Design Flaws and Consequences

This section of my critique will be focused on what I consider to be EverQuest’s design mistakes, but also negative design consequences that may be an intentional side effect or may not be avoidable, at least not without radically changing the gameplay.

I will offer some superficial commentary of possible solutions to EQ’s problems but a full discussion on this is beyond the scope of this essay. I’m aware that some of these suggestions require more detailed explanations as to how they might work and to avoid creating more issues.

Mudflation

Mudflation is the age old problem of Diku hack and slash MUDs. It has long been debated but I’ll include it here for completeness and for those who have not read or thought about it before.

Mudflation is simply the world filling up with equipment and it becoming much easier to obtain over time. This has two big problems: one is the game becomes easier because gear is better, and two is that the desire to go out and obtain your own equipment vanishes because you can get it by trading much easier, or it’s simply given to you.

Mudflation is a large reason why playing on fresh servers is a much better gameplay experience. On a mature server you will gear up a low to mid level player by farming money or something rich players need but won’t bother to farm themselves like tradeskill mats then using the funds to purchase gear from the tunnel or bazaar zone. That is not how designers want players to gear up generally, and they instead want players to dungeon crawl for their equipment. A lot of gear is simply given away for free because it becomes so plentiful.

EverQuest completely lacked an item sink until GoD added tribute, and equipment doesn’t get worn out in this game so mudflation was rampant. EQ’s only defenses to mudflation were to make drops rare enough that the game didn’t get flooded with them, lore tags and no-drop restrictions on items. No-drop tags certainly work but the popularity of the Firiona Vie server shows how it’s not the most popular solution. The ease at which a high level could solo camp equipment from mid-level dungeons was a large part of the problem.

Mudflation is why WoW’s bind-on-equip limitation exists. It’s one of the better ideas but also far from solves the problem.

I think it’s a common misconception that ‘mudflation’ only refers to gold or game currency. It really mostly applies to equipment but it does apply to both. EverQuest did have some money sinks in the original game, which include stones to teleport into Hate and Sky and Tinkerer bags. I didn’t like them adding tinkerer bags because those bags made any bags dropping in future content as rewards much less impactful, but it’s an optional money sink which is easy to swallow. If the game were just full of taxes to do anything it would feel bad. I think rechargable items could have been implemented much better and that would have been a nice sink. Vanity items or micro upgrades for super rich people are good ways of implementing sinks. A healthy economy requires sinks.

Lore vs. Lockouts

EverQuest originally added the ’lore’ item flag for quest items, as it was seemingly not logical to have multiple heads of a bounty you were to turn-in. But Sony/Verant soon found a new use for it: to mitigate camping and spawn monopolization. Brad says in a June 3 1999 post that using Lore flags to mitigate camping was done in lieu of the trivial loot code because he said the TLC would be too disruptive to implement after the game’s launch. (he says this two and a half months after launch) Of course they would later implement the TLC in some zones before reverting it.

Newer games employ something else to mitigate camping: lockouts. If you’ve looted an item, you may not loot another one until after a certain amount of time has passed. This is a better solution although somewhat jarring and immersion breaking. It is also still somewhat easily avoided if powerleveling is allowable. One issue with doing this is that it makes it more tempting to increase the drop rates, which sort of defeats the purpose as then your economy gets flooded with these items and then players have no reason to camp them.

Early Game Rot

EverQuest had a problem of the leveling game getting worse as the server aged, and this was more extreme the lowest levels. Mudflation is a big part of this but not the only cause or factor.

One factor is simply overcrowding. If you stayed ahead of the crowd you had a much better time generally. Camps were wide open to you.

Powerleveling is another factor. EverQuest had few limitations on powerleveling, which certainly can be enjoyable for the powerlevelee but comes at a high cost to the game. Powerleveled characters will take up camps and refuse to group with others, and people playing the game legit will have a very bad experience if they encounter this.

When the Mass Group Buff AA was put into the game, Sony neglected to check the levels of the buffed players, so level 1 players could receive the best buffs in the game. This made players routinely cast these buffs in town for newbies after announcing it to the server. This was a fun communal activity but comes at the cost of early game rot: the lowest levels then become a complete and utter triviality and players will refuse to even play until they get these buffs. Sony did nerf this seven months into PoP however.

The Al’Kabor server was a case study in early game rot. This was the Mac server which was stuck in PoP for ten years. Newbies on this server blew past the first 50 levels with level 65 buffs and free gear. Players who wanted a more genuine experience had to form guilds with self imposed rules to not accept or use these advantages in a pinky swear honor system. Players are much more invested in their characters and the game if they have to earn their way through and meet friends along the way. This kind of generosity is akin to feeding wildlife in the forest: you’re not really doing them favors in the long run.

Quests added to the game put high level, often aggressive NPCs into older zones. Lesser Faydark is a good example of this and that zone became much more dangerous to players of the level the zone was originally designed for. Jaggedpine quests in particular added many obnoxious mobs to the game for quests nobody did. Epic quests as well.

Fabled mobs were a later addition but I think they were a terrible idea. Why would you prevent level appropriate players from venturing into the dungeons that were intended for them? Insane.

Experience bonuses for holidays, special occasions or potions went in much later but I am firmly against these. They ago against the egalitarian spirit of EQ and older games. They make playing the game when they are not active feel worse. Leveling too fast has downsides in the form of skills not catching up and players skipping content because they’ve outleveled it. Giving out large advantages like this diminishes the player’s sense of accomplishment. If you think that leveling is a ‘chore’ that needs to be fast forwarded then you don’t like the game and you should play something else.

Level Racing

I touched on this previously, but another big problem in EQ was that it was too much of an advantage to level quickly, and this is bad for several reasons. One is that players will level in the fastest way possible and this often means doing something tedious, like guard camping. Also players will miss out on a lot of the content, such as exploring dungeons, and they will outlevel their friends and lose the ability to group with them. It’s simply toxic to incentivize ‘poopsocking’ behavior, i.e. playing 12+ hours a day, every day.

Game design should instead disincentivize fast leveling, not encourage it, and the game world should be charming enough that players don’t feel aggrieved for not being able to level race. The game should have a lot of interesting things to do besides grinding exp, like exploring cool looking zones or questing for items that will last a long time. The modern meta of racing to level cap is naturally going to happen on TLP servers but in a brand new game you can get away with restricting this. I would even still argue for restrictions on TLPs/emus, really.

The obvious means to restrict fast leveling is simply by capping the max level players can achieve and then raise it after some amount of time, but how this is done exactly can vary more than you would first think. It doesn’t have to just be X levels every Y number of weeks. A softcap could even be done such that if the really hardcore players wanted to get ahead then they could but only so far. For example, every level after this cap could reduce the amount of experience gained exponentially, i.e. the player receives 50%, 25%, 12.5%, 6.25% of normal.

In 1999 the first level 50s arrived after about 2 months. Instead you might have applied a staggered cap in the original game such that every two or three weeks the cap is raised, with the caps being levels 20, 30, 35, 40, 45, 50. The idea is that the vast majority of players will not hit these caps.

My personal favorite idea is to raise the level cap after players achieve some kind of done-once world event (akin to the AQ gate opening in WoW but on a much smaller scale) which requires many level capped players to accomplish, with the event difficulty carefully tuned such that the time it took to overcome was relatively predictable. I.e. when 10% of the server population is at the cap they will have the strength to potentially win. Although this may not work if your caps are too close together. (maybe 10 levels apart instead of 5)

When WoW Classic was in beta Blizzard had temporary level caps in place to force players to test content because players blow past it otherwise. I recall Youtubers commenting how they enjoyed being kept at a lower level so they could experience the content they would otherwise skip because players are so strongly compelled to just level as fast as possible. I think a lot of hardcore players would end up appreciating staggered caps even if they balk at them at first.

The original design intent of EverQuest was that soloing would be possible by some classes but be slower than grouping. Sony/Verant failed in this goal as soloing was faster for several classes due to some abilities being too strong. (snare) Had the game been much less soloable then this would have mitigated the level racing issue.

