Tobold's Blog
Friday, November 19, 2010
How I would design raiding

The way raiding works in World of Warcraft, like other features of that game, is the result of the iterative approach to game design of Blizzard. The developers started with something, in part based on previous games, and refined, balanced, and modified it in major content patches and expansions. That is a solid approach, but it is timid, and unlikely to lead to major changes. In this post I would like to shine a light on the shortcomings of the current system, and give an outline on how I would design raiding. That is of course a pure thought exercise, I doubt my ideas will ever be realized, but maybe they can serve as food for thought.

Everybody agrees that the key problem of raiding is its “difficulty”, which has been endlessly discussed all over forums and the blogosphere. While everybody perceives “difficulty” as a problem, there is strong disagreement about whether raiding is “too easy” or “too hard” and not accessible enough; and of course that discussion flares up again with every modification making raiding “easier” or “harder”. So instead of making a controversial judgment call on raid difficulty here, I’d like to point out some less controversial facts about raid difficulty: I think everybody would agree that raiding is more difficult than the other content of the game, and that during all the years of WoW there was usually a big step change upwards in difficulty between whatever you could do to prepare for your first raid, and the first raid itself.

I might get more discussion with another observation: In Wrath of the Lich King, raid difficulty did not go up from Naxxramas to Icecrown Citadel. That is somewhat unintuitive, because of course a raid group in ICC gear would have a rather easy time today on many encounters in Naxxramas, like Patchwerk; but that is caused by gear, and the raid encounters by themselves are not much more complicated in ICC than in Naxxramas (with the possible exception of the last bosses of ICC). I’ve even seen raid group able to do the first ICC bosses wipe on some of the more complicated encounters of Naxxramas, in spite of their better gear. The “difficulty” of raid encounters in Wrath of the Lich King is mostly a function of how complicated the special attacks of the bosses are, and how well coordinated a raid team needs to be to avoid the huge damage done by these special attacks. This difficulty is specific to each individual raid encounter, and being for example able to beat Lord Marrowgar in ICC does not mean that you are also able to do “the dance” at Heigan the Unclean in Naxxramas.

It also has to be noted that in general it is more difficult for players to deal with special attacks that require coordination (e.g. “no player should stand closer than X feet from another player”, or “players with a random positive charge need to stand away from players with a random negative charge”), than those which only require the player to watch whether there is a spot of fire where he is standing.

Based on these observations, here is how I would design raiding:
  1. The overall difficulty of the first raid dungeon in terms of gear requirement and raid group coordination should be only a small step up from the hardest 5-man heroic dungeon of the initial release of an expansion.
  2. The encounters in the first raid dungeon should not be too complicated. I’m not calling for only tank’n’spank encounters here, but the boss special abilities in the first raid dungeon should be more about each individual player having to react to something, without adding additional requirements of coordination between players.
  3. The second and subsequent raid dungeons of an expansion I would then design to be noticeably more difficult in all of those aspects. That is they should require more damage output, damage mitigation, and healing, regardless of whether that better performance is reached by better gear or simply by playing better. And the encounters should become more and more complicated with each subsequent raid dungeon, so that the encounters in the last one would really be very difficult and require a lot of coordination between all players of the raid.
At this point some people will complain that “Burning Crusade proved that players hate raid progression”, which is a common misconception. The problems of Burning Crusade were mostly the too high difficulty of already the first raid dungeon, Karazhan, and a system that needed too many repetitions of the same raid before players had enough gear to be able to progress to the next raid. Players did not hate “raid progression”, but the fact that many of them couldn’t even get into raiding at all, and others “got stuck” and didn’t progress at all any more.

A relatively easy first raid dungeon would ascertain that “raiding” as an activity would be accessible to most players, even in pickup groups. That first raid dungeon would also train players in fundamental raiding skills like “don’t stand in the fire”. And it would provide players with sufficient gear to be able to progress towards the next raid dungeon, where they would be able to train more advanced raiding skills. Players unable or unwilling to learn raiding skills would significantly “slow down” in raid progress at some level of difficulty, but that is totally okay. I’d even say that it is a necessary feature to encourage players to learn, because if you can faceroll all the way to the last boss of the last raid dungeon there is no incentive to play better.

