» the skewed views of a large opinion: Persistent World Design
Thursday, February 26, 2004
  < The Casual Gamer Gamble >        
Most of my suggestions to this point have centered around the idea of trying to attract and retain more casual players. Across the board, I recommend lowering the design barriers that keep casual gamers out. An interesting offshoot of this, is realizing those same barriers are what keep many hardcore players in. By designing a game to entice in the casual player, we're gambling that our hardcore players won't leave.

Furthermore, If we lower the time investment required to jump in and have fun, we risk lowering the attachment powergamers feel to their characters. The testament to this attachment as it stands is the numbers of hardcore players who maintain their subscription to persistent worlds long after they've ceased playing powergamer-level hours. They continue to pay $15/mo for the right to play a game they play only casually, purely because they can't bear the thought of their character being deleted.

Even when disgruntled, hardcore players are much more likely to 'tough it out' and continue their subscription while waiting for expansions, fixes, or promised future features. Hardcore player attachment to their character 'hooks' them into providing easy money to the publisher. It provides a 'safety net' of sorts against balance issues, as players know that if they cancel, they risk 'losing all their work' and might have to re-level when the fix/expansion/update they want goes live.

This effect is already demonstrated by the more casual gamers who currently play persistent games. They haven't put in the same hours as the hardcore players, and they don't pay for a subscription very long, if at all, after they're ready to move on. So if we entice more casual gamers to our game, we risk turning even our hardcore players into casual players economically - so what happens to this easy money? What happens to our safety net? We're gambling with it by designing a game for the casual player.

All this gambling for a class of player that isn't too 'hip' to the monthly fees of our genre in general. These fees are often the pivotal requirement that convinces casual gamers not to play our games. If a player has to put in 100 hours to get to the 'cool stuff', and casual players only have 20 hours a month, then they'll have to pay $75 to get to 'cool stuff'. Meanwhile these powergamers, who already frustrate casual gamers, can get to the 'cool stuff' in $15-30. On top of everything else, casual gamers have to pay more to play the same game. The monthly fee is an added insult to the injury from our designs.

If we are to be successful, it will be because we've lowered the time necessary to get into the game and have fun. Ideally, we would strive to keep a strong attachment between player and character. There would still be money, possessions, status, land, pets, and prestige on the individual character that the player would feel attached to. Yet, despite any mechanisms that we might code, lowering the time required to jump into the game and have fun with a new character will result in players being much more willing to cancel over 'smaller' issues.

I say: encourage it. We need to find a way to make it easy for them to put the game down for a while and come back later when the spirit moves them. Reassure them that canceling won't likely result in their character being deleted. We can't stop casual gamers from behaving this way - so why not make it easy to put your account 'oh hold' and, more importantly, easy to come back.

I've been thinking about giving DAoC another shot lately. People say they 'fixed' a few of my old issues (CC dominance & Paladin worthlessness in RvR), and they've certainly added some new things that look interesting. A big part of why I won't go back though, is because I remember what happened when I went back to EQ, and UO before it. I had the itch to play - but it was more casual than before. The core mechanics still prohibited me from doing enough cool stuff in the time available to justify the cost. Furthermore, resubscribing for a short period of time is a hassle. Because these games demand so much time, a cancellation is not made lightly, and is generally followed by an uninstall long before the itch to play again arises. So a player must reinstall, resubscribe, patch up, reevaluate play time vs fee, and inevitably cancel again because it costs too much to be played casually.

A recent discussion of a lecture by Scott McCloud stirred up a Micropayment discussion on games/.. That thread helped me recognize and formulate a solution to both the problem of the monthly fee barrier for casual gamers, and the hassle of resubscribing. Don't worry - it's not the 'micropayments' he's been trumpeting for the last year. Well, not entirely; but sort of.

Basically, it's a suggestion to go back to hourly billing. *ducks* Not hourly billing the way it was - but a much nicer hourly billing with a new twist. The first important addendum to the old style would be monthly caps. As I've said before, hardcore players are too important to alienate. We don't want to punish them the way straight hourly billing will. These monthly caps should be in-line with current monthly rates (~$15/mo). In fact, we should maintain the mechanism to 'subscribe' to a recurring billing package outright. We should even go so far as to make the monthly fee a dollar or two lower than the monthly cap.

Now, if we keep hourly fees small enough ($0.25), we may well be able to lure the casual players into giving us a more honest shot. The problem with hourly fees and end-of-cycle billing is that we run into higher costs for more frequent contested charges, chargebacks, support calls, etc. Furthermore, we're 'out' the server resources players consume (bandwidth, processing time, etc) before they pay up. That provides incentive for players to deny payment. We then must eat the cost of tracking them down and collecting (not fun, not profitable). The logical solution then is to charge them these tiny amounts as they use our service: micropayments.

The problem with micropayments in general is that they require a secure public key infrastructure that currently only exists in restrictive proprietary implementations. As anyone who has dealt with Visa/MC knows - being vendor locked for payment processing is not a desirable situation for the merchant. Instead of a general micropayment scheme, why not maintain a preloaded 'balance' and run our own internal micropayment-ish system that simply deducts from that balance?

Instead of giving Sony my billing information, so they can charge me $15/mo - why not allow me to preload a balance with Sony with, for example, $20? As I play EQ the hourly charges would be deducted from my balance (up to any monthly/daily caps). This way the casual player won't feel ripped off because he isn't playing as much. We may still lose the 'easy money' from the hardcore gamers by designing for the casual players - but we can more than make it up in the unused 'balance' of the casual gamer. It's the same economic concept behind Gift Cards. The cost required to manufacture and advertise the cards, process those transactions, and maintain virtual 'accounts' is made back 10 times over by the number of people who buy a card but never use the entire balance.

Similarly we could implement a system where balances are decremented a nominal amount every year, so that eventually all balances will go to zero. This wouldn't be a system motivated by greed. Rather, it is a system designed to protect ourselves from eternal data growth, and running the risk of being sued by a customer whose $3 balance from 3 years ago got 'lost'. This is the reason why Bestbuy 'expires' gift cards after a year, and other vendors deduct $1 a month or so.

We can still require a valid credit card to be present when a player creates their account. This allows us to maintain the benefits of positive player ID (to limit repeat problem players) and legal verification of parental consent for any minors. However, with no hard and fast concept of an 'open' or 'closed' account: when can we delete characters and their possessions to free up space and reduce query times? We certainly have an obligation to maintain the characters of a player who has a positive balance, but how long do we store data for players with a $0 (or negative) balance? Storage is certainly cheap, though every persistent world to date has still felt it necessary at some point to clear out old, inactive accounts. By making a game more friendly to the casual player, we are inviting many more of these stateless accounts to be created. The storage for these stateless accounts may well become a considerable problem.

Ideally we could leverage cheap storage by simply creating an 'archive' database machine that mirrors the character and item table definitions for the live system. We then could simply move 'inactive', $0 balanced, accounts over to the archive (where they aren't a detriment to performance, and can be compressed). Should the user ever recharge their balance, we could painlessly move them back to the live systems. Such a system might work so well that even 'inactive' accounts with positive balances could be moved to the archive (so they aren't dragging performance down unnecessarily). We would then of course have to define 'inactive' with a reasonable window of time, and bring that data back to the 'live' system once an 'inactive' positive-balanced user logs in.