The alternative to caps is to punish players for leveling too fast somehow. The Trivial Loot Code is one way to do this. If mid-level dungeons had loot that didn’t become obsolete at higher levels and players couldn’t camp this once the mobs became green, then this forces players to camp it before they outlevel it or trade for it later but trade isn’t much of an option on a fresh server.

Violations of the Prime Directive: RvR

Scaling risk with the reward was a core design imperative for the game, but Sony/Verant failed at this in a number of ways.

Kiting

In 1999 kiting was a hot topic. The first level 50s were almost entirely solo guard campers who kited. Later in the era guards were nerfed when they were granted high magic resist and longer respawn timers. Brad really didn’t like kiting:

“I’ve always said, and this dates back to before Beta 1, that EverQuest is primarily a group-oriented cooperative game, but that certain race/class combinations would be able to effectively solo, and that soloing would typically be more time consuming. And I stick by what I said. As the game has matured, however, certain methods of kiting now allow some classes to solo NPCs that con yellow or red with little to no risk. We, the designers of EverQuest, feel this is not only unfair to other race/class combinations that by design can typically only solo blues, but that it is also threatens the long term health of the game.” – Brad McQuaid

At one point in late August 1999 on Sony’s test server, they made all DoT spells do no damage on pursuing NPCs. This is the same nerf Sony would years later apply to bard AoE DoTs. However 1999 Sony did not take that particular nerf to live servers. They instead implemented the well-known 1/3rd reduction two weeks later, which was removed mid-Luclin. I think it’s safe to say that Sony would have gone with a much stricter nerf if they had known about it during beta.

Kiting is a tactic that is something EverQuest is known for, so it’s not easy for me to argue against it but looking at this objectively I can’t not criticize it.

EverQuest is a game where NPCs are overpowered and certain player skills are also OP. In this respect NPCs might demolish you or you might demolish them. Snare is one of those OP abilities. It cannot break early, it lasts an eternity, it costs almost no mana, and it slows the NPCs enough to allow you to cast on them like they were rooted if you simply did a little jogging prior. It’s OP AF. One reason why melee felt relatively underpowered is simply because some spells were too strong and snare is one of them. Spells like this made EQ PvP kind of a joke.

The RvR for kiting is abjectly terrible. Kiting flies in the face of the group-centric intent of the game. Leveling a druid is ez-mode for sure. Compounding the problem was that druids also got Spirit of Wolf. Sony later admitted that SoW was a mistake and that they would have made the spell fade in combat situations. Spirit of Scale was a SoW “upgrade” spell in Kunark that offered faster run speed but faded if the player attacked or got hit so players didn’t use it much.

Kiting was so advantageous that at one point druids outnumbered rangers 10 to 1 on a new server launched in 2000.

In an ideal game, the amount of experience you were awarded would scale along with the difficulty of the combat, no matter where you were. Dungeons have Zone Experience Modifiers (ZEMs) to make harder dungeons grant more exp. (Sony of course screwed that up, but anyway) If some super smart AI could analyze how difficult every fight you did was and award exp based on how much you struggled, that would make for a much more entertaining leveling game. I strongly believe that should be striven for but it’s very difficult to pull off because it’s near impossible to program a system that can accurately gauge danger and risk. Awarding experience based on the number of hits a mob lands on you doesn’t make any sense because that rewards playing poorly and intentionally allowing mobs to hit you more and waste healer mana is just stupid gameplay. (defense skill raising aside)

Kiting reached its highest level of absurdity in PoP with bards. It is well known on emulated servers that bards can round up a bunch of NPCs and sing PBAoE songs on the train while running in circles around it. That was actually a rare thing until PoP when Sony improved the pathing logic, plus PCs got powerful enough to not drop FPS, server hardware was upgraded enough, broadband became available, and bard AAs extended PBAoE song range erroneously until December 2003. Also I believe that Sony may have relaxed a stuck NPC check when they improved the pathing at this time; prior to this NPCs could warp on you if you circled tightly around them. On emulated servers, NPCs “bunch up” much more readily as well.

Bard PBAoE was simply much less reliable (maybe impossible in many places) prior to PoP. It was not done very much on Al’Kabor even though AK had the extended range on PBAoEs from AAs. Bards in PC PoP were rounding up large numbers of NPCs and getting nutty amounts of exp from them, so Sony tried to nerf this by making NPCs start teleporting back to their spawn points and mem wiping if more than 10 NPCs were around in some PoP zones, but that caused issues with things like swarm pet AAs so they reverted it and curbed it via other limitations I’m uncertain of, at least in PoP. (pulls of 10+ mobs still teleport in HoH to this day however) Sony finally had enough and nerfed bard PBAoE swarms into the ground in September 2004 when Omens launched by making the songs do no damage on moving NPCs. Sony would have nerfed this sooner had it been a thing earlier. If Brad didn’t like druid kiting then he certainly would not have liked bards killing 50 NPCs at once.

Camping

Believe it or not, camping was not something EQ’s designers anticipated very much. Their expectation was that players would dungeon crawl and move from room to room.

Camping was done because you had to do it to keep your claim on a unique spawn, but also for the safety. Once a room was broken into, the respawns would be staggered and the risk declined dramatically. This is a bit ironic because on MUDs, areas respawned all at once, not on an individual NPC basis. EQ should have had linked respawn timers somehow. At the minimum the entire room of a unique mob should repop simultaneously, if not the entire immediate area. One might code the room to respawn any mob that had been killed more than five minutes ago when the timer was up to prevent immediate repops. Timers should also be partially randomized so players couldn’t set a stopwatch. This design would encourage players to move from room to room as sitting still would become somewhat dangerous.

Another anti-camping design choice would be to have NPCs from nearby locations in the dungeon begin to move into the room if players stay in it too long. (or just spawn some) This would make some logical sense if you think of the NPCs as being able to reason: the longer you sit there, the more likely the rest of the dungeon would find out and send in reinforcements to try and dislodge you.

More patrolling NPCs should have been put in to eliminate safe spots. Dungeons should have no safe spots, anywhere.

The revamped Cazic Thule zone has a somewhat different design than previous dungeons in terms of rare spawns. That zone has a lot of named spawns in it, much more than typical, but the spawn rates are significantly rarer. (honestly too rare) This at least allowed and encouraged groups to clear more than one named mob placeholder instead of merely sitting on a single placeholder and clearing nothing else, although they would likely still sit in one spot.

I think many named spawns should not spawn in the same location. That’s the simplest way to stop camping. It also makes sharing a lot easier because the server will drop the named on a random group somewhere instead of one group monopolizing it. Random spawn locations can’t work for every spawn however; it wouldn’t work for the Ghoul Lord, (or I suppose you could but it would be weird) however it could be done for the ghoul arch magus, especially if you had multiple rooms in the zone with bookshelves in it. Guk would need to be somewhat larger to accommodate this.

Single Pulling

In EverQuest, there was no bonus or consideration put toward the amount of experience you got if you killed multiple NPCs together vs. killed single. The lack of this ended up making killing mobs one at a time the most efficient way to grind experience. This is bad because it rewards easier gameplay over more challenging gameplay. Other games have attempted to deal with this by applying exp bonuses to multi-pulls and I feel that this is something that is needed. This is hard to pull off due to the exploit potential of implementations however. This is also needed simply to make dungeon crawling attractive instead of outdoor zones to level in. Mez spells make this tricky to implement– you don’t want to grant the full bonus when all the NPCs are mezed but you don’t want to discourage the use of mez either.

The Best Exp Wasn’t Always Dungeons

The highest experience flow should always come from deep (not near the zone-in) dungeon delving in a tank, healer, CC, DPS group. If that is not the case then there is a problem somewhere.

Other games try to semi-solve this problem by making huge group bonuses, so at least the best exp is in groups, but I think that’s a bad idea. The synergy of different classes grouped together should be most of the increased efficiency and not some artificial group multiplier. If your game needs a large group multiplier then may I suggest that your class interdependency is insufficient. That is not to say that I think there should be no group bonus multiplier, but it should be small, and it was in EQ prior to June 2003, so old EQ got this right. Soloing should not be curbed by group exp multipliers, it should be curbed by skill and zone balance. If the NPC is easily soloable then it shouldn’t award much exp to groups either.

