User generated game design.

[Nachimir]

Me and Scrobbs were having a conversation on Twitter today that I thought was really intersting. This post about user generated game design going really, really wrong sparked it:

http://www.eldergame.com/2009/05/11/user-generated-quests-and-the-ruby-slippers/

* Players subconsciously calculate the cost-to-benefit ratio of content when deciding if it’s fun. For most MMO players, more reward = more fun. (This is a bitch of a lesson to learn, too. “My custom-scripted quest was so incredibly cool! Why aren’t players doing the quest? Well, yes, the reward was a little sub-par, but so what? You’re telling me they aren’t playing it because of THAT? Players can’t be THAT shallow!” Ha ha, newb.)

* Players aren’t objective reviewers. If you ask them to grade content, they will grade more rewarding content higher than other content even if it isn’t as good by other metrics (like plot, writing, annoyance factor, or originality).

* Many players spend incredible amounts of time finding ways to min-max the system so they can get more power for less effort. That’s part of the fun for many players. So there are tens of thousands of people actively looking for mistakes, loopholes, and gray areas in your game. All the time.

I find this really interesting, and think it's the exact reason Eskil Steenberg is wrong to say "fire your game designer". Short term goals can short circuit long-term value, i.e. if we have an option of two missions with the same reward, and one is easy, we tend to go for the easy one.

Tweets:

Scrobbs: @nachimir I remember limited game modding that I did way back - little effort for great reward was fun for a while but got boring very quickly. It's similar to using cheat codes in SP games - fun for a bit, but when you can beat the game in seconds, there's no point.

Nach: @Tonsko IMO second bullet in 1st list really interesting. People judge things differently over time. we like immediate reward, but look back on difficult challenges fondly. Longterm thinking not inherent to UGC but is to good design?

Scrobbs: re: longterm thinking - is this a problem with the immediate gratification required of our society now?

Nach: I don't think it's societal or a pop culture thing, though "social decline" camp would have us believe it. A lot is even genetic IMO

Scrobbs: it could kill the game. 1) The instant levelling will turn players off long term. and 2)they won't keep getting new players to fill that gap.

Nach: any sensible modder/player/developer gets bored of instant gratification, though UGC = constant throughput of newbies. it definitely would kill the game. UGC game design is behaviourally similar to phosphates and health on Nauru: http://u.nu/88a5

Scrobbs: @nachimir That's a grim story

Nach: @Tonsko *very*, but in several ways a priceless illustration of how motivation/emotion lack long term perspective and can be self-defeating.

And I missed this one, sorry Scrobbs

Scrobbs: but I wonder if a company successfully get UGC working - will that not make the job of a developer/designer less important?

[Scrobbs]

Good work!

The part about players aren't objective reviewers - sometimes they're not objective players either - like for instance the people who, when a new game is released, race to be the first to finish it, or spend their time in the game world trying to find all the glitches and bugs that will give them an unfair advantage.

I remember Remo talking about the latter while playing Hatsworth - although I think he is to be forgiven for exploiting where possible just because the game is so ridiculously hard.

I wonder what it is that drives people to do that/be like that - with most of us here, I imagine, you're happy to let ourselves be immersed in the game if it lets you to enhance the experience as much as possible. After all, that's why you play it. Some players obviously get their gratification from destroying or exploiting their environment - but, I had no idea of the sad evidence CoH has produced - that those kind of people are in the majority, and would do it if they could get away with it.

[OssK]

Very interesting link there thank you.

My humble take on the matter would be that you'd have to implement a really borring quest generator based on time to have it work.

You pick from 10-15-30 minutes quests (that you don't explicitely name that way)

then pick the type (fetch, kill, carry, fuck with etc...)

then pick two points at a given distance from one another on the world map etc...

Only by making tools narrower can you lower the exploitation of such a UGC IMO.

Then you'd have to think real hard to get enough content into your generator for it to be emergent but always keeping in mind that it has to stay very formated for it not to be over exploited.

You could make a similar tool of quest making and instead of giving it to the players, let the developpers play with it, that way you can rapidly implement quests in the game not worrying about players exploiting it.

And if you're looking to find out why Eskil Steenberg is wrong you need only read what he said on the matter to find out that not everybody has 5 years to come out with a game like his and not everybody is skilled enough. Now Love does seem cool but it's not the best game in the world that's pretty clear, so if hiring game designers makes better games than LOVE, we should definitely do it more often...

[Scrobbs]

The point about narrowing the creation tools is a valid one - but you risk making it pointles for the players to use - if it's just going to knock out a set of identikit quests that are the same as the dev ones, ("Go here, kill this many of these, collect this, go back"), they you may as well not bother. That in itself is a large part of why I stopped playing WoW.