Another concern would be: what happens to a player whose balance hits $0 during a play session? We probably wouldn't want to outright boot them from the game, or freeze them until they pony up more dough. Still, we can't simply enforce the balance logic at the attempted start of a session, as a player might just leave their account logged in. At some point we'll doubtlessly reboot the server, but we wouldn't want to encourage people to see what they can get away with. So likely we'll have to notify players in-game should their balance hit $0, and kick them if their session runs past a reasonable window of time (8 hours?). Similarly we should let them know what their balance is at the end of each session, so good players can avoid the problem altogether.

With this 'balance' payment system we could also easily facilitate any item or character sales a player might want (if your design allows/encourages it). UO for instance could have an in-game mechanism to allow a player to pay for a character-buffing - the charge for which would be automatically deducted from their account. Similarly requests for a server transfer could utilize the same mechanism seamlessly.

It may seem like an off-the-wall idea, it may be unproven - but it just might work. 
  < / The Casual Gamer Gamble >        
Tuesday, February 24, 2004
  < Quests: Examples >        
Based on the previous post, here are some samples of leveraging task misdirection, rumor, and varying levels of 'failure' to improve questing in persistent worlds. These examples are separated into the categories of: retrieval, delivery, kill, protect, explore, and create. These ideas can even be seen as potential goal/action selections from a Planning Quest System as outlined by Richard Bartle in his Dawn of Time column.

The item the player is sent to retrieve is also sought by an NPC opponent.
This NPC is actively working to obtain the item and should they get it, will bring it to some location. The player is informed of his rival, and is instructed who they talk to for more information about them. If the player delays in retrieving the target, they may face the more difficult fight of retrieving the item from the rival directly. The player could track down this opponent first, and try to ambush him in an advantageous setting. The player could let the NPC acquire the item in question, and then attempt to steal it away from him at his home, or again, through ambush.

The item to be retrieved is in one of X possible locations.
Even the quest patron does not know which location the item is at. At each false location is a random clue that would lead the mindful player to the correct location without wasting further time. Several possible clues should exist for each location should it be the 'correct' one. We should ensure that it is not possible to receive duplicate hints from two or more false locations. (X locations means each location should have ~(X-1)*2 clues that refer to it, should it be the true location.)

The player is unwittingly stealing the item from its legitimate owner
As with the other deception-based quests in this list, random clues would be present on the quest patron, and drilling down into the conversation might reveal additional clues. Further, rumors regarding the item in question might reveal that there are rumors its about to be stolen, etc. The player would be allowed to expose the plot to the authorities, betray the patron to the target directly, or simply steal the item.

The item is a lure, and the player is being sent into an ambush
The player likely will have offended some faction, and the quest patron is an intermediary for them. The player is sent after a worthless trinket simply so the enemy faction can attempt vengeance. The fight should be balanced in favor of the questor/group -- they should be set up to be surprised, not hopelessly outmatched and killed.

The package to be delivered was stolen from another faction.
The quest-giver will gloss over the inherent danger unless the player presses for details. Agents of the victim will pursue the player while they carry the item (will spawn at various waypoints to pursue the player). If the player should die at the hands of these agents (or if one should successfully pick-pocket the item), they will take the item back to their (instanced) home, where it must be retrieved if it is to be delivered. The faction of the quest giver may refuse to do business with the player until the item is retrieved.

The package to be delivered is a trap.
A villain is trying to fool the player into being an unwitting courier. The contact NPC has a few random clues that might lead the mindful player into recognizing the ruse and reporting the go-between to the authorities before the device is ever delivered. Delivering the trap incurs a personal price-hike (and faction hit) from the now suspicious target vendor. The target/alerted authorities might then issue a quest to track down the perpetrator and locate a list of further targets. If we have a thriving backstory in our world we could easily create this quest randomly from a selection of possible vindictive originating factions and a random selection of their enemies for the target. Seeking the deceptive quest-patron might lead to an (instanced) temporary hideout perhaps with traps, guards, etc.

The package turns out to contain a sentient creature (eg. fey, imp, pet, etc)
The player isn't told of this, and needn't be forced to discover it. Some clues to odd weight distribution, or some shifting about inside the package should be given. Finding a way to open the package could be a lateral diversion within the quest itself. The player may need to consult a locksmith or a mage to discover how they might open the package. The creature itself may simply flee, may fight back against the player (it won't necessarily be friendly), or may offer the player a reward to cross the original patron.

The charge to be escorted is an assassin.
The player is (un)knowingly protecting and helping this assassin gain entrance into the home of the target. The player may be allowed to assist in the actual assassination (instanced home invasion) for an additional reward and faction adjustment, create a diversion to get the authorities attention (less faction hit, less reward), betray the assassin and protect the target, or deliver the assassin and report him to the authorities. Again, a random selection of originators, targets, go-betweens, and clues would be helpful to maintain a reasonable amount of depth.

Opponents will attempt to abducted the charge en-route.
If the opponents get to the charge, (or defeat the party) they will spirit him away to their (instanced) lair. The questor(s) would have to then rescue the charge before they could deliver them. The originating faction might cut ties with questors who give up after the abduction, or they may offer them a ransom to be paid to the abductors.

The charge doesn't want to go to point B.
The quest could be a prisoner transfer, during which the charge will attempt to escape. Alternatively the charge might simply be an NPC in an arranged marriage (or some such), and would prefer not to go to point B, but won't resist physically. The player might be offered a bribe to 'lose' their charge, or be offered a reward to safely deliver them to an alternate destination. The charge may even be lying that he can pay them when he arrives. The prisoner being transferred may lead the unwitting party into his (instanced lair) for an ambush by his cronies. Again the player who is engrossed in the world and recognizes who the charge is, and what his situation is will have a leg-up on the quest -- incentivizing players to become engrossed in the world.

The target offers the player payment
The player can choose to take a new reward to go back and kill, expose, harass the originating NPC.

The target is on someone else's hitlist as well
The player must try to assassinate the target first, or failing that, defeat the rival assassin so the quest-giver's faction can claim responsibility.

The assassination was a test
A friendly faction may be testing the questor to see if he is truly trustworthy. The player will discover that the target is a faction-ally and may report the request to a higher-up. They may tell the quest-giver that they refuse the charge, or they may kill the faction-ally anyway. At such time the ruse would be explained, and clearly there will be a faction adjustment. More serious consequences may follow for those who kill the ally.

The player is scouting for contested resources
The player has been sent to look for a new vein of ore in a dungeon, and can use the informant network to find a different buyer for the information.

The player may stumble onto another quest
They may see a crime occurring at the target location, and can choose to intervene for a possible double-reward. Intervention would not be required for completing the exploration quest itself.

The player is sent to spy on an enemy convoy
Instead of going to a single static location the player can either follow the target to find its destination, or it could sneak into the convoy and steal the travel route. The player may also attempt to approach the convoy and offer to sell-out the quest-giver, and perhaps even return faulty intelligence.