Low Dark Blues Was The Best Exp

This is a big one. In EverQuest, at higher levels, the experience formulae worked out such that the best exp grinds were killing the lowest dark blues for your level in quantity. Killing higher level NPCs in harder locations was pretty much a complete waste of time exp wise. This was a betrayal of the core ethos of the game and it was not like this at lower levels. At the very lowest levels, killing the highest level mobs you could was well worth it because even one more level on the NPC meant getting double the exp for it. That quickly stops being the case after about level 5 however. At the high end, a level 42 mob will award half the exp of a level 60 mob, which is rather crazy. This problem was mostly fixed shortly before PoP opened.

Powerleveling

I touched on this previously but powerleveling is a huge RvR violation. Leveling with level 50 druid buffs on you is basically cheating. Powerleveling can of course be fun and enjoyable but as often in game design there are pros and cons to many if not most design choices and these choices are not cut and dry right or wrong.

Pros to allowing powerleveling:

• It allows you raise up an underleveled friend into group range.
• It allows you to switch mains faster or at all if you don’t have the time to level legit.
• It allows you to experience other classes as alts at higher levels.
• It’s fun, at least in the short term.

Cons to allowing powerleveling:

• It’s a very negative experience for the nearby players who are trying to level legit to have powerleveled characters around.
• Players who skip the tribulations of leveling legit are not as invested in their characters.
• Leveling fast means the player may quit sooner as they have little to do at the level cap.
• Powerleveling is fun in the short term but playing the game as intended is more enjoyable long-term.
• Few classes can do it well, which causes a class balance issue.
• It’s highly solo centric gameplay in a game designed to be cooperative.
• Hardcore players having stables of alts will hog content and powerleveling allows them to have stables of alts.

A consequence to EQ’s powerleveling is that it resulted in some players making druids as their first character (or perhaps enchanters) even though they didn’t want to ‘main’ the class. They made a druid first to powerlevel and gear their intended main later. So the players leveled the druids solo then leveled their melees solo via powerleveling. That is not healthy game design.

Anecdotally I played a druid on Project 1999, and powerleveled several associates of mine to level 30 or higher. What ended up happening is almost all of them quit playing shortly after.

Another problem with being able to level characters up very quickly is that people will create stables of characters for a single purpose and this destroys the core interdependence design of the game. One particularly unsavory thing players did on TAKP is guilds created a small army of mages to use as CotH bots to move the entire raid around Vex Thal to avoid clearing trash in the zone. They camped these mages in safe spots then used an exploit to get targets for group invitations.

Tradeskill Materials Farming

This one is a bit of a nitpick and subjective but I think there is a good argument to be made that the best tradeskill material farms should be in difficult places that only higher level players can go. Spider silk for tailoring for example was best farmed from East Karana Crag spiders. Spiders in Lower Guk should drop much more of this than some level 19 outdoor mob.

Bad Zone Experience Modifiers

In the original game, the indoor and outdoor ZEMs were so close together they didn’t really matter all that much. This didn’t make sense because killing NPCs outside is so much easier and low risk. Zones like Kedge and Siren’s Grotto should have had significant ZEMs. (Kedge had one but maybe not high enough) It made no sense that Dragon Necropolis had the same ZEM as Velk’s Lab either, which were both not much above baseline anyway.

There is some evidence that city zones had an elevated ZEM, equal to that of Kedge, which I feel the jury is still out on but if it was the case then it was really stupid.

In Luclin, Sony massively jacked up ZEMs for some of the lower and mid level zones. Some of these ZEMs just didn’t make sense. High Keep exp was double that of a baseline zone, and on the Al’Kabor server it was the go-to zone to level up from 25 to 52. The risk vs reward here was extremely poor because the risk was very low. Sony saw players killing guards and decided to reward them for it much more than dungeon crawling which is dumb. To top it off guards had good cash loot. Higher level zones got small increases or none at all. CT 2.0 didn’t get a higher ZEM which was a huge oversight.

Sony did however slap massive ZEMs on PoP zones, without increasing ZEMs elsewhere. While it made sense to give PoP zones an elevated ZEM, the combination of these large ZEMs plus the high level kill exp bonus made PoP zones so much better than previous era zones that it became a waste of time to level elsewhere if you could be in a PoP zone. The RvR in PoP zones was somewhat poor due to the lack of summoning on NPCs. Ideally you’d want a zone like CT 2.0 to be at least a reasonable place to level after PoP, but even after the high level kill bonus PoP zones crushed it.

In PoP Sony started putting experience multipliers on individual NPC types. One of these was put on non-summoning frogs in Plane of Storms, which made absolutely no sense. If anything these should grant less exp than the frogs that do summon. The flameheads in PoFire also got large bonuses, and since they didn’t summon they become large exp pinatas for kiters. What Sony should have done is put exp bonuses on mobs deep in the PoNightmare caves, or the Ghoul Lord, or places like these. Mobs that are hard to get to in dangerous areas should award more exp and more than mobs near a zone-in.

Stats Were Near Useless

It’s common knowledge that stats in EQ didn’t do very much, but players didn’t quite realize how useless they were. I think Sony got away with this because stats in EQ were somewhat mysterious and players didn’t know the exact combat formulas like they do in modern games, otherwise players would have complained more loudly.

Why was the game made like this? Several reasons, that I’m aware of:

• Gnome warriors had to tank well. It’s that simple.
• Brad himself mentioned that stats were designed to not do much to limit twinking.
• If you had to CR and fight naked or with backup gear, you would struggle to kill anything.
• They made simplistic algorithms and didn’t expect to be making content after 50 when they made them.

Remember that in the original game, players had basically no gear with any stats on it until like the 30s. Tanks replaced their statless banded armor in the 30s with statless bronze. Maybe you had some Unrest drops with some stats on it by the 30s. This was Lady Vox loot, i.e. one of the best drops in the game.

Brad’s excuse is a bad one. There are very simple and effective solutions to curb twinking. Sony in fact implemented some in the combat logic. For example, weapon damage from equipped weapons are capped under certain level thresholds.

I would note however that stats worked well enough that a level 1 ogre levels a lot easier than a level 1 erudite paladin. It was rather brutal as a level 1 erudite paladin actually.

Resists actually hard capped at 100, before level difference modifier. (against a level 70 raid boss as a level 60 player the cap would be 150, or about 225 for lure’d AoEs) 100 is quite low even for vanilla EQ. By the time players were doing Velious content, players could outright ignore resist buffs if they were wearing resist gear on, or they could ignore resist gear with resist buffs if their normal gear had some resists on it. Although on the flip side, getting an item with MR on it before you had 100 MR was very significant as every point of resist up to the cap did a lot.

I don’t think I need to go into detail as to why stats doing nothing is bad design.

Itemization Mistakes

Good itemization is absolutely critical to a MMORPG and the impact is perhaps underappreciated. The second R in RvR is comprised of two primary carrots: exp and loot. Both Rs need to be properly calibrated or the game fails and players are unsatisfied.

I think that if you left EverQuest’s design untouched and did nothing else other than perfect the itemization RvR you could improve the game considerably.

Bad Zone Itemization

It does not matter how awesome you make a zone, if there is nothing to get inside then players will ignore it. On the flip side if you put good loot on boring content, then players will torture themselves on this bad content for the loot then they will complain how bad your game is.

Case study: Siren’s Grotto vs. Velketor’s Labyrinth
Velketor’s Labyrinth was one of the most frequented zones in Velious, compared to Siren’s Grotto which only had people in it because they were traveling to Western Wastes, or were doing the shawl quest. Velk’s was full of nice loot, full of juicy low hitpoint dark blue mobs for exp, wasn’t hard to pull, and wasn’t very far from town and a ring. Grotto was inconveniently located, the mobs had poor exp/kill, had high MR, the zone was full of enchanter NPCs which were hardcoded to ignore your MR and it was a semi-water water zone. Grotto should have had the great loot plus a ZEM, not velks.

