I finally hit upon the right combination of moves.
I won't do it again, just to give other players a chance!
I think there's a little bug with blueprints:
At the time of writing, Howard Mark Jones' house has two electric floors with fried chihuahuas on them. However, in the blueprint interface these squares are described just as "electric floor / chihuahua", misleadingly suggesting that they're alive. (I read the debug output to see that they're fried - the code is "110,71:3!". Also, I fried them when I robbed the house ;) )
(iv) Yes, maybe wait to see if there's actually a problem with "trapped" paintings.
(i) Wouldn't it be easier to get your head around if it were clearer that something had changed? You might not notice a $2000 difference, but you'd certainly notice only having $50! The exact length of the timeout, with whatever approach, is not going to be very discoverable, but it could (eventually) be presented in the client.
(i) Yes, that could work. I'd make it a simple binary - you're incarnated either rich ($6000) or poor ($50), and only your first life per [time period] is rich. That should suffice to encourage people who start rich not to throw their riches to the winds, while having houses buffeted by waves of poorly equipped desperate indigents, which fits the theme well.
(iii) Having recently been forced into the position myself, I can warrent that the role of enbittered hermit whose motivation for keeping money in the vault is simply to lure into his deathtraps those who, like the murderers of his family, would dare invade his home, is a role easily adopted.
(iv) Of course as you noted elsewhere, this leaves the problem of paintings. But returning them to the pool after a few weeks' inactivity seems reasonable.
(v) cool.
(vi) ok, it's only a minor annoyance.
Yep, the design is seeming quite solid now. Note however that making a house truly one-time robbable is harder than it looks - a determined thief can use 8 tools in their first run to soften up the house for subsequent runs, so you need to make sure that in the post-robbery state, family and vault are protected against 16 tools (and moreover that clever use of tools on the first run won't make the "post-robbery state" rather different from that you'd intended).
One last little point (to save anyone else from having to sourcedive): when there's a tie, a mobile prefers the direction furthest towards the end of that list (N,S,E,W).
Also: cats generally use the same rules to flee from your new position, but if you move directly onto a cat, it instead flees from your *previous* position.
Having just watched someone come in with 8 guns, exploit an inadequacy in my security to shoot down my wife and daughter, then suicide rather than taking out half of my >12000 savings... and then another come in to kill my son for no reason... perhaps I should retract my suggestion that we don't have enough users self-motivated to commit senseless unmotivated acts of violence.
Jason: you do realise that this game you've written is essentially an engine for destroying whatever faith in humanity its players may have previously clung to?
Let me start by saying: huzzah! v6!
This is now much more the game I'd always hoped it would be.
Most importantly, the fact that it's now possible to protect the family means
that you finally have a motivation beyond cruelty or avarice for robbing
others. Just as importantly, it provides motivation for keeping a non-trivial
amount of money in your vault: with the wife alive and with things set up
right, a successful robbery will leave the house unrobbable, and so leave you
with half the money - money which you'll need in order to fix up the house and
make it harder to rob in the future. Without money, successive robberies could
lead to a broken up house with an easily reached vault, meaning there's nothing
to stop a determined murderer using arbitrarily many tools and reaching your
family.
Meanwhile, although it's too early to say how well it will work, I get the
impression that players are starting to learn to produce the interesting
puzzles which maps force them to make. Those of us who were originally
sceptical of maps should prepare their hats for consumption.
The new wiring is also pleasing.
So thanks to Jason for his work on this.
Now, premature though it might be, I'd like to mention a few problems I'd like
to see addressed in v7.
(i) Life is still too cheap. As in v5, we're getting a lot of people leaving
their house unprotected on startup and going straight out to rob. This
leads others to (I can only assume judging by the speed with which they
disappear) sit there obsessively refreshing the house list waiting for
such a cash-filled unprotected house to pop up. Although I haven't seen it
yet, I would expect also that some will continue the practice of
deliberately repeatedly dying in a house they know how to solve so as to
leave a big stash of guns for them to steal back. All this skews the
economy. As discussed previously, a timeout before you can respawn would
help with all this.
(ii) There's something missing in the above outline of player motivation based
on wanting to protect the family. Why, once there's nothing in the vault,
would anyone want to go after your family? As in real life, the only
motivations would be perverse - and this is the internet, so perversion
shouldn't be too hard to find. But judging from the fact that many of the
broken $0 houses on the list still have easily reachable alive family in
them, murder isn't currently providing much of an intrinsic thrill to
whatever perverts we have in our userbase.
To "fix" this (perverse though such a suggestion might seem in itself!), I
suggest one simple change: a tally of the number of family kills made by a
player should be kept, visible only to the murderer. This should of course
be done in a neutral way, not making it look like a score. I strongly
suspect that this would nonetheless be enough to bring out the perversion
in those prone to it, and meanwhile act as a continuous pique to the
conscience of those who aren't murderous by nature, but who were lured by
money into such grisliness.
(iii) Children are still a pure liability, which conflicts with natural
motivations. This isn't as much of a problem as it was, but it still is
one. A suggestion for a (rather drastic!) possible fix: force a player who
has lost *all* their family to suicide.
(iv) As discussed elsewhere, the house list can get cluttered with houses left
in an unsolvable state. I actually don't think this is so much of a
problem, and to the extent it is I think a simple interface addition could
fix it: allow the player to mark a house as 'uninteresting', and have it
persist in this state until the house is updated; uninteresting houses
could appear in a different colour in the list, or not at all.