The player is sent to get a map of an enemy structure or their defenses.
The player could opt to cross his patron, map the structure by casing the place manually, or break into an authority's office to steal the existing plans, or make a copy of the plans. Perhaps there would be a varying level of reward based on the quality of a questor's cartography skill.

The player is tricked
Once again a nefarious villain has tricked the player into giving aid to his faction enemies. Random clues and in-context story hints that the quest-giver is asking for something that is known to be a jealously guarded secret may tip off the player. Should they deliver the item, they may be allowed to report it to their own faction or the authorities and springboard off into a kill/retrieve quest.

The item to be created is flawed
The player is sent to find a recipe, and then craft a certain widget. The instructions, however, were intentionally wrong. The player can optionally seek out the proper recipe, find out who sought to fool the original patron, research to determine the correct recipe themselves, deliver the flawed item, etc.

The key to these scenarios is in keeping them broad enough to where they won't feel static. Two players may likely both have a quest to race an opponent to retrieve an item -- but they should have a good chance of having completely different experiences. Again, rewards for all choices should be made so that players do not invest time in quests for no reward. Failure should contain a penalty that isn't insurmountable, and should never destroy the ability for a subsequent player to complete a similar quest. The prince may be killed when a trusted group fails a covert escort mission, and we may wish to make that permanent and write it into our fiction. But we should not let that prevent other groups from attempting a similarly balanced high-profile-escort quest. (Though it may be merely a high-profile merchant, or visiting emissary)
  < / Quests: Examples >        
Thursday, February 19, 2004
  < Quests >        
In the tradition of pen-and-paper roleplaying, the idea of Questing is central to most persistent worlds. Quests can vary in difficulty, length, context, objectives, and consistency - yet players routinely lump them together logically into a single classification. The derogatory description for every hated quest under the sun is simply: FedEx.

We know what FedEx quests are, and we hate them. They leave the player to feel like an unimportant errand-boy who is secondary to the actual story or, at best, a participant with no influence over the course of events. Failure is nearly always literally impossible; success may be put off indefinitely but officially losing does not happen. They never allow us to leave our mark on the world. Whether I rescue the Princess from the Foozle ultimately doesn't matter. The Princess will always be kidnapped again, and almost any persistent world we might play in will need our help in rescuing the generic Princess.

Before we go too far, let me state that I don't think that the ability to irrevocably 'fail' a quest - or a direct punishment for losing is necessary for a fun quest experience. The Evil Baron doesn't have to be able to kill The Princess for the fight to free her to be enjoyable. The goblin menace doesn't actually need the capability to burn down the entirety of Peaceful Village for the player to care about trying to stop them. In no case should the player be effectively penalized for trying (via a time investment in a quest that yields no reward). Nor should the player be unable to correct what he did.

A well-seasoned developer once stated, and I'm paraphrasing here: All quests are FedEx. He was speaking with regards to the actual mechanics underlying all quest systems and he was absolutely right. There's no way around it. If a player doesn't feel engaged by a quest, they can always dismiss it as 'FedEx', because the underlying preconditions for success can always be reduced to: <player> at <location X> with <flag Y> = <reward>.

However, abstraction to this extent renders even The Lord of the Rings story a FedEx quest: <Frodo> at <Mount Doom> with <hasOneRing> = <middleEarthSpared>. As an exercise for the doubtful reader, try to identify a fictional narrative in which this abstraction cannot be made. Few people ever stopped to consider the plight of Middle-Earth to be a mere FedEx quest, yet it was. The lesson then is for us not to spend time trying to find some quest formulation which is mechanically different from 'FedEx' (which is fruitless). Instead we should be concentrating on ways to engage the player in the quest to begin with.

Improving the average quality of the story surrounding persistent world quests is a universal goal. Quality in dialogue, quality in mechanical polish (no broken quests), and quality in immersion within the world (tying in our elaborate backstories and built-in social structures). Improving writing quality and story is well outside the scope of my little diatribe so I won't deal with how to improve that. Suffice to say: it's very important, it's beyond most coders & designers alone, seek professional help.

So we are left with looking into the mechanics themselves to improve our quests. Let's consider the capabilities that 'merely' FedEx structure still allows. Note also that Hint sites will always boil down any quest or operation into easily digestible steps and possible rewards. Good and bad quests alike will be reduced in such a manner, so we shouldn't waste too much effort trying to block it. Sony went to great lengths to obfuscate the path to unlock their Force Sensitive slot in SWG, only to relent, essentially self-publish the walkthrough, then reverse themselves and redo the entire process.

Stepping up an analytical level from 'FedEx' - we have several contextually different gameplay categories for quest building blocks: retrieve, deliver, escort, kill, explore, create. With these we can construct a virtually limitless series of quests. As goals by themselves any one block is certainly simplistic and in most cases boring. So what can we do to spice them up?

Primarily, we can explore the situation where the player is informed that the task is of one type, but the quest can branch off to become another (eg. deliver the package - package is damaged - retrieve/create a replacement).

We should also leverage our in-game factions and socio-political structures to bring the player into our world via the quests. We can create multiple possible objectives for any one quest, and allow them to determine which way they would like to handle it based on how they feel about the groups involved, and who they wish to curry favor with. (eg. a simple retrieval for the merchants guild might allow the player to sell the retrieved item to a band of thieves)

We needn't tell the player directly that these choices are there when they sign on to take the quest initially. Indeed that would be fairly lame if the quest patron went ahead and outlined all the ways the player might cross him. We simply need to design for these possibilities, and allow the player to be aware that such choices are possible. Of course, we then take on the responsibility of having to provide for logical alternatives to most of our quests. If two rival merchant guilds were in a bidding war over the last package, it would make sense that they should both be interested in a similar package. Consistency however, can be left up to the quest designer and dealt with on a per-quest basis.

Now, we must include a mechanism to facilitate information gathering; a way for the player to learn how different factions feel about a potential job. We could do this by creating a 'rumor' table specific for each quest, with each rumor flagged by the factions that might know it and weighted with a 'reliability' score. The NPCs of a given faction are given a 'connectedness' stat and thereby have a chance to know their faction's rumor for the quest the PC is on. This chance may be weighted by a 'gather information' skill.

The NPC likely shouldn't share the most reliable rumor they know with just anyone, or even admit to interest. They may withhold it depending on the PC's standing with their faction. Informants may charge a fee or have racial or profession bias as well. They may also effectively act as a two-way courier of information. Should the PC ask a thieves guild member whether they are interested in a ruby goblet he recovered for the local mage, the contact may say yes. But he can also set off a trigger in the quest signifying that the guild is aware that the PC asked. We can then adjust the PC's faction with the guild depending on whether he sells it to them, or he decides against it. Again, we can embed all this logic in the quest itself, leaving only racial bias, faction standings and a 'connectedness' stat on individual NPCs.

A game in and of itself would likely evolve from the player trying to gain favor with a particular network of informants, or enough faction for a particularly well connected NPC to give them the most reliable information. We could even create a mechanism whereby a player could designate another as a 'broker', and we transfer the quest flag to this player broker. The broker could then ask his contact for information relative to the quest on behalf of the person who hired him. This information brokering system would doubtlessly include fees. We should also then adjust the broker's faction standing in kind with the actual questor's decision. If I ask around the merchant's guild to see if they'd like to buy back a stolen wagon, but the wagon turns up in their rival's yard - my faction should decrease, though perhaps not as much as the player's who actually sold the stolen wagon to the rival merchants.