Case study: Skyshrine vs. Kael
Skyshrine actually had some loot in it if you weren’t aware. Of course hardly anybody did the zone because the layout made fighting in there an enormous PITA, even for full raids. The RvR of Skyshrine was so bad that Skyshrine was one of the least explored zones in the game at the time. Kael on the other hand was very convenient, not very dangerous and had a lot of good drops including full armor sets for dwarf factioned players. The entire zone of Skyshrine was just to make players run a lot to turn in quests. Of course if you put BIS items in such an obnoxious zone, it will force players to punch themselves in the balls for them, but the solution to this is to not make obnoxious zones.

Case study: Chardok vs. Sebilis
Chardok had so little inside it that Sony decided to revamp the zone before Luclin opened, and then proceeded to put in drops so strong they would need to be disabled on TLP emus because they would obsolete too much Kunark era gear. Item wise the difference between Chardok and Sebilis is about as stark as it gets: Sebilis was very well itemized and the zone was a big success. Chardok was a bit linear but the zone was otherwise pretty good if not for the complete dearth of items. Chardok was unfortunate because it made everybody crowd into Sebilis and the goal of designers should be to spread the population out.

Overly Strong Items

If a designer makes an item that is too good then it will invalidate all previous items in that slot, making them obsolete overnight, especially if this new item is easy to get. Care must be taken to preserve the RvR because new items can destroy the RvR of old content (entire zones even) overnight.

Case study: Chitin armor in Velketor’s Labrynth
The armor that dropped in Velk’s was so good for chain classes that it obsoleted previous era raid armor, and was comparable to to the Velious quest armor. Crystal and Silver Chitin dropped from the most frequented zone in Velious thus was very commonly available and tradable.

Case study: Bracer of Benevolence
This bracer is from a dragon faction Velious quest and has two problems with it. One is the stats are so good it obsoletes everything that isn’t a Vulak or Dozekar item. Some people might even prefer this over the Vulak wrist item since hitting the AC and strength caps are so easy and MR is a valuable stat. The second problem with it is that the quest was easily trivialized with the harmony spell, allowing groups to single pull the captain, which was also near the zone-in. The quest was also just a single fed-ex step and one fight. The only thing preventing this quest from being completely ridiculous is the 10 hour respawn rate of the scout. The bracer isn’t even lore!

Case study: Jade Mace
This mace drops from the first named mob in Karnor’s Castle which is the most frequented zone in Kunark and perhaps the most camped unique spawn in the entire expansion, thus the economy had a steady supply of them from day one of Kunark opening. The weapon is 9/18 with a fairly low delay which is preferable for the damage bonus. That ratio is better than god loot: Briarzephyr is 9/19, same as a Lamentation from Sebilis; Blood Fire is 11/24. It’s not far off from VP weapons in the main hand because of the lower delay. Even if the drop chance of it was somewhat low, the RvR for it is stupid because it’s such a safe camp next to the zone-in. Supposedly when Kunark opened it dropped from almost any drolvarg in the zone before being moved to the captain.

Case study: Elysian armor
This is the druid armor set from Umbral Plains. Sony in their infinite wisdom decided to make the mobs this comes from not summon or cast spells. They’re easily soloed by druids who aren’t even near level 60 yet. (albeit will take awhile) The armor quality is superior to Velious quest armor. The RvR on this stuff is terrible. In this case the issue is more so that the mobs the armor drops from are way too easy rather than the stats themselves.

Most Tradeskill Equipment Went Uncrafted

Sony admitted to putting in a tradeskill system as an afterthought because they felt they needed one as Ultima Online had one. Frankly it shows. There are vast numbers of tradeskill recipes that produce weapons and armor that nobody ever used, although there were some bright spots.

The Kunark patch that put in epic quests also put in the first cultural tradeskill armor and allowed a single tradeskill to cap at 250. Sony made crafted armor sets with the exception of adding a deity requirement on them and they had to be combined in a racial container. The idea was to differentiate crafters, as before this anyone could max all tradeskills and make everything, so there was no interdependency. While the attempt was admirable, (the one skill to 250 was a good idea) the problems with this were that the armor sets were just carbon copies of each other with a different deity requirement on them and the effort to make them wasn’t worth it. Also some races couldn’t make anything (half elves) or couldn’t make much (trolls). Also having the wrong deity or race locked some people out of gearing options which is unfair. It’s OK if options are limited as long as there is some kind of balance instead of one race+class getting all the good stuff.

Honestly this was a lot of wasted labor and recipe materials cluttering treasure tables. Had more effort been put into the stats of tradeskill items, or better trivials set, or drop rates more reasonable, then a lot of effort could have been spared or not been in vain. EQ was timid in this regard and was afraid of making recipes produce good stuff because all the mats were from low risk places.

Fierce Heraldic Greaves was one of the items players actually desired (a Luclin recipe) because it was better than anything outside of Vex Thal and the High Priest. The mats to make this came from Hill Giants and Griffons and the plat sink was Blue Diamonds. Sony just didn’t make recipes with mats that came from raids and so the resulting equipment had to be inferior for the most part.

I think a better design would be if like half the gear players wore (including raiders) came from tradeskills and half from drops. In this scenario, raid bosses would be dropping a lot of tradeskill materials instead of loot or both. This would actually give content developers some more flexibility, as they could for example, require combinations of bosses to be killed for a finished piece of gear. It would also give players more flexibility, as they could exercise some amount of control over which items resulted from a kill. Although these items would need to be tradeable or bind-on-equip but there are workarounds to that. (e.g. no fail combines)

Drop Rates Ruined Some Items

It’s not enough to make a good item placed in a location where players have to work for it; if the drop rate is not set correctly then it can ruin the item’s RvR because the reward shrinks if the drop chance is low.

Case study: Planes of Power lower tier drops
For Planes of Power, Sony made rare spawns and drops for casual players in the expansion’s lower tier zones, but they set the spawn and drop rates so low (literally 1% in some cases) that some of the items didn’t even end up in spoiler site databases. Darkpaw increased the drop rates on some Plane of Disease drops in 2019 to “fix” them.

Case study: Cazic Thule 2.0
I actually like the CT revamp: it’s the only serious dungeon crawl zone in Luclin and PoP and they made it hard and the drops are quite good. The problem with this zone is the rarity. Spawn and drop rates are too low. Rare spawns were 5-6% on a 20 minute timer. So groups will kill roughly 2.5 PHs an hour and would get about 3.5 named in 24 hours of camping. They put in a lot of named spawns to compensate for the low spawn rates and 20 minutes is fast for a high level zone but being fast just meant you couldn’t clear much of the zone before repops. These mobs are tougher than juggs in Sebilis. They added a bunch of tradeskill recipes as well, but the Flawless Tae Ew Hide for example has a 2% drop chance off of mobs that are level 58-60 deep in the zone. This makes armor that is about comparable in quality to Umbral drops, which are much easier to get, especially for druids. Realistically you’d need 4 flawless hides to make a tunic as the combine failure rate would be half or more. And you’d need Tae Ew bloods which are also rare and is needed for more recipes. Basically these were so rare they were only used for the weight reduction bag.

Case study: Akheva Ruins
This zone is full of ultra rare mediocre weapons, many of which are no drop. Akheva Ruins was really lacking in reasons to go inside other than raiding and they made a bunch of mid loot that doesn’t even drop due to rarity.

There are some notoriously super rare drops that are valuable in the game, like the Bottomless Box from Twilight Sea, and the Rend robe. I think having a handful of ultra rare items is a good thing. It’s like winning the lottery. People like to gamble. But it’s pointless if the items are worse than much easier to get options.

Case study: Bronze armor
It’s well known that small bronze armor is much easier to get than medium and large. I remember the location of medium bronze armor being such a mystery in the earliest weeks of the game. When we discovered it in Solusek B it was like a revelation. By the time medium armor had been discovered, small bronze was already cheap to buy from other players. Solusek B dogs are the same level as wans in gukbottom. By the time you get bronze in Solusek B you’ve already outleveled rubicite camps and could be doing Lower Guk camps for better stuff. Large bronze isn’t in a better situation since it only drops from two rooms in the game from infrequently camped NPCs, albeit the NPCs are about 5 levels lower.