What should be avoided is players getting stuck simply through gear requirements, like it happened in Burning Crusade. That is actually not so hard, since Wrath of the Lich King already provided the perfect solution: Improving gear with a system of tokens. I would design rewards in a way that raiding would give more tokens than running heroics, to preserve the same risk vs. reward ratio, or even a bit more to compensate for the bigger organizational hurdle. The token system is useful to make sure that there is a constant improvement and progress, to make sure that it is really skill and not gear that determines how far and fast a player can progress. What I wouldn't want is token rewards from heroics being so plentiful that they make the first raid dungeons completely obsolete, as it is now the case. Thus a "pickup raid", while harder, and harder to organize, should give better token rewards than farming heroics.

What I think my raid design would achieve is the best of both worlds: Accessibility of raiding to most players, even casual ones, and faster progress and status for the more hardcore players. In such a system an automated Raid Finder functionality makes perfect sense, especially if like the current system for the Icecrown heroics it makes sure that you can’t sign up for raids you are completely undergeared for. Pickup groups *should* be able to do the first raid dungeons, even if they lack the coordination necessary for the later ones. “Pickup groups can’t do every raid” is a much better situation than “pickup groups can’t do any raid”.

Of course no raid system will completely eliminate the whining of the two extreme groups: The players who want everything handed to them on a silver platter without having to exert any effort, and the players who want all of raiding to be reserved for an exclusive elite. But in spite of the amount of noise these extreme groups often made, they are actually only small minorities in the overall population of World of Warcraft, their size often exaggerated by their opponents from the other extreme (“All raiders are elitist jerks” vs. “99% of players are morons & slackers that only want welfare epics”). It is the eternal naysayers who often create the impression that stagnation would be the best option. But a system with a low barrier of entry and rising difficulty level in the end serves the needs of everybody better than the current “one size fits all” raid difficulty model with its endless discussion about what exactly that one size should be. Players generally not only understand skill-based progress much better than some people claim, but they are actually attracted by it and encouraged to up their game.
Extra Credits recently covered the same topic on a more generic level.
I started to comment, but then it grew too long. You can read my comment here.
I've always thought why can't there be a "lore" group (my name for it). Where basically it is tuned to a 5man heroic. This allows players like me to go through the story instances (like Lich King etc). The item drops can be slightly below 10man stuff. Then for people who want, they can then start their 10/25man raiding career and go for the big epics etc. For people like me, I'll just continue on doing my own thing, starting alts, playing with friends in other areas, crafting, etc etc
I would then be satisfied that I've seen the "ending" story for that expansion.
That sounds good to me. But it might disappoint the group of players that, without progressing in skill, enjoy progressing simply because of gear upgrades. I think they are an important reason for wow's success. There's always a way to get better in wow without upping your (arrgh...gotta...resist...using*puff*...the...worrrd...can't...) skill.

@Amphigory: good idea! Blizzard has already toned down required group size from 40 to 10, so maybe this will be implemented, too. Generally speaking, I find it amazing how they've come from offering only random drops to offering a whole array of different ways to obtain gear. Amazing differentiation.
I like your idea for raiding except that I'm not so sure it would be easy to implement.

When an expansion launches it typically only has one or two raid zones. Blizzard isn't particularly quick at rolling out additional content so it would take a while for the more difficult/elite content to be available and in the mean time those players won't have anything challenging to keep them interested in playing WOW.

As an example, a lot of my RL friends played Wrath with me and were raiders from Vanilla/BC of the 'elite' type - they progressed well raiding in the version of wow they had played (Original Naxx or Sunwell). When we played Wrath all together they very quickly became bored with Naxx because it was so easy and there was no challenging content for them and they had all quit by the time Ulduar came out because they were so disappointed with Blizzard's new raid philosophy.

I did notice that Cataclysm is launching with 3 small raid zones to start and blues have said that they expect to be releasing new raid content quicker for Cataclysm then they had previously, so hopefully they've learned their lesson and will be able to provide content with a ramping up difficult this time around; but I can't see this working out unless one of those 3 small initial raids is adequately challenging to provide that 'step up' from the starter raids.
One issue I see is that you'd still have problems late in an expansion, when a lot of the new 85s are alts of experienced players. If progression is based on skill rather than gear, these characters would go straight to the latest raids (as they already know them). Will there be enough "true" fresh 85s to run the starter raids?
what if raid drops were minor skill modifications instead of equipment....and such that you had to do different raids to boost different talent trees.
Will there be enough "true" fresh 85s to run the starter raids?

If the risk to reward ratio is balanced right, yes! People being generally too skilled and geared for heroics hasn't stopped people from running heroics until the end of Wrath. In my model heroics would give a bit less rewards, but starter raids a bit more, making them eternally attractive.
What if instead of starter raids you have them scale to the amount of people going in. A long time ago I played a game called Diablo II, whenever a player joined your party you got a message saying the monsters just got stronger. Why couldn't this be implemented for some raids? I've never understood the obsession with fixed party sizes.

