End Game: Progression vs Buffet

So you’re an MMO developer and your players have hit max level. Now what?

We’ve seen a lot of MMORPGs out there, so you would think we’d have seen a lot of different ways to handle end-game by now. But end game across a large number of these games is dominated by instanced dungeons with bosses designed for large groups. The main difference between these games is just how many bosses they have and the order in which you’re allowed to fight them. This blog is about restrictions on how and when you’re allowed to experience end-game.

Some games let players do any of their dungeons in any order. For example, Star Wars Galaxies let players attempt any of their “heroic” instances in any order as long as they completed a pre-quest for each one and are within 20 levels of max level. This makes SWG’s end game feel more like a all-you-can-eat buffet. Sure, there’s not a lot of selection, but they can eat whatever they want and skip whatever they want while still watching their character get stronger from each run.

Other games handle end game like Star Trek Online. If you want to do STO’s “raid episode” called “The Cure” then you must first successfully complete the “Infected” instance. Like being served a first-course and then a second course at a nice restaurant, it’s a linear progression. Not only does all of STO’s end game instances have to be completed in order, but players are absolutely required to be in a full-group of max level characters in order to be allowed to enter for each one.

For the first year, Lord of the Rings Online took a more ‘buffet’ style to their end-game instances. The small group instances of Annuminas, the 12-man raid called the Rift, and the 24 man zerg-blitz known as Helegrod could be completed and farmed in any order. But like a buffet, not everyone bothers trying everything. Players knew what they wanted and made a bee-line straight for it: the best armor they could possibly get. Developers hate watching content “go to waste” so they decided to change things for the first expansion.

The first expansion to Lord of the Rings Online saw a more linear progression style of end-game. Sure, there was some freedom to chose the order in which they farmed Moria’s instances, but everyone farming radiance armor knew they had to polish their plate clean before they were allowed to experience the main course: raiding. Attempting to enter the new raid without the previous armor set would give extreme debuffs to stats and cause you to lose control of your character every few seconds. And if players wanted to move on to the second raid, they knew they needed to win the roll for the helm and shoulders from the first raid.

By the time Lotro’s first expansion was sun-setting, players were fuming about the forced progression. Some new small-group instances with new armor were added as another option to gear up for raiding, but it was still understood that players had to finish farming all their non-raids before being allowed to raid. So along came the second expansion for Lotro, Mirkwood. Suddenly the system changed again. All non-raids in the expansion drop the same tokens which can be saved up to buy the gear required for raiding. It still didn’t let players run right into the new raid, but players are now allowed to farm just one instance dozens of times, if they so choose, as their means of gearing up. Ironically, less content went to waste in their old buffet-style end-game. Players like to check out the different rewards and challenges in instances. When the reward is exactly the same, but the instance is longer and more difficult, it’s harder to get players to even try it.

Lotro isn’t the only MMO to experiment with different models. World of Warcraft has shifted over the years as well as many other MMOs. Sometimes the linear progression is disguised in different ways, but players always see through that ruse. Even within a raid cluster, players will figure out if they need to farm boss X for the stat bonus’s that make beating boss Y realistically possible. There’s no getting around it, developers have to make a choice between on model or the other.

There are clear disadvantages to a linear progression model of end-game because it forces players to swallow content they don’t like. If something which is required is overly difficult, long, or frustrating, it can kill the excitement players have for the game. For many of them, it was already too much to ask that they be forced to level to max level before seeing end-game.

On the other hand, the buffet model lets a large amount of content go to waste once it’s become out dated. Players tend to seek out the armor that will make them the most powerful and they want to skip the content that won’t. There’s the worry that once a player has everything they want, they’ll quit the game. It’s awfully hard to justify spending months making something that will be ignored or forgotten by the player base within week’s of it’s release.

Ultimately, I say there is one ideal choice: the buffet. That being said, there are good buffets and bad buffets. The bad buffets are seen when a linear-progression style reward system is adapted to a buffet model. If every single raid offers you a different armor set, well you can only wear one at a time so you’ll probably only care about the raid that offers the best one. If every single raid offers you just one piece of an armor set, you’re forced to care about all the raids. Armor sets are just not a good reward mechanic for a buffet style end game because armor is all about replacing previous sets. It’s linear by design.