Not all of our rumors have to be direct information either, though the inclusion of deliberate red herrings is of arguable benefit. If we do include them, such red herrings should lead directly back to what the player was searching for to avoid too much frustration.

In creating such elaborate quests it is important that we reward the player for the work they're doing. If one player delivers a stolen goblin idol without surprise, they should not get as much reward as one who had the idol stolen, and had to retrieve it, and then deliver it. If the player chooses to leverage their sneaking ability to retrieve an item, they should be rewarded for sneaking. If they gather rumors to lead them more directly to their goal, they should be rewarded for gathering rumors as well as accomplishing their goal.

Even should players attempt to meta-game these quests via Hint Sites -- we can make the quests so that they'll have to be engaged in the story just to determine who the players are before they can look up where to go for potential rewards. In short order the most effective knowledge that hint sites will contain, would be our in-game back-story: who hates who, and who wants what/ Again, we needn't permanently remove any content on failure - and several possible paths available to the player enrich the story.

It's when our quests fail to engage that they become labeled 'FedEx'. It's when the actions taken are routine, and the possible experience static that they are derided. By allowing the path from Start to Finish to vary based on the player, his abilities, preferences and a dose of dumb luck, we can help to make Quests fun in and of themselves.

Quest examples to follow.
  < / Quests >        
Tuesday, February 17, 2004
  < Power Gamers >        
It may seem as if I'm staunchly opposed to powergaming. Most of my suggestions seem aimed at destroying the systems in which they dominate, and obviating much of the traditional advantage their time investments grant them over the rest of the playerbase. While that is true, I still regard powergamers as the single most important group of participants in any persistent world.

Simply put: Powergamers are always there. A world of only casual gamers is a community strangers. Were it not for powergamers our players community would have little common ground, few 'friends of friends'. Our worlds would feel empty and there would be no-one with the time or drive to organize the more enjoyable elements ( events, raids, etc ).

Yet there exists a palpable derision between the powergamers and the casual gamers and frankly: this is all our own fault. Our systems make the level 10 casual gamer completely worthless to the level 20 powergamer. They can't contribute enough in combat to be worth dragging them around (much less taking a group slot from another potential powergamer), they can't craft any item that the powergamer might need, and they won't have enough money to buy anything a powergamer might want to get rid of. Casual gamers are frustrated by this situation as well, because they can see all the power, wealth, and 'cool stuff' heaped upon the powergamers - and they see how worthless they are, and their skill is, in comparison.

In strictly level-based systems, a Good group will have a level variation of typically no more than 6 levels between highest and lowest power member. Most systems actively discourage larger variations to prevent 'powerleveling'. If we can broaden this range (while still protecting against the undesireable effects of powerlevelling) we can keep casual gamers 'useful' to the powergaming set for longer. Powergamers are not inherently rude, they simply have no desire to suffer organizational downtime waiting for less useful casual gamers to get up to speed. They are shrewdly aware of useful resources though, and we often see a relatively 'newbie' powergamer as the driving force behind a group of higher-level casual players. This demonstrates that when the casual gamer is useful, the powergamer has no inherent distaste for sharing their company. Even though the casual gamers might not have the best gear, use the best spells, or fight with the best tactics, if they are helpful, they will be leveraged. In turn, this includes the casual gamer more in the community and the game.

Player-economic markets similarly segregate the playerbase. If only powergamer crafters can provide the products and services powergamer adventurers need - casual crafters are useless. If only powergamers can find loot that powergamers desire, casual adventurers are useless. By designing items and services that casual gamers can provide to powergamers, we can keep them from being wholly ignored.

I do not hate powergamers, nor do I advocate trying to design systems which will remove them from our persistent worlds. They are absolutely necessary, wonderful, resources and simply need to be more appropriately leveraged. Our status quo is placing undue stress upon the casual gamers. They see the power gamers and realize how much time needs to be invested just to be able to tag along or make an item that might be purchased. The powergamers create the core of your community with their ever-presence, and if the casual gamers feel outcast from this core - they certainly aren't going to stay long.

Again, the relatively flat level hierarchy I advocate isn't a silver bullet, and isn't the only solution. All we need to do is widen the range of usefulness to ease the friction in the playerbase. Again, the hardcore gamers are the core of our communities. Changing our systems so that powergamers can achieve their goals more efficiently by including the casual players is one of the most important things we can do.
  < / Power Gamers >        
Thursday, February 12, 2004
  < Death >        
Nearly every type of game, aside from a narrow subset of puzzles, predicates 'loss' on the game concept of Death. The protagonist never truly 'loses' until he has met his end despite that, increasingly, it is anything but permanent.

In persistent worlds, it is nearly anathema to even raise the possibility of permanent death. The overwhelming majority of players don't want to even consider a persistent game in which death could be final. Indeed, given the realities of modern networking, and hardware/software (in)stability, it is too easy to imagine a situation in which a character will die outside of his own faculties. As surely as we might code a system which 'forgives' permanence for those unfortunate souls who are 'linkdead' at death, we will have players recognize and meta-game this system. Permanent death is not only difficult to code, but it is nearly impossible to implement 'fairly'.

At a fundamental level, players do not like the idea of losing what they've worked to attain, particularly not in 'merely' a video game. In our level-based standard, players spend too much time on their characters to get to The Fun™, and they find it nothing short of maddening should that character's death be seen as too harsh. In my previous essays I've argued against level-based design, against the level treadmill that keeps 'cool' forever out of the reach of the casual player. Yet if we reward achievement in our game in any discrete way, an attachment will certainly form between our player and their characters - and the straining or breaking of that bond will not be acceptable.

So, permanent death aside, how should persistent worlds implement failure? The very concept of fighting an enemy in physical combat nearly necessitates that the loser will almost certainly be killed. Current death systems range across a wide breadth of philosophies, but nearly all carry a unifying concept: the Death Penalty.

Common game-design wisdom suggests that death must carry some penalty, lest it be trivial and by extension obviate any emotional investment in the underlying struggle in the game. Who would care about a character death, if they simply spring back up from the ground again like any mob? This may seem like an exaggerated case, but there is little difference between such a system and the traditional die/reload system behind most off-line CRPGs. The only tangible difference is that play time may be lost between the most recent save-point and the point at which death occurred. There is little in the way of actual 'risk' for the player aside from that timesink of having to 'replay' some part of the game.

Yet does being able to save nearly anywhere in Baldur's Gate remove the desire to win, the tension of the battle, or the need to stay alive? Did NeverwinterNights' death system alter the tension level? In many ways NWN only decreased the penalty for death in those situations where the timesink and xp/gold loss was preferable to reloading a save game. The only difference was that the timesink penalty was replaced with an experience/gold loss penalty, and it kept the tension and excitement of that fatal combat fresh in the mind of the player.