One time I had 6 friends wanting to do ZF so we switched to a raid... for that web were punished with really low XP and unable to complete quests. It would've been a lot cooler if the instance would have been harder instead.
I'd remove being able to earn raid gear in 5 mans, no matter how slow. Otherwise, you will still get those with minor skill 'geared up' because they AoE'ed a 5 man 15 times a day, 7 days a week. Those players would ruin PUG raiding for everyone else.

If the intent is to keep 5 mans active, give them tokens for items not found in raid dungeons (minor stuff like trinkets, but still best in slot). Scale the amount of tokens needed for that BiS item as needed, or make a 'raid' level of difficulty for the 5 mans that give more tokens.
I think an overlooked mechanic is the date on boss kills.

I.e. you can make a raid where only 1% of the players can do it on day 1 and 90% at the end. So everyone sees content and the epeen measurers can ridicule when you killed deathwing, not whether. Both are necessary: the majority of the customers should get to enjoy the content and the insecure need to be able to compensate by feeling good about their twitch skills relative to others.

I think it would be easier for Blizzard, and less complaints from both ends of the normal distribution curve, if the easing were more mechanical and open. I.e. Blizzard announces that until the 4.99 endgame, the raid bosses get weaker, asymptotically, based upon the % of player success.
So this week, DW get 1.4% less powerful not because "Blizzard is catering to hello kitty M&S ...." but because the goal for this phase was 10% success rate and players are doing 8%.

Also I think CynicalBrit's comment is that normal and heroic is not granular enough. There is some merit that in WotlK the gap is pretty big and that adding a mode half way between would allow better progression, perhaps it would also allow normal to be 10% easier or heroic 10% harder as well.
I would love to see 2 things:

1. Add a new tier to dungeons with each patch. Rather than just gifting better emblems for the same (increasingly trivial) level of content, increase the challenge. This would also mean the new harder tiers could have better actual drops, to keep the drops relevant.

2. Increase the AMOUNT of gear dropped in old raids with each new raid. When Ulduar releases, make the bosses in Naxx drop twice as much gear and twice as many emblems. When ToC releases, double the Ulduar drops and double again the Naxx drops (now 4x the drops). This keeps old raid instances relevant through sheer volume of drops.

I like your raid design idea, but I do want to point out that you seem to be forgetting a very important issue - TIME.

As you've noted in the past, the time committment that players feel that they are required to invest in any game is directly related to "progression" and overall feelings of "player satisfaction" in terms of the playerbase as a whole. I think that the badge system is a fluke in that regard, simply because a player still has to invest a certain amount of "time" farming heroics in order to get those badges/points. The issue then becomes one of how easy those heroics are and whether or not they offer a challenge to the LFD selected group, or whether it's considered a faceroll for free badges/points. I think we can see clearly here why the social space has erupted into the current ideologies that it has.

The most critical thing about content and advancement within WoW is how the developers have handled mudflation up until now. From all accounts WoW's very design lends it to always be bound by mudflation dynamics, so why even bother with addressing it? I think a simple and elegant solution for raiding is to just implement a percentage rule much like Hagu suggests: A 90/10 rule, so to speak, where 90% of the end game content can be accessible by anyone and 10% of the content is only accessible by guilds/players who are willing to put in the time and effort to learn and defeat the encounters, earn special gear/achievements/titles and whatnot. But design this raid content, tune it, make sure it is as bug/exploit free as possible, release it then leave it the hell alone. Dont make changes later on to nerf the content to where it becomes "facerollable" in the current expansion. The mudflation effects of future content will serve to trivialize the content soon enough.

To support this, I really like Hagu's suggestion to alter the achievement system to where a "date" is used, or a special title is awarded to players who complete the encounter under certain gear/time/date/group requirements(much like is being done with titles and group sizes in ICC raid content where normal/heroic modes are concerned).

The developer isnt going to be able to directly control the social space, but they can sure engineer the framework that supports and influences it. After that, let the social space sort out whose e-peen is actually bigger. And, if done properly, I think it will split itself along those 90/10 lines that I spoke of above. Which, in the end, will be a healthy thing as players can feel comfortable with the amount of time they have invested versus what they have been able to attain.
It seems to me that at the start of Wrath Blizzard actually did do something along the lines that Tobold suggests!