Sorry to harp on about EVE again, but one of the guys in our alliance has managed to create the first player run NPC 'mission giver' in the game. The way it works is that you'd go and see the agent who would give you a mission, but instead of the standard boring ones of, "See this bit of space here? It's infested with bad guys. Go kill 'em", or, "There's a shortage of Quaife in that system. Please take this there!" and so on, the agent will give you missions that mean something to our alliance and carry the story forward in the bit of space we live in. It' pretty good, and it generally results in PvP combat as the missions invariably extend into enemy territory

It seems to work ok as well, as you're not just fighting the game - you're fighting people.

[DanJW]

My first reaction would be to suggest that while the quest is UGC, the reward is procedurally-generated based on the resources used in the quest. This would be no mean feat and might lead to exactly the kind of 'narrowness' you described.

Problems include things like spacing. Travelling a long distance makes a quest take longer and take more effort. SO it should give better rewards. But not if the two locations are near fast-travel points.

Another: number of enemies. You could drive up the 'point cost' of rewards by throwing in lots of small enemies. But lots of weak enemies spread out are easy to defeat, whereas exactly the same enemies are lethal in a single swarm that attack together - more dangerous even than a single very strong enemy.

I'm optimistic that such as system could be devised, but it would have to be incredibly sophisticated to be both balanced and comprehensive.

[OssK]

My first reaction would be to suggest that while the quest is UGC, the reward is procedurally-generated based on the resources used in the quest. This would be no mean feat and might lead to exactly the kind of 'narrowness' you described.

Problems include things like spacing. Travelling a long distance makes a quest take longer and take more effort. SO it should give better rewards. But not if the two locations are near fast-travel points.

Another: number of enemies. You could drive up the 'point cost' of rewards by throwing in lots of small enemies. But lots of weak enemies spread out are easy to defeat, whereas exactly the same enemies are lethal in a single swarm that attack together - more dangerous even than a single very strong enemy.

I'm optimistic that such as system could be devised, but it would have to be incredibly sophisticated to be both balanced and comprehensive.

My guess would be that what you described is exactly what they tried to do, but as you try to diversify your game, characters do specialise and are more effective in one kind of quest resulting in low-profile (if not full scale) exploit :

If you're better against swarms of creatures, there is simply no way you'll ever get into a troublesome quest for you... so you just spec all the way into something and quest exclusively on that peculiar game spot, on the long term, specialised specs emerge, some for which the cost-reward ratio is high (for instance, characters that are specialized in big boss battle action can earn more <insert here reward currency> in one single quest than others for an equal amount of difficulty so every swarm-spec'd and average-monster-spec'd will do a boss-spec'd character)

The richer and more complex the system, the more flaws it shall allow, so if you build it out of pre-made blocks, the emergence* will be hard to create if it were ever to come... and only by making a fucking huge amount of function-blocks could you provide enough content for the quests to be not-exploitable...

Of course this does not resolve the specs dilema, what would however are limitation mecanics (you have to pick at least one fetch+one boss+one swarm items (and each one has a cooldown so you can't pick the same again) to make a 15 minutes quest but it's of course much more valuable on a time/benefit ratio than a 5 minutes fetch quest). To me, time ought to be the major common ground to all these quests, that way you can always find easier quests for your character but at least everybody spends the same time. Difficulty can be included as a weighting mecanic but time's really my focus on that subject.

*hmm... Game Design term here, does everybody know the technical meaning of that word ?

[DanJW]

Yeah, it was a bit obvious really. But now I have idea number 2! Are you ready? Try to look ready then.

The User tools should not include rewards. Perhaps they can have a small fraction of the auto-generated rewards, as described previously, but they should not approach the official developer content rewards no matter how large the quest.

Next step: user filtering. As noted in the article, volume is a problem when it comes to vetting. The only sensible approach here is the Little Big Planet one, which has user rating and tagging. It's not perfect, but along with 'player education' initiatives it can work.

Then the developer live team can evaluate the top rated user content, and each week or month designate an 'editor's choice', lock its content and assign it suitable rewards. Perhaps some kind of in-game prize for the author also.

The user vetting is open to vote-rigging as ever, but is mitigated by the developer's choice. This is still quite a bit of work for the live team, as they have to judge the top content and then create more in the form of balanced rewards.

Also the system will break down if the smooth running of the game is interrupted, eg the number of players casting votes drops too low, or if the live team is suspended by the publisher. Any other foreseeable problems?

[Nachimir]

I was thinking along these lines too: a systemic relationship between tasks and reward, with the user either making a quest have having little to no control of the reward beyond maybe a category selection, or absolute choice over the reward, but the quest editor then makes them build a quest that matches up to the reward cost before it can be published.

Some kind of distance+enemies+average completion time measurement of quests might work. Lots of people would still build dull missions (Beat up all the dragons in this square room. You did it, hoorah for you), just like they make shitty box maps for their FPS clans, but these would get voted down. A system like this would still have loopholes* and bugs in it, but it might reduce blatant abuse.

* For instance, light levels can enormously affect the difficulty of something, but if a quest editor were to rate dark quests as more difficult, there could be all sorts of abuses of the lighting to create inconsequential darkness. People meta-gaming an editor designed to balance things would be interesting...