next »

[Jason Rohrer]

Been playing this a lot recently, and it's pretty amazing:

http://decisionproblem.com/mercury/

First, he distilled a core, interesting essence from a roguelike and turned it into a score-based game (see classic mode).  That alone is pretty cool.

But on top of that foundation, he built a really weird, experimental structure where the best players of the game (top two from the leaderboard in each 4-day cycle) get to change the game itself in substantial, permanent ways.  If you're at the top at the end of a cycle, you get to design a new monster type, object type, or character class for the game OR remove one of the existing monster types, object types, or character classes.  And there really are no "sensible" limits in place in terms of what you can design.  If you really want to break the game (by making an unkillable monster, or way-overpowered item), you can.  And then everyone competes to master the broken game over the next four days, and the best players in that broken game get a chance to repair the game then (or break it even further).

The idea is that, in order to get a precious chance to modify the game, you have to be really good at the current incarnation of the game, and study it, and understand it deeply.  Thus, you are best positioned to change the game in meaningful and interesting ways.  Kinda like letting a Magic tournament winner design a new Magic card.

There's also a "chaos" mode where anyone can add anything at any time (which I haven't played around with), but looking at the creation interface will give you an idea of the breadth of design options that will be available to you if you win the current cycle (which ends in one day):

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

(Also, studying this stuff will help you understand the game better, so you can score better).

This is the first released game from James Lantz, son of Frank, who is only like 22 or something.  Kindof a juggernaut of a first game, but the apple doesn't fall far from the tree, I guess.

[e_va]

looks dum

[man of doom]

Sounds like a cool idea. I like user generated content as it's a good way of keeping things feeling fresh and building a community around a game.
I'm appallingly bad at roguelikes though so I can't see myself ever adding stuff to this 

[s0]

I can't try this right now b/c im on a phone but how does it prevent people from just making "ENEMY WITH 99999 OF EVERY STAT XD LMFAO" and "LOLOLOL PENIS AS WEAPON" over and over again if at all? Because this being the internet, the game is probably going to get swamped with shit like that if it gains any sort of popularity.

Also I hate to be that guy, but almost every roguelike in history is score-based.

[Jason Rohrer]

Well, if you have to invest FOUR DAYS plumbing the depths of a hard system, and you beat everyone else who is doing the same thing, and you finally get that one, precious chance to modify the game... well, that's a pretty severe filter against internet idiots who are going to add 99999 HP penis monsters.  So far, no penises have been added, and everything added has changed the game in deep, interesting ways.

And I didn't unpack the "score based" aspect much, but it's not just "gold collected" or "level reached"---it's a more intricate system with multipliers and clearing bonuses.  And because it's a competitive game, the play is all about figuring out how to max that score with the resources that you have before you die.  Which has a different feel than other roguelikes that I've seen.

[Malky]

I just picked this up, but it's already got me hooked.  One hell of a find.

[Udderdude]

As soon as a group of dedicated gaming griefers gets a hold of this (SA goons, 4chan /v/, myg0t), the whole thing will guaranteed go to shit.

[Graham-]

It's an interesting concept.

I imagine it will be tough to sustain, the concept, whether it's with this game or not (though not undo-able). A given player may change the game in deep ways but it might be too much, alienating a lot of players. Playing the game can become disheartening because you don't know what to expect. The design has literally put in constructs that breaks the player's understanding of it every so often.

Though I like the idea. Funnelling creative output from players is an undeveloped direction. There's a lot of room for growth there. The next steps are...

1.
Allowing changes but controlling how much of an effect they can have.

Maybe there could be different levels of "power" given out for different levels of commitment. A player could be forced to earn their way up to an influential state. For example, say you made several good changes to the game, you would be given more freedom the next time around. To make that system work though the constraints for being given any power would have to be loosened a little.

2.
Giving more power to not only the players that master the game but who master the act of creation. This would require "scoring" creation by measuring how deeply the game is improved.

3.
Communicating the changes more effectively, or building infrastructure for the community to do it itself, in-game.


Wikipedia does a lot with user-controlled content. They just have a very well-developed system for self-policing.

Or maybe I know nothing about rogue-likes and the idea is excellent. It's interesting either way.

[Nillo]

Maybe you could allow the winner to choose between adding in an element, or removing an element that someone else added? This means that there would be a way for dedicated players to get rid of things they see as harmful to the game (and people still wouldn't do this lightly, since it means they don't get to add in something new of their own).

[Graham-]

I believe it already allows that. Check out the original post.

[TeeGee]

I think it's an amazing idea. Who cares if it's going to be broken or not -- that's the point, it's an experiment to see what happens when you give the most dedicated players the power over the game.

[Graham-]

As an experiment I think it's excellent.