Nevertheless, there is a clear disconnect in player preference between those who feel that death should carry 'some weight' and those who simply want to have fun. Personally, I did not have inherently more fun or feel more tension during combat in Everquest (large death penalty) than I did in Dark Age of Camelot (comparatively small death penalty). The most emotionally intense combat I ever joined in DAoC, was realm-vs-realm combat (timesink death penalty only). Dying in realm-vs-realm combat in DAoC simply resulted in returning to my bind-stone. I lost no items, I lost no experience, I lost nothing but the time it takes to run back out into the field. Yet that was enough. The fight may likely be over by the time I get back; The Fun™ may be gone.

It is The Fun, primarily, that we seek out in our persistent worlds. However we subjectively define Fun, wherever we find it - we aren't playing these games with a gun to our heads after all. For myself and most others, any penalty that removes us from The Fun is penalty enough that death will be avoided at all costs. Emotional weight is lent to any challenge mostly by the players themselves. Losing in and of itself is penalty enough to make most try their best.

An interesting side effect of low-penalty death, evident in DAoC, is the use of death as a means of travel. Travel in Dark Age was slow, even by horse. The penalty even for player-vs-environment combat death was so low, that often people would die simply to be instantly transported back to their bind point. For simulationists this was seen as undesirable, even though those who complained about it had few qualms with leveraging it to get themselves to The Fun™ more quickly.

It is also important to note that nearly all persistent games have spells available that nearly completely obviate the harsh death penalties that players ostensibly risk. Very quickly it becomes the norm, rather than the exception, for players to not suffer our true death penalty but instead this watered-down version. As characters advance in power, it is customary for their resurrection spells to remove more of the penalty. A tragic reversal of common sense and good design that is too often overlooked. Why is it that we punish a group of level-teens in Everquest without a rez spell, so much more harshly than a similar group who simply happens to be a few levels higher? At best, the status quo has some fairly rough edges. In my not so humble opinion, this is simply another case where our level-based standards are outright screwing the little guy, and coddling the powergamer. Is it any surprise that most players leave, and those who stay, stay for a very long time?

Experience penalties and item loss are fairly common death penalties, and being transported back to a 'bind' location is nearly universal. Some games however do, or seek to do, something more clever. As a whole, these fall into the domain of 'death quests' - travels through a purgatory meant to lend more context to the resurrection mechanism. For the most part they act purely as extended time-sinks. They aren't Fun themselves, they just result in more time being spent before the travel back to The Fun can even begin. UO's ghost system was in many ways simply a death quest. Instead of just running back to The Fun, one must quest for a healer and then run back.

Personally I think these can be harmless, and perhaps even atmospheric if done well. However, such death quests tend to be repetitive and drastically increase the time-penalty associated with death. Death Quests, if used at all, should be used instead of an experience or item loss penalty. Possibly, when the death quest is successful, the player should be transported to their corpse, rather than a bind point, outright removing the double-time-penalty. Indeed any downtime due a death results in a net loss of fun. The player will earn less experience, less money, and participate in fewer events or quests when they spend time dealing with death. Their parties will likely also lose time, experience, money, and Fun. Should we necessarily double the timesink punishment, simply for context?

Indeed death quests tend to remove the ability for death to be used as a means of travel. Yet many propose them as a way to make death, in and of itself, part of The Fun. To me, this appears entirely backwards. It might be enjoyable or neat the first few times through, but being separated from one's party is something that is entirely undesirable in persistent games. Any mechanism that exacerbates this organizational downtime will be unpopular, or will result in a distinct lack of grouping. Any mechanism to foreshorten the quest (a rez spell for instance) trivializes the effort we put into making death fun in the first place as most people will get resurrected, and our purgatory will be as wasted as any desolate zone in Everquest. These practical issues all stand beside the philosophical question of: should we make death Fun?

If it was at all possible to make death Fun, would we need a penalty for 'failing' the death quest? Would we need to risk 'double-death', and multiply our penalty again? How far into the logically recursive mess of repeat death could we go? If we agree that death could be Fun and engaging without having to risk 'death' during the quest itself - then what is the point of having an elaborate death mechanism, when dying is not a prerequisite for Fun and engaging content?

Death systems are extremely personal for every team and every project. Still, it is quite important to examine the longview of what our systems are supposed to be doing, and what they actually are doing.
  < / Death >        
Tuesday, February 10, 2004
  < Customer Service >        
I've had the unique (mis)fortune of seeing customer service from nearly every angle. I've been the frustrated customer, the support tech, the developer of the software in question, and the developer of the customer service software tools.

I've seen a nauseating quantity of call centers and customer service departments, and I'd like to think I learned a thing or two about the common pitfalls and roadblocks to effective customer service, particularly as it applies to multiuser software. While the actual phone support centers maintained by persistent world providers are generally of average to exceptional quality, there is definite room for improving the entire experience. I recognize that some providers outsource much if not all of their technical support staff, so my following suggestions may not all be possible.

Primarily, persistent world providers mismanage their different support channels. Why should I reach a different type of staff if I /bug a problem in-game, than I get when I email a problem, or call phone support? Separating these channels results in in-game staff who know nothing about client-issues, email staff who know nothing about in-game trouble tickets, and phone techs who don't have a clear lane to escalate reports of in-game item loss, or TOS violations. We fragment our knowledge base unnecessarily.

For our purposes it would be sufficient to having a rotating call center staffer deal with email inquiries each day. This position should ideally rotate so that no one staffer becomes too specialized, or integral to the process. We wouldn't want service to suffer because that staffer got sick, or as my colleagues like to say 'wins the lottery' (as opposed to the 'gets hit by a bus' standard). Experience has also shown that phone techs are generally eager for diversions from the routine. Email support requests and phone calls will tend to cover much of the same problem domain, and the existing escalation structure would allow us quite easily to filter out 'wish lists' or harassment claims, etc to the appropriate design/GM staff.

Most in-game /bug requests are simple problems of being stuck, or bugged in a known manner with a known fix. Once again, by creating a rotating position for our existing service personnel, we can allow them to handle such mundane requests, without having to interfere with our specialized in-game GM staff (whom, ideally, should be working on enhancing the experience, not unsticking players). They could easily escalate game-specific problems to the appropriate GM staff. For example: a harassment request could be previewed by a standard CS staff, who then initiates logging (if available) and escalates to the GM-staff. However, we shouldn't think we could simply escalate based on customer-defined request category. Our players have been known to submit less-immediate concerns under more 'urgent' categories attempting to 'game' our support queues, and it would be useful to have someone giving all requests a 'once-over' to prevent such abuses. Our GMs would be reserved for the more game-specialized 'judge'-like CS duties: handling griefing, harassment, kill-stealing, item loss, etc.

This is not to suggest there is any magic behind a GM's insight into the game, and that ordinary CS staff couldn't make such decisions. Rather, time has shown the wisdom of maintaining a small group of trusted, monitored individuals as the one's to handle such grey-area determinations.

Indeed the players would doubly benefit from this arrangement. For one, the support queues would be quicker, and access to GMs for urgent issues are more likely to be handled promptly. Secondly, the phone-techs would be intimately aware of all known issues, be it with the client, geometry, compatibility, etc. Any /bug request regarding such issues would result in an immediately response as to whether it's known, how to avoid it, what any workaround might be, and requests for more information as necessary. Our phone staff would have access to their normal logging/tracking system, and would be appropriately trained to handle such requests immediately, with standard procedure. Our development staff then hears about technical issues from a single source, and the design staff hears of balance issues/complaints from a single source.