Gooderer Items

At some point starting in Luclin, items devolved into what I am calling ‘gooderer’ items. These items are just like the items that they replaced, except they have more of each stat. This is mostly evident in hitpoints, so items in one era might all have 100 hp on them, then the next era they have 150 hp and so on. Items started getting some of all stats on them too, so an item will have significant amounts of AC, HP, STR, STA, all resists etc. You gotta go pretty far back in EQ to find raid items that only have a handful of stats on them.

This is partially because stats didn’t do very much. Items in the earliest eras were much more unique and standoutish. Since stat caps were hard and easy to hit, item designers in some cases just slapped silly amounts of stats on items to make them look much better than they actually were– Vulak loot is the best example of this. Of course in Luclin they made the AC cap a bit soft then in PoP they made cap even softer, so the result was that some Vulak and VT items were better than PoP items because they slapped so much AC on them.

Lack of Item Diversity

EverQuest items lacked in diversity. What I mean by this is that players rarely had to make a choice as to which items were better for them. In other games, the choice of what to equip is not so obvious. There should be trade-offs. Do I want to gear for survivability or damage? WoW did this better. Smart use of negative stats would make itemization much more interesting as well.

In EQ entire sets of armor were just the single option you had. Elemental armor in PoP for example. It was just a no-brainer that everybody would wear it. The notable exception to this is the Kunark sets, but I doubt the statless armor being unappealing was was intentional.

Dependance on No Drop

In EverQuest almost all the raid drops were not tradable. I would concur that raid gear should be earned and not bought for the most part, and this goes a long way to curb RMT, however the downside of having no real economy for raid stuff is kind of a bummer. Guilds who reached the elemental planes in PoP could make a lot of platinum by selling bows, and it was part of the reward for achieving elemental plane status.

I propose another option: design a game such that you may trade raid items only to other players who have also defeated the content. It doesn’t make much sense from a continuity perspective but it would allow for some limited trade and would somewhat alleviate the problem of the random number generator’s streakiness. It also encourages guilds to get along.

No Long-Term Planning

In the early years, EverQuest’s designers were working under the assumption that their current era/expansion would be the last one, and so they didn’t do any long-term planning which resulted in serious problems down the road. It’s hard to fault them too much because nobody can predict how well games will sell, but I will outline some of the issues that arose from this.

Complete Heal

The story about Complete Heal is fairly well-known. Sony/Verant made spells based on D&D and implemented a spell which fully heals the target. The downside to the spell was the cast time of 10 seconds. The problems with this spell are that the efficiency of it skyrockets as player maximum hitpoints increase, which results in cleric mana never running out while using the spell if only the tank took damage, ensuring that other heals are not used; also the 10 second drawback stops being a drawback when you coordinate other Complete Heal casts with other clerics.

The spell was fine in the original game, but when Velious rolled out and fights became much longer, using single target heals became too costly to use and so players developed the Complete Heal rotation/chain which became not just standard practice but necessary. In Luclin fights became so long that the single job magicians had was to conjure modulating rods and drop them on the ground for clerics to use.

The rotation strategy blindsided Sony/Verant and they did not expect it. Their answer to this was to simply make the bosses hit harder, instead of nerfing the spell which one might have expected them to do. I would argue that Sony’s response to this actually more or less worked, at least for some time: the game remained challenging and organizing rotations tested a guild’s ability to cooperate which is something you want in a MMOG. It’s not the design you would go with if starting from scratch but it works as a game.

The main problems with rotations (aside from the gameplay not being very engaging for the clerics) were that clerics started not running out of mana with mod rods. Mod rods were nerfed about when PoP launched but it didn’t take long for clerics to again be able to rotate forever. You could simply have more clerics in the rotation to accomplish that regardless. The second problem is that druid and shaman healing became absolutely worthless on warriors. Sony had to give them CH-like spells but they were no substitute until the druid PoP heal. Then there’s the issue of clerics forever using the same heal spell they got at level 39. Sony did cap the heal at 7500 HP when PoP launched. Then there’s the problem of mobs obliterating any non-defensive tank before the rotations could readjust to the new tank which was a challenging obstacle which required macros and very fast reactions that are counter to the game’s otherwise slower nature.

Another huge problem with the spell is that healing aggro was capped in early Velious, which made using the largest heal possible very advantageous. I feel that this was a big mistake and that aggro should have remained a downside to using the spell. (also it made heal aggro utterly trivial in general as player power increased) Due to heals being subjected to witness checks, what happens in boss fights is that heal aggro is high for some clerics to start with but over time healers end up closer to the bottom of the hate list. If heal aggro was not capped or had much higher caps then this would force the use of smaller heals to start with before transitioning to a CH chain.

An alternative way of nerfing the spell might be to scale the mana cost with the amount healed: minimum 400 mana but all HP healed above X will subtract even more mana from the cleric depending on the amount healed. You could still leave the spell being somewhat (not a lot) more efficient than normal heals this way because a downside to this for CH rotations would be that some clerics end up going out-of-mana much sooner than others due to the randomness of melee combat, which would balance rotations as a strategy.

32k HP

Sony originally used a 16 bit signed integer for hitpoint values, which has a size limit of 32,767. This is an obvious mistake even in 1999, but Sony didn’t anticipate players would be forming raids of even 30 people back when they were designing the game, combined with them not knowing they would be releasing any expansions. In the original game Sony granted some raid bosses a death touch ability because of this hit point limit, which I would argue isn’t very good design for reasons I don’t think I need to go into.

The less forgivable mistake was not raising this to a 32 bit integer in Kunark, when they had fully realized the problem for at least a year. Sony tried to compensate for the low hitpoints by granting mobs regeneration and lifetap attacks. The problem with regen is that it highly encourages zerging NPCs down and makes killing them with smaller raids impossible. Kunark bosses blew up quickly in high DPS raids. Trakanon in particular was vulnerable to zerg swarms because his breath AoE has a long recast time on it and he might die before he even got a chance to do it three times.

There was another little known problem with the 32k hp limit however. Hate generation from spells is determined from the target’s max hitpoints, not level. The standard hate amount for most spells is max_hp / 15. When Sony removed the 32k limit in Velious, they neglected to cap the hate generated by spells, and this made spells like the Tash line do absurd amounts of hate. Sony started capping hate from spells two and a half months after Velious launched. Even on 32k hp mobs however, the best warrior weapon was the tash stick from Sebilis before hate caps went in, as tash on a 32k hp mob would cause tree times the hate amount as a single Enraging Blow proc.

Skill Caps

Skills in old EverQuest were limited to 252 because Sony used 8 bit integers for these. But more importantly Sony’s skill check algorithms were linear in nature which would prevent perpetual increases in skills as at some point skill checks would reach 100%.

An example: a double attack roll is a random number up to 500, and if the roll is less than your double attack skill + level then the double attack succeeds. This makes a level 60’s double attack rate 61% for a warrior. If the skill had gone up 5 per level after that then the warrior would reach 100% double attacks at level 93 with 410 skill. This is why some skill caps even on Sony’s servers today are still in the neighborhood of 250. A design with many expansions in mind could have instead put in a logarithmic rate of increase, thus allowing for an increase in skills for every level.

Percentage Based Spell Effects

Spells like snare and slow were percentage based afflictions. A level 1 snare spell costs 15 mana and hinders a level 80 target as much as a level 1 target, for the same duration. Unsurprisingly snares and slows are some of the most overpowered spells in the game. Ensnare and Turgurs could not be replaced by more effective higher level versions.

One way to deal with this is to create level limits on them like charm and mez have, but druids have enough spells in their book. I think spells like these should have mana costs that scale with target level. They also should have been weaker in general. More about class balance later.

AC Hardcaps

EverQuest had some rather strict hardcaps on AC in the early years. AC was a valuable stat until these caps were hit then the player got zero benefit from more AC on gear. That the caps were hard until Luclin was arguably a mistake in itself, but the caps were not particularly set very wisely.

In the original, pre-Kunark game, the cap was 350 for all classes. This made no sense because that meant that casters essentially had no cap because it was so hard for them to hit it. Casters were already much stronger than melee at the time and this was just gasoline on the fire.