(v) I might be wrong about this, but it looks like the server doesn't get pinged
while you're exploring a blueprint during a robbery. At least, I found
that after an extended session thinking through how to handle a house,
after coming out of the blueprint I found that my character had died due
to inactivity. This doesn't seem reasonable.
(vi) Another minor complaint: after a robbery, a chihuahua was left, alive, on
an unladdered trapdoor. In order to move it off, I had to buy a new trapdoor!
It already seems unfair when it's a ladder rather than a chihuahua, but surely
this is a bug?
I also don't think periodic resets are a good solution.
I would suggest instead:
(i) remove the rule that house state is saved when a family member dies;
(ii) pay wages only on login.
That way the only impossible houses would be those with nothing in their vault.
(i) would also solve the problem that children are currently purely a liability.
If (ii) leads to problems with the economy (and with maps, I'm not sure it would), I'd suggest finding other ways to solve the problems (increasing salary, say).
Cost was something like 8000; I got the money by not playing for a few weeks. Not only crime pays!
All we need is for people to pay attention to the coming of v6, and start designing houses which don't rely on hidden information.
I've done that... my house's logic isn't hard to expose, and the puzzle is tricky but fair. It has around 2-3 grand in it currently, and is distinguished by a big square of doors at its centre. It's been there for a few days without anyone solving it (or even really trying, as far as I can tell from the snuff videos). Feel free to pound on it if you're getting annoyed with the samey top houses. I'm interested to see how long it lasts.
That still wouldn't prevent it though. A modded client could just simulate the
game rules locally so that the player can test his solution, and upon reaching
the vault, it would replay the solution by sending moves to the server in the
right sequence with "humanly reasonable" waits between requests.Yes, you can't really AFK for too long inside another house, but it could be
modded in such a way to allow you to play a house offline locally (the server
would think you are in your house). The other feature would be to replay your
last solution to the server while actually being inside the target house.
Yes, the only way I've thought of to *really* solve this problem is to
(i) not allow players to try the same puzzle twice
(ii) ensure there's no way to look up a particular puzzle from another account
(this to frustrate attempts to share observed puzzles with friends / sock-puppets)
(iii) for actual live solutions of the puzzle: put in a strict time limit, and either
(a) make unlimited undos an accepted feature of the game
(b) adopt a Fischer-clock style system, such that the player has to keep
submitting moves to avoid timing out, along with a limit on the total
number of moves allowed; add to the client a "consider" mode, allowing the
player to play out possibilities from the current position with full undo
before actually submitting their next move.
For CD, (i) and (ii) would be plausible, and might be enough to stop most
cheating of this kind. (iii)(a) is inconsistent with the permadeath theme of
the game. (iii)(b) could work, though obviously it would complicate the
interface and would require changes to the networking (a separate http request
each move doesn't sound like a good idea). Hopefully the current trust system
will suffice!
You could have the client attach already-completed moves to the pings, making the player commit to them. Not a complete solution, but would give data on which to base suspicions.
A forced delay before you can respawn (I'd suggest 24 hours for experienced players) would go a long way towards dealing with the unfortunate addictiveness of the game, as well as with brute-forcing-across-lives.
Zed: please share whatever kind of unpreventable cheat you've discovered.
I just had in mind the possibility of hacking your client to allow undo. Can't do much about that without server-side processing (and even that isn't an immediate solution, you also need some fairly harsh time limits on thinking time between turns)
5) these are people who've stopped playing the game but are still getting their paychecks. The paychecks all come through at the same time, bumping lots of houses up to this kind of money level.
...although thinking about it, there's another essentially undetectable form of cheating, which I won't even describe lest it give people ideas, which maps wouldn't render legal. So perhaps that isn't a strong argument in favour of maps.
...although free maps have one big advantage: they render the main form of cheating, which can't easily be prevented, not cheating at all.
Maybe a choice between two packs... one with 32 slots for
scouting, but you can't carry a loot sack to take anything
That's actually a really nice idea. Much more interesting than maps. 8 tools
might then be too generous for 'rob' mode - with such extensive scouting
opportunities, you could reasonably force people to actually solve most of the
house's puzzles.
With any of these "cheap vision" solutions, I'd suggest balancing it out by
adding a rule that you can't rob a house in which any of your previous
incarnations died.
And don't get me wrong, I also support the idea that the game itself should be changed, but even combining all compatible suggestion within this thread still allows a player to make an unbeatable combo lock house.
no wires through walls + animals only move within 6 spaces
still possible
You're right. I thought this wasn't possible in a tool-proof way, but I see
now that it indeed is.
So, unsatisfactory though it is in some ways, introducing maps still seems to
me the best solution suggested so far.
One worry is that the effect might be that rather than give up on combination
locks, players make the logic of their combination locks as difficult to
decipher as possible. With clever use of the details of electricity and pet
processing, you could make this pretty complicated; plausibly sufficiently to
withstand the scrutiny of those few who will be able to afford maps. You could
also make even easily decipherable combination locks just one part of a
difficult puzzle, making sure that only those with maps have even a chance at
solving it. So I'm sceptical that introducing expensive maps would actually
entirely remove combination locks from optimal play.
I fear that the only real solution I see is to drastically revise the design
of the game by introducing *free* maps.