Thanks everybody! Wow minmay, that's a long reply, I appreciate your effort! I'll get to your points one by one; they are really interesting.
minmay wrote:Your "Creative Points" system sounds interesting except for one thing: I really, really don't think they should be awarded for playing the game, even if that happens to include rating content. The most active players and the best designers are never the same group. I have more Crawl playtime than dpeg but do you think I would be a better developer? I think your willingness to grind for points shouldn't be a factor in your "power." The fact that you were not only able but willing to use the word "farmed" while describing it is a red flag...
I see what you mean. The idea, however, is that rating only gives you so much Creative Points. The actual, sizeable source of Creative Points is actually creating good content. The necessity to get into the system by playing Experimental and Default first is just to make sure every designer has a basic grasp of the game and the type of content the community wants. Then, once two or three ideas have been taken into Default, the monthly or weekly points revenue should be much higher compared to "farming". However, all of this is debatable and should be carefully tuned to become a system in which creativity isn't suppressed, but enforced, and in which rating other people's content is desirable enough than people who wouldn't be naturally bothered to engage in that sort of gameplay will be motivated to give it a go.
minmay wrote:So I'd just get rid of that, and allow adding content to Experimental mode for free, with the option to spend points to increase its frequency. Instead of "make X copies," use a weight system, like Crawl has for vaults - when the level wants a vault, it picks one from the relevant list, and the chance of any one vault being picked is (its weight)/(total weight of all vaults in the list). So every item would start with a default weight, and you could spend points to increase it. Also, getting good and bad reviews should probably increase and decrease the weight, respectively.
I do like this idea! I think it's nice that everyone is entitled to get their content reviewed; however by paying more, you get it reviewed faster. I also like the snowball effect of positive reviews increasing the weight. Thanks!
minmay wrote:Don't worry about a lack of investment. They invested the effort to create the content in the first place. Sure there are going to be a lot of flying cock-and-balls monsters but unless you make reporting really cumbersome somehow, those users will be banned pretty fast, and you'll be recruiting moderators from the player base, so the percentage chance of an Experimental player encountering a cock-and-balls monster is going to remain near-constant regardless of the size of the player base, and if it's too high, you can always lower the default weight or the threshold for banning.
True. However, offensive content is not what concerns me the most. I'm more scared of broken content, like a vault that kills you no matter what or an item that completely destroys any semblance of balance. These may not have any bad intention by the developer, so it's not fair to ban them on those grounds. However, they need to be moderately punished so they don't release half-finished content. I guess downvotes swiftly reducing weight is a good way to do this.
minmay wrote:My other major concern is learning the game. Considering the size that you're aiming for, if you're not in Canonical (and perhaps even if you are), you're not going to see the same thing twice. How are you going to know what it does? Hiding significant information about anything is pretty obviously undesirable - if the player doesn't have a spoiler, they won't know what it does, and won't be able to actually meaningfully think of tactics. If they do have a spoiler, well, your game now has 1 minute of reading a wiki for every 5 seconds of play.
So you need to come up with a way to communicate properties in the game, without it taking a long time (1 minute of in-game reading isn't much better). This means you need to have limitations on user-made content that are perhaps tighter than you would like, in order to keep it explainable.
I guess discussing this point involves going over the interface and other game details, but I have a couple ideas in that respect. First, I thought the player should have the ability to pause the game (although its implications on multiplayer should be explored) and read the game logs. Despite being a real time game, I wanted to keep it text-based through a series of different logs such as a OS log (which reports you with your own internal status) a sensor log (that feeds you information of your surroundings) etc. Once paused, the player can take his time to read the logs, and even run some risk analysis software (which is instant in game time) to evaluate enemies in the vicinity and so on.
Any content that enters canonical should be as non-spoilery as possible. This implies detailed descriptions of the creature or item in question, reachable through direct inspection or whatever measures the player has to analyze its surroundings. Also, canonical should axe content as fast as it introduces new pieces, so the average quality always increases and the total amount is understandable.
minmay wrote:Finally, your "STRICTLY qualitative stat system" is going to be quite a pain for designers...I'm supposed to make a new gun, but I'm also not supposed to know what it does? That's going to make balancing it awfully tough. There are lots of people who enjoy playing games, roguelikes included, unspoiled, but I can't imagine wanting to develop one unspoiled.
I find this point interesting, because part of the reasoning that got me to consider only qualitative stats is making it much easier on the developers.
I'll give you an example: say you're making a gun and want to decide its weight. You could code a number in Kg, but... What would you do? How much does a robot gun weight? What is heavy and what is light in this context? Or if you happen to be from the US and used to thinking in pounds, how does that translate to Kg? Instead, having a slider with 7 states that reads "Weightless, Very light, light, medium, heavy, very heavy, extremely heavy" is, in my opinion, much more intuitive and direct.
Categories such as weapon damage can be mentally assumed to be a number for all purposes. However, the end result is more intuitive. But most importantly, what I like about qualitative stats is that they help clearly and concisely stating game mechanics. For instance, a passive may read "Your attacks with
heavy melee weapons deal
more damage". Heavy and More Damage would be bolded or displayed in a different colour to make it clear they're key words. Going back to the weapon design phase, by tagging it as "heavy" you can clearly picture the interactions with other game mechanics, rather than slap it some arbitrary number and hope for it to be balanced or to work nicely with future content.
minmay wrote:Other than that, it sounds pretty good, although you clearly like Diablo's interface a lot more than I do.
Thanks! About the interface... Eh, I have mixed feelings. The problem is, I have to admit the roguelike interface is something people have trouble getting into, and in a game so reliant on a healthy community, it's important to make sure it's approachable. Also, real-time makes racing more fair, I think.
Sunday, 10th March 2013, 10:43 
)