Even if we consider the case where our phone support is outsourced, we can still realize many of these benefits. Anyone with experience handling outsourced customer support recognizes that we'll have in-house support professionals either way. When we outsource level-n support to a service bureau, our in-house staff is simply reduced. Nevertheless, if we organize them to handle the support channels in a similarly media-agnostic way, we can leverage these processes to improve our quality of support.

We can still combine the support channels through them, maintaining a centralized repository and interface for phone, email, and in-game support requests. Likely, we can even find a service bureau who could handle level-n support for our email as well. We still have a centralized support staff with solid knowledge of our entire problem domain, keeping GM time CS to a minimum. Who knows, we might even be able to free up enough GM-time so that they could concentrate on enriching the customer experience through content.

We likely may have a software disconnect between the reporting system/database our service bureau uses, and the system we use internally to handle in-game and escalated requests. However, since we're likely rolling our own, or customizing something for that specific purpose, it would be trivial to extend it to handle our combined service channels.

I'd like to think persistent world designers have been doing these things, or at least intend to do this on their next projects. However, from my time working volunteer support on UO, and EQ, and dealing with customer support in DAoC and AC - it certainly doesn't seem so. In general, in-game support is slow and less well-organized, and email is all but the red-headed stepchild of customer service. As an industry, we don't even seem to have recognized this support disconnect as a problem, preferring instead to throw staff at the issue (volunteer or paid) and be content to maintain the status quo.

Frankly, as our playerbase becomes ever-larger, this problem is going to grow into quite the costly monster. Never mind the fact that our potential customers are becoming less tolerant of second-rate customer service as monthly fees continue to rise.
  < / Customer Service >        
Thursday, February 05, 2004
  < Communication >        
Typically, creators of persistent worlds communicate very poorly with their playerbase. I believe that Mythic Entertainment, and largely Sanya Thomas personally, have managed a level of communication leagues beyond the status quo. I truly treasured their unmatched system as a player, and respect it even now that I've moved on. In examining why myself and my circle of friends unanimously point to Dark Age of Camelot as being the pinnacle of persistent world customer service, I've naturally formed a bit of an opinion on the matter.

Messageboards are great places for the internet culture to converse. The messages are generally searchable, and conversations can persist easily across time-spans that obviate any other media. They are useful however, only within a peer group. 'Official' messages will always be drowned in the crush of feedback that follows them, and are generally lost amongst the mundane. There are many mechanisms that try to counter such problems: sticky notes, dev trackers, et al - but they are largely ineffective. After all, not all visitors read the faq, so how many will check a devtracker before posting, or keeping up on sticky thread? How many 'sticky' threads can we even have before we choke out discussion by fanning the flames of the "It-fell-off-the-first-page-so-i'll-bump/repost-so-I-can-be-heard" crowd?

The usefulness of messageboards for game designers is in getting feedback from the players. Marketing will talk about 'branding' or 'community' - but as we can see, official messageboards are not unnecessary for community growth or longevity. The drain on designers however, is much larger. Posters consistantly vie for developer attention in a bizarre sort of nascent celebrity chasing. Official posts draw more noise to discussions, as people try to backdoor their pet issues into the developer's line of sight. The bulk of messages is a recycling of the pet peeve du juor. Whether it be a nasty bug or the playerbase's latest balance issue emergency. This is not to discount the validity of their claims, but rather to point out that the issue being repeatedly raised is not at all helpful.

Quite simply, only one or two customer service reps need to be reading message boards to locate player concerns, and even if you have 'official' boards, to be thorough, they'll end up reading community boards anyway. The reason developers feel the need to post to the message boards as it is, is because they realize their other channels generally suck. Messageboard posting just isn't a good default solution to that problem. 'Official' messageboards are an even worse solution, since their very existence gives posters the expectation that they will receive personalized feedback. On community boards, there is little to no expectation that the developers are reading (though they nearly all do), let alone that they'll respond.

The information that an official source needs to make available simply doesn't benefit from being displayed in a forum setting. It need to be prominent, clear, and current. Garnering feedback is important, but it needn't happen in the very same space. Further, official comments are of interest to a larger crowd than those happening across a single thread.

So what is it that developers need to convey to players, and why do players post the same issues repeatedly? I know many developers cringe or react poorly to insinuations that they should justify their decision-making process, and indeed I'm not even calling for that. What I am, is calling for more transparency on what decisions actually have been made. It is beyond the scope of communication, and into the realm of PR whether you decide to discuss your decision-making processes with players.

Generally, posts repeatedly raise the issues of: known bugs, known balance issues, and wish lists. They tend to be repetitive, because there is largely no acknowledgement by developers that they've been heard. Message boards are entirely ineffective at getting this message out. So what we need, is a few pages that are official, concise, and current.

There should be a 'Known Bugs' page, that lists all known bugs in a game. Exploits may have details omitted to prevent widespread abuse, but they should always be acknowledged. Indicating the number of accounts banned for taking advantage of a particular exploit might not be a bad idea at all. These known issues should include all relevant details, a priority assessment, and an ETA, if any. Yes, players may disagree with your deciding their pet issue isn't a priority. Yet if you never say what is a priority, everyone by default assumes that their issue is not..

There should be a 'New Features' page. Even if we have no desire or ability to implement a feature, if players keep asking for it, it behooves us to post your official position. Simply list the idea, how we feel about it, and if any, a priority, and what state it's in: undesireable, impossible, expansion-only, brainstorming, designing, coding, testing.

There should be a 'Balance Issues' page. We should maintain a database of common balance complaints. We should clearly state our position, its priority, and state: investigating, coding, testing.

There should be a 'What's Cooking' page. This page should be a collection of bugs, balance issues, and features that are currently being worked on. If an exploit takes development time away from player housing, the players will be much more understanding if we tell them so in no uncertain terms. Furthermore, players will be able to immediately see all the problems that they might never experience that are deemed more important than their own and may be more understanding. Perhaps not, but the disgruntled assume we're doing nothing as things stand, so it can only be an improvement. Indeed much of Mythic's player-relations success has centered around their excellent Camelot Herald, which is (at least initially) simply a well-written, up-to-date, and honest 'What's Cooking' page.

Another channel of communication that's largely ignored by developers, is pre-game communication. After login, before play: right about when they're clicking off on our silly EULA. Many developers have rightly recognized that putting a narrative update in front of patch notes is a net positive communication method. Many players simply don't care to read message boards or fan sites, and this is likely the only update they're exposed to. However, developers haven't been taking advantage of leveraging this time to prompt player feedback.

Why not ask a player to fill out a 'how are we doing?' questionnaire on their monthly anniversary date? So long as it's optional players won't be bothered by it. Any chance to take the pulse of the average player, who is not at all represented by the messageboard-posting community, is worth the effort. Naturally though, we would have to wade through all that information, and hopefully reply to it when necessary. Yet I believe it would be worth the effort; a pound of prevention and all that. By keeping the prompt relative to anniversary-date, and not asking all players at the same time, the stream of feedback is kept much more manageable.