In Kunark Sony raised the warrior hardcap to 405, but left the other classes at 350. At Velious launch or shortly after they raised caps for all melee classes. Basically Sony slowly raised caps for melee classes while leaving casters mostly untouched which is how they fixed the problem.

I feel that these caps were lazy design and instead careful itemization should make caps unnecessary beyond sanity checks. Players should be able to gear for armor class at the expense of all other stats and see benefit from it. Hardcaps in general are a bad idea; instead use logarithmic curves or piecewise functions if going linear won’t work.

Robe Graphics

The original game had very limited texture memory available due to the hardware limitations at the time, so they could only have so many robe graphics. Instead they relied on color tints to distinguish robes from each other. The problem here is that the non-tinted robes look the best, and Sony often put non-tinted robes in lower level places, like Najena and the Oracle in OOT. I think it would have been better to reserve non-tinted robe graphics for the hardest to get robes. The magician quested robe in Solusek Temple should also have had a tint on it so as not to look like the robe from a god.

Invisible Men, Hardcodes & Kludges

The invisible men phenomenon was the result of Sony being unwilling to devote time to implement more robust and intricate systems under the hood. Sony wan’t expecting the game to last anywhere near as long as it did so they used a lot of hacks and shortcuts. Content designers used unkillable invisible men as timers and state machines for dynamic events because the game lacked a proper scripting engine for some years.

When examining client decompiles it’s very obvious that Sony hardcoded a lot things that should have been softcoded, particularly spell IDs. A good example of this were resist adjustments, i.e. lure’d spells. This should have had a proper spell field added, which they eventually got in late Luclin.

EQ was just full of kludges. This made it hard to maintain. Had they known at the time that the game would have many expansions I’m sure they would have coded it better from the start.

Raid Design Criticism

This will be largely subjective but here is my take on Sony’s early raid designs.

Far Too Single Target Focused

I touched on this previously but the raid game was far too single target oriented. I.e. one NPC vs dozens of players.

Why is this bad?

• Only one tank gets to fulfill their primary role.
• Too simplistic and predictable.
• Unchallenging for most raiders; Players just mash their button(s) on a parked single target for 15 minutes.
• Melee DPS output from the NPC needs to be massive; entire raid falls apart in seconds on tank or healer mistakes.
• Silly from a lore perspective; other than dragons and gods, makes no sense.
• CC and debuff classes end up relegated as buff bitches.

Planes of Power started making encounters with more targets. Terris Thule is one that spawned gargoyles for example. The Rathe Council I thought was well designed. The Rallos Zek the Warlord 3rd phase is fantastic if it weren’t for charm trivializing it. This is partially why the expansion was successful.

There are some problems with multi-mob encounters however. One is that suddenly spawning a bunch of adds results in them insta-killing people or otherwise ganging up on people before the tanks can get them under control. One could argue that this could be part of the challenge but I think it could use some AI tweaks to make these fights less obnoxious. When players are put in front of a wave of spawned NPCs, they cannot instantly select their targets and focus fire in milliseconds and neither should NPCs. The AI logic could be written such that the mobs refuse to target the same player as another mob, at least until after some time passed, to prevent this unfair behavior. Also target switches by NPCs could be delayed so they don’t all simultaneously swap to the cleric that landed the CH.

Another problem is loot and raid wipes. The loot needs to somehow not appear until all the targets are dead, and if the attempt fails, all the mobs will need to respawn. This would require a non-trivial amount of scripting for old EQ but is certainly doable, especially with the emu codebase. The corpses could be locked until all of them are dead.

Imagine a six mob raid encounter that mimics a player group: a tank NPC, a melee DPS NPC, a caster DPS NPC, a cleric NPC, a enchanter NPC, and a whatever NPC. (maybe even randomize the last one) And each of these is strong enough to require at least a full group of players to handle. Deciding which target should go down first would have pros and cons: leave the cleric alone, and it heals the others. Leave the enchanter alone, and it mezes people. Leave the caster DPS alone and it does a lot of AoE damage to the raid. Leave the tank alone and it starts protecting the lowest health mob with /shield or knocking people away from it. This is a lot more interesting than a large sack of hitpoints walled into a corner.

One Mistake Wiping A Raid

This sort of goes with the singular target design criticism, but I feel that raid design should generally follow a principle of mistakes from any one player not resulting in guaranteed wipes. Mistakes should be recoverable if other players notice them and act quickly. I feel this way because your guild should not be held back simply because you guilded some dude’s wife. It puts a lot of pressure on everybody which makes the game feel unenjoyable to people who aren’t hyper focused achievers with a lot of time on their hands. It’s anti-social and will result in players yelling at players. And it will limit raid sizes because the more people you have in raids the more likely one of them will wipe your raid with a Baron Geddon bomb. This is one reason why raids in WoW went from 40 to 25.

EQ was more forgiving than WoW in this regard but if you were a warrior or cleric you could not be an unreliable player. Ideally the collective effort of the entire raid should be what wins the day. The burden of performance should not just be shouldered by a handful of people in the raid, it should be spread out.

Trash Killing As a Time Sink

EverQuest evolved into a design pattern of large numbers of trash NPCs before bosses were reachable. Plane of Growth and Vex Thal were perhaps the most noteworthy examples of this. For Plane of Growth at least I can kind of give it a pass: Tunare was one of those “we kill it every once and awhile, not weekly” mobs because the loot table was not great. Vex Thal however was not like that: it was the BIS drop land end-game zone, so players were either going to do that zone or almost nothing.

The problem with this design is that it’s not enjoyable, obviously. VT seemed designed such that the bosses were easy to kill and clearing was the hard part. If clearing is so arduous then failing to kill the boss afterward feels really bad so I’m not sure what’s worse here. You can’t make super hard bosses with a large amount of clearing before hand if it’s content that is supposed to be routine. You could but it would be a bad game.

Ultimately you need some trash clearing in a game but not too much. And it’s important that the trash be able to drop some valuable things so it doesn’t feel like a complete waste of time. VT trash had a handful of drops but not enough. Keyed raid zone trash mobs are the perfect place to put tradeskill mats for raid quality crafted loot. The trash mobs themselves should also require some thinking about how to dispatch them instead of just being sacks of hitpoints. This is another thing Blizzard did better.

Dearth of Content for Smaller Guilds

EverQuest had a lot of content for single groups and full raids but not much in-between. I feel this was a mistake. WoW solved this with 10 man raids. EQ needed raids aimed for 2 or 4 groups (depending on era). PoP tried to provide some of these but priorities meant they were poorly implemented.

One way to provide this is to not obsolete previous era raids by making single group content provide gear that is superior to last expansion raid gear.

A problem with these is that a subset of high end guild players will end up knocking them out between their main raids. Either for alts or to kill time. Something will need to be put in to stop this. I honestly don’t have an easy answer to that.

Lack of Boss Diversity

One thing I wish EQ had done more is design bosses such that different classes got to be the rockstars. What I mean by this is that a role that is typically best done by a certain class is instead, on this specific boss, done by a different class. Examples of this would be knights tanking Time Cazic Thule, or Rangers DPSing Fennin Ro. Warriors didn’t tank Time CT well because they couldn’t be fear immune and Time CT feared. Knights could become fear immune with AAs. CT didn’t hit hard enough to disallow knights from tanking it. For Fennin, melee DPS couldn’t even hit it due to rampage, and fire nukes (the best nukes) didn’t work on him. Tank transitions were also very difficult on him. Rangers with low aggro trueshot were the rockstars on that fight. That PoP had some class rockstar bosses is a reason why PoP was a great era.

Sony could have designed bosses with more variance when it came to defenses so that either spell damage or melee damage worked much better on it. (resists and AC) Sony did make spell immune stuff in Sky but they failed to make melee immune things in the zone to counterbalance. You could conceivably do something a little wild like make undead NPCs that were immune to (or highly mitigate) all damage except anti-undead spells such that clerics had to nuke and druids and shamans had to be the healers.

Plane of Sky

Plane of Sky was perhaps the least popular raid zone in the game. I’m going to give it its own critique here.

I do admire the attempt of this zone, as it’s quite novel and challenging in some respects, but I feel the execution was poor overall.