By contrast, stat bonus’s do not replace each other. Were a raid to ever give you +100 to health for completion, that bonus would stack with the raid that gave you +50 to health for completion. Small cumulative rewards for instances give players a reason to care about all the content to some degree, but don’t force players into feeling like any raid or set of raids is “mandatory”.

In the linear-course meal vs buffet analogy, we could argue that food is always going to waste. Either you have a buffet where people skip the more unsavory dishes, or you have a multi-course meal where only 1% of the customers make it to the main course. In both situations food goes to waste because customers get full. The question is, are your customers going to get full because they’re full and satisfied or are they quitting because they’re full of rage?

Published by


Suzina is a 27 year old who usally plays the same MMOs as her husband. Games played: UO, EQ2, FFXI, SWG, LOTRO.

7 thoughts on “End Game: Progression vs Buffet”

  1. I totally agree. I prefer the “buffet” as well – even if it means more work for developers.

  2. I dunno, I think the way WoW currently does it works kinda well (I would though) with the random dungeon option (that is better than picking specific dungeons) and the weekly raid quest and the achievement system. There’s often a group going for a classic or a TBC raid just for the achievement (not that you end up seeing how that raid is supposed to be played).

    And I always felt that end-game was a buffet anyway, as there are tons of different things to do.

  3. This is why I miss ShadowBane. There were multiple types of end-game content. Sieges were the most popular, following by hunting for elusive Vorgrim armor, which spawned on 7 creatures in the entire world (at a time, and there was one of each creature), which reset once every 30 minutes when killed, and had a 5% chance of dropping. Imagine that when hunted by hundreds of people at a time. Then came the R8 mobs–one had NEVER been killed, others took hours; all were worth the time and effort. Owning and building a city was another.

    This was one game where there were countless ways to occupy your time. Hundreds of character builds. Many FOTMs at any given time. Despite its bugs, ShadowBane was the best game I’ve ever played. Hands down.

  4. I’d say there are actually a great many more approaches to this than just buffet vs. linear, as long as you are willing to consider paradigms (can’t type that without picturing Dilbert’s boss) that vary from the EQ/WoW/etc. formula.

    The 1st question, really, amounts to the question of whether or not this game is intended for mass appeal or a niche audience. Although it’s terribly expensive to create a good MMO, it is less expensive than it used to be, and the consumer base has grown such that I think a lot of the recent crop of MMOs have a made a mistake in shooting for a mass-appeal, WoW-killer type of game.

    The standard, currently accepted MMO experience is accessible to the most people, yet leaves a lot to be desired for many of those same people, which the question of buffet vs. linear endgame helps to illustrate.

    The linear model, for example, allows to developer to create a rich (hopefully) storyline. The buffet model eschews story in favor of facilitating player progression. Each of those models is compelling to different types of players.

    And both of those models rely on the current MMO standard of a level-based system that ends in an arbitrary cap, at which time the game changes dramatically. Players seeking to advance and improve their characters cease to care about experience and instead are directed to focus on equipment, if they haven’t already.

    I know this is basically sacrilege in most circles, but level-based systems are just not necessary in an MMO. Levels, gained by acquiring experience points which then allowed the player to choose new abilities, was really a compromise predicated by the need to balance gameplay and realism in paper and pencil games, where it would have been too time-consuming to do more realistic incremental calculations.

    If you’re thinking I’m referring to skill-based systems a la UO/Darkfall/etc., and then don’t they have a “skill” cap, and doesn’t that leave you in the same boat, you’re partly right, except that you can have a soft-cap (this is getting kinda long and a little left of topic, so I’m not going to go into detail on that). Even better, character progression should ideally be driven by player progression.

    What I mean by player progression, is the player, through learning game mechanics, or discovering a new mechanic, increases his characters effectiveness without any traditional metric being incremented. A quick example of this kind of advancement, would be figuring out a new spec and/or combo in WoW that is more effective than the currently accepted cookie cutter response for a given situation. This type of advancement is great for the developer, as it often requires no additional work (provided it isn’t game-breaking and requires the nerf bat).