When Wrath began, there were four raids: Naxx, Sarth, Maly and Archavon. I would argue that Naxx and Sarth (no drakes) were very easy. Our guild, which was a medium raiding guild on a pretty backward server, were in Naxx as soon as ten or twelve of us hit 80, and cleared 10-man Naxx before Christmas.

25-man Naxx was harder, and Maly was much harder. And Ulduar was *way* harder than Naxx.

I don't know that you could make Naxx much harder and still make it harder than 5-mans. Remember, 5-mans were harder then than they are now! (Most of them anyway, I don;t think UK could ever have been called hard....)
Well, I'll bite:
"In Wrath of the Lich King, raid difficulty did not go up from Naxxramas to Icecrown Citadel."

The metric for difficulty you use here is the number and nature of deadly boss mechanics, specifically those that require group coordination. I disagree that Naxxramas has parity with ICC normal in this regard, particularly before the stacking ICC buff was implemented.

Deadly boss mechanics not withstanding there are other dimensions to difficulty that your summary doesn't account for. I think any measure of difficulty of raid content would have to include individual player performance requirements. For instance, Deathbringer Saurfang and Festergut were far harsher DPS checks than Patchwerk ever was. The former required not only high sustained DPS but fast target switching, slows/stuns, and high burst DPS. You might find a group in far superior gear failing to mechanics in a Naxxramas fight, but most likely they are failing solely for lack of familiarity with said mechanics. "The “difficulty” of raid encounters in Wrath of the Lich King" is not just a function of special attacks but also personal performance benchmarks that in my opinion ICC had in much greater abundance than Naxxramas at the launch of Wrath.

As for your general proposal, I like it but I fear it would be less accessible than the current raiding scheme, unfocused as it is at times (looking at you ToC). Placing raids themselves (versus the fights within them) along an ascending complexity/difficulty curve over the life of the expansion means that at a certain point, casual players will no longer get to participate. That is not a statement about the ability of casual players to improve their own performance, but the nature of WoW raiding. It's a team game and by your measure one increasingly dependent on group coordination at higher difficulty levels. The kind of person with 5-10 hours a week to devote to the game on an irregular schedule will have a hard time finding a stable group to make progress in increasingly difficult raid tiers.

Is it "fair" to them, then, to lock the dramatic climax to an expansion's storyline behind an unbreachable wall of a difficult raid? Compare that to the implementation of ICC in which the stacking buff served as the rising tide that lifted even the most casual of boats.

While I often liken WoW raiding to sports - and in sports, not every team gets to the playoffs let alone to their finale - I wouldn't want a raid design that ostracized newer or more casual players after a certain point.
I think time is a bigger barrier than difficulty for raiding tbh. I also think Naxx was significantly easier than the later zones. I think Blizzard is going down a binary "regular mode is pretty easy, heroic mode is pretty hard" difficulty-tiering.
I think the main problem with the raiding in Wrath is that the difficulty ramp was actually rather good for newer raiders. Naxx was a good starter raid (despite being undertuned), Ulduar had lots of options, TotC had a couple of fights that really made everyone perform, and ICC again has a couple of more interesting fights.

But that's not so great for the guys who had been raiding for years. They get bored of the starter raids almost immediately -- I guess Blizzard is going to put in some deliberately obnoxious hard modes to try to handle this, but basically you have different groups of raiders with different requirements and there isn't enough content to keep them all happy.

My solution is just allow raiding to get easier and focus the harder stuff for more elitist players on smaller groups.
Beside gear the other problem the TBC demonstrated with tiered progression was splitting the splitting of the population. Those who progress slower will be left behind as there are less and less people running the lower progression dungeons (and those that do still run them prefer more geared people). Unless your token systme offers the same token across all tiers of raiding - but people will ending up farming the lower raids and leave to burn out as guld try to raid every night to maximise badges.
Wasn't the TBC era a hugely successful time, in terms of suscriber growth? How was it decided that the TBC way was the wrong way to do raids?
Why would subscriber growth be equal to success of the raiding system? I would say that during TBC there were the *least* number of players raiding regularly, less than both in vanilla and WotLK.
I agree with the general direction you are suggesting, but the premise that raids didn't get harder is very wrong. Sure, doing ICC doesn't mean you can do the dance for Heigan, but Heigan was so undertuned that as long as you had three people who could do the dance you would win. I would say ToC was easier than Ulduar, but other than that things generally got harder from one tier to the next.
Post a Comment

<< Home
Newer›  ‹Older

  Powered by Blogger   Free Page Rank Tool