Since modelers can only make so many NPCs for the game due to labor cost, designers took every flying creature and put it in this zone. This made the zone lack its own character and it felt like a clips episode of a TV show. I can’t blame Sony too much here given they only had two dozen people working on the game but it is what it is.

My primary gripe with this zone is that it required a bunch of cheesy, borderline exploit tricks just to make it feasible: corpse summoning people up to later islands; making corpses to store keys; bind sighting across the zone to get targets; dying on purpose and using resurrection spells as teleports. This is not how zones should be designed nor should this be the “challenge” of doing any zone. Defeating NPCs should be the challenge of the zone.

If a guild did none of those things then they would be stuck with 24 people. While the zone is doable with 24 you’d need an A-Team to do the zone with 24, and this would almost certainly mean that some of your guild would be excluded. So what everybody did was corpse summon more people up there to get around the key limit, which is extremely obnoxious. Furthermore doing this zone to isle 8 for most guilds meant camping in the zone overnight because of how long it took.

Sirran was a big headache in general. People aggroed him all the time. He spawned on the death of the boss that was just being actively attacked so aggroing him on spawn was common. He assists other sky mobs. He could attack by talking to him. He wouldn’t spawn again if he was up somewhere else. Sirran kept everybody walking on eggshells.

The RvR for this zone was terrible. The problem is, the zone was such a PITA that if the rewards were all super BIS then it would make players miserable because they would be compelled to do the zone frequently. So in a way it’s good the items weren’t that great. Of course not being great means people rarely did the zone. They just did it for epics. It’s like a catch 22.

My proposed fix to the zone:

• Remove keys entirely
• The island boss drops a single no-rent quest item that when given to Sirran turns on the teleporter for 60 minutes
• The teleporters only work for the people in the group who turned the quest item in and members of the guilds these people are in, or if the raid window is in just instead use the raid ID
• Limit the number of times the teleporters will function
• Make Sirran unaggroable, or make him mem blur every time he kills somebody so he doesn’t go on a rampage
• Nerf corpse summon to only work on island one

Some kind of limit on teleporter use should be put in place otherwise a guild could just invite effectively unlimited people along to island 7 for crowns. It also acts as an anti-zerg check. Although if players chose to, they could camp out on higher islands and wait the next day for another batch of players to go up and then join forces, so even a limit as low as 24 could still work that way and this should be allowed, although I think the limit should be higher.

This would make the zone so much less of a PITA that they could then buff the quest rewards and then you have another viable raid zone that isn’t just a “sometimes” zone while preserving the spirit and feel of the zone.

Content Monopolization

EverQuest was an uninstanced game and therefore players had to compete for resources, which in EQ meant NPC spawns. We now have 25+ years to look back on to understand what player behavior will be like on these kinds of servers.

I went over possible mitigations to a lot of this under “Alternatives to Instancing” in the Good Design section.

The Play Nice Policy

In 1999, Sony/Verant apparently had a non-interventionist mindset and didn’t get involved in player disputes over content, but a year post launch that changed and Sony created and implemented a “Play Nice Policy” which included this language:

“There are cases where two or more groups wish to kill the same thing. In these cases, the groups are required to compromise. If an equitable compromise cannot be reached between the players prior to EverQuest Customer Service Staff involvement, the EQCSR will mandate a binding compromise. Refusing to abide by a compromise mandated by an EQCSR will be considered disruption. It is therefore strongly suggested that the groups make every attempt to reach a compromise that they can live with prior to involving an EQCSR, who may mandate a compromise that does not suit you to the extent that a player-devised compromise would.”

EverQuest in the early days had much more GM involvement in server affairs. A GM might just randomly show up and start chit chatting with you. EQ only had 200k subscribers a year after launch, and this smaller playerbase allowed for more player-employee interaction. A result of this is that Sony relied on humans (GMs and guides) to resolve conflicts instead of implementing server code to prevent disputes. Sony’s policy was, more or less, don’t be a dick or we’ll punish you, and we’ll decide if you’ve been a dick or not. This was quite vague and error prone.

Ragefire was a very contested mob and rather easy to kill, so older Usenet posts can be found about how contested spawn resolution worked in the worst case: GMs simply made the parties roll for the loot, regardless of who got exp for it.

For raid bosses that required a significant raid force, servers generally did one of two things: rotated them amongst the guilds capable of defeating them, or going with a First-in-Force policy. FiF meant that if a guild saw another guild gathering for the kill and had several groups there already then they would back off instead of attempting to contest it. Server rotations I think were more common than not but I have no way to verify or prove this. My server (Druzzil Ro) rotated for the most part but had periods and cases of FiF.

Project 1999

P99 is probably the most extreme case ever when it comes to competition for raid spawns and what players have done in order to get the spawns. (granted I’ve not played TLPs) It resulted in the most extreme forms of camping possible. It’s everything the original designers would hate, and yet the game design they made pushed players into this kind of behavior. This is mostly because the playerbase on emulated servers is so much more hardcore than players were in 2001. EQ was simply designed for a different audience. MMOGs need a casual majority or the server becomes toxic when it becomes crowded. If you could wave a magic wand and manufacture thousands of casual players then you could make more servers and spread out the hardcore players and the problem starts to disappear.

I’m quite sure that if P99’s player behavior had happened in 2000/2001, Sony would have strictly enforced rotations or perhaps coded in something to mitigate the problem. I am about as classic minded as it gets but in this case I feel it’s entirely fair game to implement custom behavior or rules to deal with this problem. P99 seems to have some kind of race system now but I’ve not experienced it because I played there 15 years ago. To be honest it sounds less appealing to me than standard FTE or FFA.

The Al’Kabor Project

TAKP is the polar opposite of P99. This community came from Sony’s Mac server, and it was always low population. The combination of low population, Apple, and content being stuck in era for years seems to have produced a much friendlier environment. They voluntarily rotated content without any GM mandating it. When the emulated server opened the community continued the policy of rotation. The players themselves managed all of it; staff did not need to get involved beyond threatening players to not to break the rotations.

While I would say that TAKP’s rotations were successful, there are some notable downsides to it:

The first is that because just about everything is rotated, getting in on the rotations can be difficult and require the metaphorical filling out of forms to to play the game. This isn’t much of an issue on large targets but for single group targets like Dozekar drakes, this starts to get problematic. Strict rotations lead to cartel behavior. This almost forces players to join an established guild to get anything noteworthy.

The second issue is that rotations will compel guilds to kill content that they might not otherwise, just because “it’s their turn” and so players who need or want it more and would have otherwise gotten it are denied. On a FTE/FiF server the guilds who want it the most will more often get it.

Even if a guild might not want to kill something, because it’s older era content, a subgroup of super enthusiast guild members will consume the content for alts, bazaar loot or even streams. This denies older era content to new and casual guilds and players. A good example of this would be Tormax which is killable with a single group (and thus two players) on TAKP, and has good droppable loot. Rotations make it easier for the enthusiasts to content hog.

The Alt Problem

Players having stables of alternate characters creates a number of issues. This becomes more of an issue the older a server becomes. This problem is more pronounced in EverQuest than it is in other games because of EQ’s slow paced nature allowing for easy boxing and EQ’s permissiveness allowed for powerleveling or ‘boosting’ in a way other games don’t allow for.

Alt stables are bad because they allow players to get around restrictions that force players to share content. Multi-boxing allows solo players to camp things they couldn’t otherwise. Alts allow players to get around lockout restrictions. Alts motivate players to consume content they otherwise wouldn’t because their primaries had no need of what comes from it, or it allows them to “double-dip” and funnel items to their primaries. Hardcore players will fight tooth and nail to get that item for their 10th alt and beat some casual player to it. Alts allow players to skip “eating the vegetables” before getting the desert, i.e. by skipping VT trash with mage coth bots or simply camping out a character deep in a dungeon some place. Alts reduce or eliminate the core interdependency that is central to the game. If you can swap to any class that is needed at the time, then you have less need to guild somebody of that class and it starts getting confusing as to who is on what characters.

Guild Competition

Some amount of inter-guild competition and rivalry is often desirable. But like anything else needs to be in moderation otherwise it quickly becomes toxic.