    Another example of this type, one that requires a lot more developer involvement, would be seeding the environment with the means to create new variations…

    Lol, I should really reel this in and finish.

    I’d say my answer to linear vs buffet for endgame character progression would be some of both, if necessary, but I really think that if as a developer you have to ask that question, then you are trying to make another copy of a somewhat tired endgame mechanic.

    1. Personally I don’t think either levels or skill-based system is necessary. You could just have everyone be more or less equal from the very beginning. You could still have some sort of progression though, and solve the “leveling is tutorial” problem too:

      Replace all the “kill ten rats” quests with one quest/chain per class-ability. Give the elementary set of each class from a few simple tasks at character creation, then let people work rest of them in whatever order they feel like. You could make them so the more important stuff is easier, the really great stuff (the “high-level skills”) takes longer chains that essentially require some of the earlier stuff.. sort of soft-gating.

      Cool things:
      1. still giving people stuff gradually, which helps with the learning curve, as you don’t need to figure out 40 skills when you start your character
      2. can put more thought in the quests when you don’t need such a huge amount.. plus you got an excuse to theme them for the skill they award
      3. you could randomize some aspects like NPCs/dungeons/mobs (pick alternatives from a set list or something) creating unique path for each character

      Especially with the last, one could go crazy: it shouldn’t be hard to tag 20 NPCs as potentially acting as nodes for a chain, giving the same/similar tasks. You could also pre-randomize the whole path for a given character so that he’ll be visiting each area before having every skill, yet do it in different order every time, for different reasons, doing slightly (or drastically) different things, minimizing wasted content, if that is desired necessary; though with randomization replay would be more interesting, optimal paths would not be set in stone, and being “viable” from day one, a new character wouldn’t be so daunting tasks to undertake (especially as leveling alts is boring: you can’t even chat with your friends, as you’ll be flooded with requests to log on your main to come play an instance ‘cos they have a group that “only” lacks a healer/tonk/whatever and I’m too nice to never say “no”; which is why I don’t play anything currently, having tried “casual” gaming several times, and always somehow ending up with a 5 days/week raid schedule on top of the 5-10 dungeons/day).

      Arbitrary order wouldn’t need to be a problem, if every character past tutorial would be approximately equal in power. Yet your character would grow in flexibility (which in practice equals power without bigger stats; no need to necessarily scale mobs, yet in PvP and tricky situations you’d have more and more options _actually feeling more powerful_) every time you did another quest for another skill (though you could still go run dungeons, or practice PvP or whatever on the side).

  5. I’d just like gameplay that was so innately enjoyable it was self-sustaining without any progression or reward required.

    What I’m looking for is the equivalent of knitting, whittling, throwing a frisbee, flying a kite, walking the dog (four-legged and yoyo version), bouncing a tennis ball against a high, flat wall. The kind of activity that you do just because it’s fun.

    Only I want to do something that I can’t do in the physical world. Something that involves magic or mystery. My favorite end-game activities so far are sailing my boat in Vanguard, flying my flying mount in Vanguard, decorating my houses in EQ2 and organizing all my stuff in any game.

    After that comes going to places where things used to be too difficult to explore and sightseeing. If the sightseeing involves beating up a lot of the still-aggressive but now not very threatening locals and relieving them of their valuables, so much the sweeter.

    What it really comes down to is the age-old (well decade-and-a-half-old) choice between a virtual world and a game. I want a world for my character to live in. I don’t want to have to build the entire thing for myself, so I don’t want a full-on sandbox like Wurm Online, but neither do I want a “game” with lots of parameters for success and failure.

    Interesting and surprising thing to me is that three of the big MMOs in development, FFXIV, GW2 and SW:TOR, all seem to be hanging on to the Virtual World end of the rope to some degree. I did think we might be approaching a time when MMOs turned into all game and no world, but it seems we have at least another generation to come yet.

Comments are closed.