Indeed small pre-game polls would be similarly useful. They have been implemented by several developers, but in my opinion have been vastly under-utilized, and I don't understand why. But why not ask the players whether they'd rather have a new dungeon or a new class? Whether they'd rather have a bug fixed or a class rebalanced? Or even how they would prioritize a given bug? The polls should certainly be entirely optional, and time sensitive (so a player who's been away isn't bombarded with a dozen questions). In fact, players should be allowed to opt-out of the polling system entirely, so that they don't even see new poll questions.

Many of these things the exemplary Mythic player relations team does not do. Yet it is noteworthy how much more respect they have from their playerbase even with the modest bit they do differently than their competitors. Surely my advocation sounds like an staggeringly expansive database system, and may be a bit much in its entirety. Yet I think we can see a large measure of improvement simply through the adaptation of a few core ideas. Essentially messageboard-going player-relations teammembers are essentially already maintaining lists of known bugs, priorities, standpoints on balance issues, and items in development. The problem is that they are unnecessarily duplicating their efforts by not maintaining a centralized point of reference for these items.

Still, without keeping all our communication honest, by not giving realistic assessments of where the work stands - all communication with our players is useless. Similarly, if prompted feedback is ignored, if the player never gets so much as a 'thanks for your time' email in return - gathering it in the first place is pointless. Rational players will assume we're working on something after all, and if the pages are never (or even merely rarely) updated; they won't trust them at all. The value will be lost.
  < / Communication >        
  < Rewards: Unrecognized >        
Perhaps 'unrecognized' is a better term than intrinsic. Despite the fact that all virtual rewards are by definition intangible - it makes a more clear distinction between the discrete rewards, and those the game does not officially recognize in any way.

Being subjective, players will find unrecognized rewards in behaviors you can not possibly predict. Occassionally we can forsee the fun people will find in roleplaying, or that allowing players to free-form place furniture may well result in interior decoration. Although I don't know that McQuaid and co. ever sat down and thought about how much mindless joy myself and a friend would derive from simply jumping through their game world. We would indulge in friendly competition to find the 'coolest' jumps available in the game: cutting corners as sharply as possible to shave travel time or simply plummeting from the highest places in the game.

Herein we can see the simple difference between the actions that garner these emergent rewards in our games. Some of these actions are expressly desireable, while others are purely tangential. Would the game world benefit if we coded a bonus for the 'coolness' of a jump? No, most certainly it wouldn't. Would the game world benefit if we coded a trade/rating-system for interior decoration? Perhaps. Would a game benefit if simple gambling items (cards,chess,etc) had an express recognition system? Almost certainly.

Many players discount entirely game actions that do not result in direct rewards. UO's gameboard set for example did not enforce any particular game rules, did not provide for a secure 'bet' mechanism, and did not have any method of tracking win/loss records. By and large, these sets were ignored by the playerbase. The reward from winning a game of chess in UO was entirely unrecognized even though there was a very real satisfaction for the many players who did enjoy it.

UO was undoubtedly a better experience for having such a board game in its gamespace. Yet what if the chess set in UO had slots for secure betting? What if mechanisms existed that allowed an event organizer to begin a system-managed chess tournament, complete with brackets and standings? Surely it would have meant more work - yet it would have also drawn more players to the content.

Similarly with crafting professions. What if crafters and their clients had an ebay-styled rating system that allowed them to define how 'good' a particular job was done, or whether the client stiffed the creator? Surely crafters would seek out such a method to not only establish their proficiency, but their professionalism as well. Clients would no doubt be much relieved at having some sort of independent verification of the trustworthiness of any particular player-merchant.

Such a rating system could be coded similarly to the karma system proposed a while back. There is no reason that we couldn't extend such meta-ratings to other areas as well. Players love having official 'roleplay' tags or servers, even though adherence to such a designation is impossible to enforce through code. Yet why couldn't we have a meta-game rating system whereby roleplayers could have a more appropriate expectation when they were about to deal with other players. Obviously not every subjective means by which players rate one another could, or should, be coded, but adding one or two would not be too time-consuming or difficult. After all, once we have coded one such system, any additional rating mechanism is less expensive to add, as the code and interface items already exist. We merely need to add more dataspace to the character record, and a means to visual convey a character's standing.

There is little reason we couldn't apply titles, awards or effects to the player to recognize their previously unrecognized achievements. Perhaps we grant only those with a high roleplaying score access to some advanced emotes. Why not reward those with a high merchant rating "chamber of commerce"-style awards, that they could display, for being a beacon of fair trade?

Many times questing itself is an unrecognized reward. A character may be granted items, money, and experience for suceeding, yet the player may as well have earned that reward from regular fighting. The story and the context to a quest may be a reward in and of itself - but if we want more players to take advantage of our questing system, we can easily do so by more prominently recognizing such action. After all, the more players that use our questing content, the more effective that content has become.

We generally don't want quest rewards to be outright superior to player-crafted items, but perhaps we could grant questors augmentations to whatever items they desire. Perhaps a thankful NPC imbues a character's weapon with a visual effect that confers a slight bonus? The PC could be granted a desireable (nontransferable) dye to be applied to whatever they wished. An NPC guild may even grant the PC a tatoo signifying acceptance.

At the very least we could maintain a record of quests completed, as a list of meaningful deeds performed by the character. Perhaps the NPCs in town merely use such titles in their dealings with the player, depending on their local notoriety.

Naturally, the moment we code such systems, 'unrecognized' rewards leave this categorization. It seems all I'm really calling for, is an expansion the range of behaviors that we discretely reward and punish. A large part of the playerbase merely camps and grinds, because nothing else has any lasting effect on the character. Perhaps we can draw them to the other Content, simply by showing them these other means of recognition and accomplishment.
  < / Rewards: Unrecognized >        
Tuesday, February 03, 2004
  < Rewards >        
Rewards in persistent games must primarily be recognized to all be merely potential rewards. This is not to suggest the idea that some players will never attain them, but rather, that some players will never want them. Players will find their own reward for the actions they take, and often it will not be in line with what we might expect. Some players of UO don't care to slay dragons, or amass wealth. Some players of Dark Age of Camelot have no desire to garner accolades from defending their homelands.

How then do we decide what rewards are appropriate, what rewards are desireable, and how much of any reward is necessary to make it feel achieveable, without making it feel cheap? Indeed managing rewards goes a long way toward managing the game's economy, and that is a terrible hornet's nest I don't think I'll stir any time soon. Indeed, there are entire blogs, papers, and chapters of well-received books devoted to the topic, and they do a very good job of covering the problems, theories, and heuristics.

What concerns me mostly, is types of rewards and what they are intended to symbolize. Generally, there are what I term 'discrete' rewards, and 'intrinsic' rewards. Discrete rewards are those which persistently follow the user, and are capable of being displayed. Intrinsic rewards are those which give the player personal joy, but that we can/do not do generally track, and may not be capable of tracking.

Discrete rewards are things like titles, experience points, treasure, loot, housing, abilities. Intrinsic rewards are more akin to the satisfaction of having a well-styled character, a well-decorated home, or the pride in having been to every town in a persistent world.