Years ago around the time TAKP launched I came up with an idea for emulated servers to make contesting raid spawns more cordial. (I’m sure somebody else has come up with a similar idea but I’ve not seen it) My idea was to implement a claim system whereby a raid can claim a raid target and get exclusive rights to it for a very limited time. If anybody outside the raid tried to interfere they would be removed from the zone. The catch would be that you couldn’t add more players to the raid after you did this (not until the mob died or the timer expired), players in the raid but not in the zone would be kicked out of the raid and you had to beat the timer. If the timer expired the raid gets teleported out of the zone and they wouldn’t be able to claim it again for some time. The result is that when two guilds are mobilizing for a target then the leaders need to make a gamble as to when they believe they had a large enough force to win. This would result in some occasional failures by the first claimant guild and so the 2nd guild would then get some uncontested kills even when they mobilized slower. (although the boss should probably full heal after)

If multiple guilds are camping a dragon’s spawn point before it spawns then the server could simply select one of the raids at random for the rights to kill it and teleports out the others when it spawns. To prevent poopsocking, the spawn timer could be easily known to the players either by being static, or the server could just tell them on request or a serverwide announcement could be made some amount of time (e.g. an hour) prior to repop.

I believe that any sort of system or rules needs to be relatively simple and easy to follow. You shouldn’t need to be a lawyer to play a game. Ideally the server and possibly client should enforce the rules such that nobody needs to read anything at all and GMs don’t need to get involved. This might be as simple as a player may not engage a raid target for X number of days after they’ve participated in a kill otherwise he is teleported or death touched. (i.e. a lockout timer without instancing) If X is longer than the target’s respawn time then it’s a very simple and effective way to force sharing. The value of X doesn’t need to be static and could increase as the number of guilds contesting spawns grows.

It’s ironic, but emulated severs have more freedom of movement when it comes to enforced content sharing. Emulated server devs can basically run the joint with an iron fist if need-be. I have a harder time seeing a commercial project mandating that users not circumvent lockouts with alts or multiple accounts for example.

Simplistic NPC AI

A complaint I have with MMORPGs in general and not just EverQuest is that the NPC AI is rather dumb. Of course when EQ was new, the servers were probably running on Pentium II machines so they have a good excuse. EQ did have some limited parameters to make NPCs behave somewhat differently however. Undead and animals would ignore most of the hate of the first spell cast on them. There was also a parameter to make NPCs prefer closer players to farther players. The NPC’s intelligence stat could modify NPC behavior to some limited extent but I don’t know how.

NPCs should be less predictable. You shouldn’t be able to know with absolute certainly how the NPC will behave. Smarter NPCs should be more willing to go after casters. NPCs should be more willing to flee if they’re not snared and facing certain death. NPCs in one dungeon should behave somewhat differently than in another dungeon. Smarter NPCs should grant more experience because they’re more threatening. EQ randomizes a lot of things but behavior is not one of them which is honestly rather odd considering how disuniform real life people intelligence is.

Smarter NPCs make the game more engaging. If they’re less dumb then they don’t need a truckload of hitpoints. In a long grindy game you really need to avoid making everything extremely repetitive and predictable. It’s also a bit immersion breaking if your humanoid NPCs are completely brainless. Tactics like kiting were overly strong in part because of the dumb AI.

Exploitability

EverQuest unfortunately was prone to being exploited. A 3D MMORPG had not been made before so it was only natural that some unexpected behaviors would arise. EQ was also more permissive than modern games, and features like making corpses created a rather large grey area. In contrast, games like LOTRO made me feel so confined into a box that it felt suffocating to me.

In the old days Sony/Verant had much less tolerance for exploits. They at one point publicly suspended a guild using creative tactics in Sleeper’s Tomb. Legacy of Steel was suspended for pulling mini Aten down to Blob 1 room. Some members of my own guild were suspended for transferring no-drop items to alts at one point. I miss the days when administrators cared about the integrity of their games.

Cheat programs for EQ have had many years to mature and are now a cancer on the game. Players are endlessly frustrated by bots and automated multiboxing on the Live servers. Emulated servers have static binary clients so the cheat programs never need to be updated with new offsets. It’s really nightmarish for anybody who cares about the integrity of the game. A large reason why I never made my own emulated server is because I feel the cheat programs have become too rampant and this kills the enjoyability for me.

I fear that multiplayer games may become infeasible in the near future as AI will facilitate cheating and automated play to previously impossible levels. Even forcing the client to be played with video streamed clients a la Stadia can’t entirely solve the problem because AIs can simply read the video feed. Multiplayer games may actually return to the days of being small hobby projects because when games reach a certain popularity, some player will just develop a cheat or AI to ruin it. Games will have lifespans and die once the cheats come out, or they’ll end up being akin to a private torrent tracker and be invite only with player vetting.

RMT (real money transactions/trading) is another huge problem that is a natural consequence of EQ’s design. If a game has rare tradable loot then RMT will manifest once the population reaches a certain level, particularly if the items require enduring tedium to obtain. To my knowledge RMT has not been an issue on TAKP because the community was so small there, but on P99 it certainly has been a problem. The big issue with RMT is not that players get unearned advancement, which is also bad, but rather that it will result in content monopolization and legit players being unable to advance because people are being paid to prevent them from accessing content. It’s highly toxic for the game.

Multiboxing

Users playing multiple characters at once is another scourge of sorts EQ has. It’s a natural consequence of the slower paced nature of the game combined with the interdependent design. Modern computers being able to run multiple clients compounds the issue.

Boxing is one of those things that people think they want but really don’t. TAKP voted to be a three box server and years later much of the community wished it had been only been two box. Most users simply do not have the foresight to understand the consequences of allowing it, and they only think about how great it will be to never have to talk to anybody in a game designed to be social.

Boxing is something akin to cheating, although not cheating. It does make the game much easier, even at the raid level. A simple example is leveling an ungrouped warrior and using a boxed cleric to heal it. The first level 60s were players who boxed healers while camping guards on their mains. Boxing allows players to camp things by themselves they otherwise could not, thereby denying the camps to groups of single boxed players. At the raid level boxing CH chains makes them significantly easier and keycloned armies of wizards allow a single player to output multiples of typical DPS with zero extra input.

On the flip side boxing allows raids to box a job that would otherwise be boring or tedious and not require a dedicated player to do it. Enchanters were buff bitches at certain points of the game for example, so while I prefer one box servers I do think there are some circumstances where it’s not so bad.

EverQuest’s raid design is actually not very difficult mechanically. The real difficulty in EQ’s end-game is simply managing all of the players needed to down bosses. In a sense EQ is a logistics and coordination game. Boxing is cheating the game when looked at from this perspective. There have been cases of players boxing upwards of 18+ numbers of characters to down raid targets, and not needing to interact with other players to kill bosses in a game designed with socializing at the center of it will create a very negative experience for the other players.

Unfortunately I think the best defense against boxing is to make combat more fast paced and require the player to press more buttons. I would prefer that some slower paced games existed but the slower the game the more boxing that will manifest. Boxing is very hard to detect. Emulated servers try to scare people into not doing it but in reality it’s just not caught very much.

Monetary Units & Item Value

EverQuest had four monetary units: copper, silver, gold and platinum. Sony made a big mistake in pricing the items in the game, because most items are worth at least one platinum and thus the other units had little meaning and these other coins weighed the player down and were often just deleted. EQ’s item valuations and coin drop rates should have been one tenth or less what they were such that the value of a coin becomes ten times what it was in the game. Level one players could get a platinum coin by selling staves dropped by level 1 skeletons and this is silly. Players shouldn’t be getting platinum coins until mid-game.

Skills Did Too Little Damage

Other than Backstab, skills (e.g. Kick) did very little damage to the point you could just ignore them unless you were a monk or rogue. Although monk skill damage was also rather low, and by PoP era a monk’s kick damage was less than 10% of his total. Backstab was fine because it scaled with gear but the other class skills did not, which was the problem. This is largely why rogues outpaced monk DPS. It would have been much more interesting if the other skills increased their base damage based on some stat or stats of boots or gloves. (although ATK does work on all of them)