[Nillo]

I believe it already allows that. Check out the original post.

Oh, derp. Never mind then, in that case I think this is a great idea. 

[s0]

I think it's an amazing idea. Who cares if it's going to be broken or not -- that's the point, it's an experiment to see what happens when you give the most dedicated players the power over the game.

I guess, but ultimately I'd prefer a game I can actually enjoy over 4chan the roguelike. Especially because this is a pretty nice game (finally had the chance to give it a shot). With all those "ultra-lite" roguelikes coming out recently, this one of the best I've played. But... enjoy it while it's still good I guess? Maybe some of the best winner-created content could carry over into Classic Mode in the future?

One criticism I have is that early game is too much of a gamble. It's very easy to run out of turns or die to an overpowered monster on level 1 without being able to do much about it.

Also how do you create content in chaos mode? Maybe im dum but I couldn't find it anywhere.

[Jason Rohrer]

Game setup is a bit weird---editing is done through the website, not the game client itself.  You can add stuff to Chaos mode here:

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

Agree that early game is too much of a gamble.  I think you start with 40 moves, and sometimes it takes more than that to find the exit.  This results in pointless restarts.  Of course, the strategy becomes:  "do nothing but find the exit in level 1".  Plus, the clearing bonus in level 1 is too small to be worth going after (though I usually try to kill the boss there to get the +1 multiplier).  Not sure how he could fix it, though... because turn-limit tightness is crucial for the game, including later levels. 

I've wondered how the slow reveal of a level matters, gameplay wise.  I don't think it adds much, and it subtracts from overall tactics, and it adds to the likelihood of throw-away level-1 restarts.  Of course, slow-map-reveal is a roguelike standard, so it's probably just there by default.

Point is, the game is enjoyable now, and you can learn it, and then it changes gradually over time, allowing you to learn each small change in turn (and giving you four days to learn that new change).  It's not like the entire game changes.  But every four days, there's something new to grapple with as a player. 

Comparing it to 4chan or Wikipedia (where anyone can add anything all the time) misses the point.

[Graham-]

No it doesn't. Wikipedia at its core is player-controlled development. This game is a reduced kind of Wikipedia. Saying that it belongs in its own class of problems because it changes "more slowly" doesn't mean anything. In the game people earn the right to make changes. Other people are expected to undo bad changes. The entire thing slowly develops over time. That's just like Wikipedia.

I think the game is a good idea. I don't think there is anything wrong with it. My point is that Wikipedia has probably done a better job of ensuring that changes are productive. I'm sure that his game will see some good results. All it needs is for the developer to step in if a problem arises. Community-first development isn't a bad way to do things. You give something to the players, see how they deal with it, then make changes based on what you've learned. A lot of the time that process beats guessing, especially in this case where the community has so much control.

The idea that "anyone" can change a Wikipedia article is an illusion. If you go make a bad change it will be undone in a few hours. Pushing a change into Wikipedia is actually a pretty intense process. There are all these layers of safeties and checks to make sure everything adds up in the end. You have to learn your way through that community to become a contributor in the same way that you have to learn your way through this game to make a change. The difference is that the Wikipedia community is multi-layered. In it experienced contributors are in a much better position to guide the changes of newer members. They don't have to re-prove themselves every cycle.

That means that the product, the wiki, always steadily marches towards a better state. That's not a given with this game. There are too many unknowns. If players have enough power to change the game in interesting ways that benefit it, then they have the equivalent power to ruin it to the same degree. If the changes are small then they are irrellevant. If they are large then they have to be managed. Communities aren't magical. They have to be watched and sustained like anything. Either the developer does it manually or creates tools to let the community do it themselves. If he wants to do it on his own there is a strict upper bound on how big the community can grow or how complex the game and its changes can become.

Everything starts somewhere. This game is somewhere. I wasn't criticizing it. I was just drawing a parallel. It's only the stimulating games that are worth talking about. This game is one step towards the future. I'm curious to see how it develops.

[s0]

it to 4chan or Wikipedia (where anyone can add anything all the time) misses the point.

I meant 4chan in terms of content, not structure. I understand (and like) the concept behind the game.

[Chef Boyardee]

This is the coolest idea ever. I'm going to check this out!! Thanks for the heads up!

[st33d]

Bit buggy. Item keys didn't respond and had a few crashes.

I like its coffee break roguelike approach. I could probably get into it when it's a bit more stable.

Haven't identified with any of the symbols at all. If I see "H" I instantly think, might be a hippo, probably a hobgoblin. In this it feels like I'm being attacked by an algebra equation lead by Prince.

[Blademasterbobo]

what are typical high scores? i'm not sure if i have the high score cuz of when i played it, or if my score is actually good