Many players, understandably, seize on discrete rewards. Much like any other hobbyist collectors, they derive satisfaction from their proveable achievements. They may have travelled across the entire game world to become 100th level, but it is the destination, and not particularly the journey that fuels their pride. However, not all players sieze on the same discrete rewards. Some will take great interest in questing or collecting, while others may care only about their total wealth and naught else.

Other players find a more vague joy in playing the game in unexpected ways. They weave stories with other players, they struggle to immerse themselves in their alternate identity, they organize fairs, taverns or plays, some simply enjoy things like leaping from the highest places in the game, and plummeting back to the game-world surface. Intrinsic rewards I will leave to another post.

More than anything, what bothers me about persistent world design is the feeling that any discrete accomplishment must require a significant investment of time and energy (primarily time). Even helping to slay a powerful beast or doing something as heroic as rescuing a damsel necessitate being of the highest levels of in-game power. This certainly places a restriction on the availability of the rewards, and thusly keeps them artificially valuable.

They stay valuable only artificially as, inevitably, whatever we feel is a near impossible task today, and deserving of a truly great reward, will become a more common occurrence tomorrow - and the reward considered more 'pedestrian'. There is no avoiding this -- not with our level-based, grind-centric, status quo. So then, why do we maintain this standard?

There are a few discrete reward assumptions that have been made as far back as I can remember, and truly I feel they are pigeonholing us into the same mistakes over and over.Having the trades drive adventuring is a powerful tool to facilitate a player-creation/player-consumption loop of Content. Yet all of the games that have attempted such systems (to the extent of my personal experience) have failed in achieving this goal. Simply, trades are always the red-headed stepchild of the gamespace. The adventuring types have no particular reason to seek out such contracts, because it carries no rewards they can't gather more quickly, more effectively, more reliably elsewhere.

Consider however, if adventuring was not the primary moneymaker. I understand that many people likely just muttered 'crackpot' under their proverbial breath. I'm also certain I'll have no chance to ever convince most gamers to even try a game with that kind of assumption. It has not been done before (again, to my knowledge), and simply is counter-intuitive to the philosophy that killing monsters is inherently the most fun. Most people play these games to do heroic things, so why would we want to devalue the eradication of an orcish horde in favor of ferrying crafters from town to town?

I argue that the standard for fighting monsters and questing in persistent worlds is pretty far from Heroic as they stand. The orcish horde isn't all that meaningful to fight: the battle-lines never moves either way, no victory is persistent, and no defeat is truly possible. The enemy never advances, never takes spoils - they simply stand their ground, ready for their eventual defeat and resurrection. How much pride can we take in having killed such truly ineffectual foes?

If every conflict (or avoidance for that matter) is done in the persuit of a more concrete goal (transporting goods, collecting resources), then there is more weight to our actions. The orcs may come back tomorrow, but our shipment arrived today. Also, if adventuring is not a guaranteed way to generate wealth, then we can more carefully balance our economy. Only through harvest and production does the market generate wealth - and most players will find their 'fun' is often rewarded by wealth, but isn't automatically so.

It has been a standard since the birth of pen-and-paper gaming that defeating a stronger monster should always result in the accumulation of ever-more treasure. After all, if not for the reward, why would we risk such a fight? The only thing we do in these games is destroy our enemies. We have no higher objectives or motives, nothing worth risking our necks for, other than the treasure itself. So we designers give more gold to a 5th level foozle than a first level foozle. By allowing foozels to manufacture wealth like this, we have created more than the conceptual leaky faucet for resources -- we're opening the faucet ever-further as the playerbase advances.

Lastly, it is assumed that rewards that don't confer greater aptitude at the rewarded activity will be rejected. If killing monsters does not make one better at killing future monsters, players will reject the activity. This process traps us into level-grind logic. The only way we allow people to create 'better' stuff, to fight 'bigger/stronger/cooler' monsters -- is to reward them with this capability only after they've made piles of worthless stuff, slain stacks of less-fun monsters.

Why is that? The origins in pen and paper gaming were in a different setting, in which a carefully crafted story gave context to every goblin slain - where no other adventurers were in contention with our own party to rescue the princess or retrieve the golden widget. In persistent worlds, this has nearly never been the case. Furthermore, in what other storytelling medium has the hero ever had to pay his dues in anything more than a montage or off-camera transition? The primary stories in our persisten world revolve around training, and training simply isn't fun. Over and over, we're training for the end game; trying to amass enough power to finally be able to see the cool stuff, to do great deeds.

It seems as if the designers at Blizzard have caught onto the level-grind fallacy. Their quest system seems to acknowledge that killing monsters without context is nowhere near as fun. Even though they weren't about to abandon tried-and-true spawn points and static encounters, they're doing much to give context to the play at all levels. Truly, if questing is more effective for advancement than random camping, everyone who cares to advance will quest. And we all know quests will have walkthroughs in short order.

Perhaps at Blizzard they realize that a walkthrough for a quest is merely the same as Cliff's Notes for a novel. They may help everyone power through To Kill a Mockingbird, but they don't at all cheapen the experience for those who truly engage the content. Quest walkthroughs themselves are not even functionally different than the standard player's optimal advancement heuristics: these levels hunt here, those levels hunt there.

Rewarding players for using their sword, with more skill at the sword only goes so far. Indeed the vast majority of people who bought EQ, the staggering majority of even those who paid for more than the 'free' month - didn't stick around for even three months. This is not to indict EQ - it does very well what most others are trying to do. The adoption statistics are the same across the breadth and depth of our genre. Wherever we find a level-grind, we find those who take to it obsessively, and the majority who catch on fairly quickly.

Yet even for those who take to the level-grind, we recognize there has to be something more at the top of the ladder. Something else to do. The main dread of the level designer, the thought that has us clutching our level-grinds as the proverbial security blankets they are, is what if our end-game isn't fun? If our level-grind is short, everyone will see our end-game for what it is. Many may not like it. To that I say: What about all the players you've already alienated by mandating the context-free grind in the first place? Are they not vastly more numerous than those that stay? And don't those that do grind often stay for the end-game? Why require players to grind, why insist on rewarding them with advancement in and of itself, when eventually everyone will be as powerful anyway?

We do not keep power rare by mandating a time investment. We shrink our community by retaining only the powerful. Are those strong enough to slay dragons any more rare in EQ for our efforts? No. The playerbase remains nearly uniformly powerful. We learn to balance at the top end by creating encounters that require the concerted effort of many players - we learn that the pattern of advancement does not have to go on forever. And indeed, to present a recurring challenge, it can not.

Advancement as a reward is not a bad thing, provided it has context -- provided that there is another reward for defeating our enemies than simply gaining power. In recognizing that there must be another reward for adventuring itself, it should no longer seem so extreme to consider a design where fighting that stream of kath hounds is made only rewarding when the player seeks out context. Keep in mind this does not discount dungeon-delving or fighting monsters from having rewards in and of itself. We should still retain activities such as dungeon raids, and indeed should make them more common and accessible to the average player. What this does discount, is merely the endless fight/heal/wait loops that we require of our players before they are strong enough to make use of such content.
  < / Rewards >        

Powered by Blogger