EverQuest: 20 Years of Retention

**This blog was original posted on Gamasutra.com

This year, EverQuest reached the nearly unprecedented milestone of 20 years of active development. We have players who started in 1999 and are still going strong today.  As a game designer, I’ve been able to study and analyze how EverQuest has achieved 20 years of retention.

If you’re familiar with the proposed principles of player motivation within game design, like those presented by Richard Bartle, Quantic Foundry, Nir Eyal’s “Hooked” model, and countless others, you can see the theories working within the design of EverQuest and we can observe their impact over two decades.

Lesson Learned #1: Timing is Everything

One of the most important things to keep in mind when looking into EverQuest is the time it was created. 1999 wasn’t like today when the phone in my pocket gives me immediate access to the internet and innumerable apps that let me play games with friends or jump on Reddit to share my Lord of the Rings fan fiction.

In 1999, I had a gigantic and very heavy CRT monitor plugged into a Voodoo 3 video card. I was on a trial version of AOL (America Online, to all you youngsters) to access the Internet. It got cut off every time my mom picked up the phone to make a call. On my desk was a stack of Q-tips to clean the dust off the rollerball contacts inside my mouse. When I’d log on to the Internet at 1 a.m., the dial-up sound would wake up my dad and he’d bust into my room and yell at me to go to bed (and get a job). Even with those struggles, I was lucky to be part of the 4.1% of people worldwide who even had Internet access.

Figure 1. Growth of worldwide internet access from 1999 to 2018  [1]

Figure 1. Growth of worldwide internet access from 1999 to 2018 [1]

EverQuest was an ambitious and genius game, but it also had superb timing, and I don’t just mean market timing. Yes, EverQuest beat Asheron’s Call and Dark Age of Camelot to market, but all those games benefited from the same timing that EverQuest did. The dawn of the Internet brought the shrinking of the world so that communication from anywhere was instant and always available.  This fundamental change in the way people socialize is the catalyst that paved the way for EverQuest’s success. And its creators felt it.

For a generation of gamers, EverQuest became, and is, a solution for many people’s social needs. Large scale MMO adoption is cyclical, as new generations come online and take to MMOs for the first time, this moment is recreated. Look at World of Warcraft and Runescape. Each is the EverQuest for that generation. And that is the first lesson of EverQuest’s 20-year retention: Timing and purpose are everything. Find your novel vision that delights players by filling a need that isn’t be met.

Lesson Learned #2: Stay Focused on Player Motivations

This sense of community and belonging was exactly what I needed at the time and EverQuest’s game design provided it. I was (ok, still am) weird. Defined as a dork. I loved fantasy novels, video games, and making GeoCities webpages with obnoxious flashing text. I was a year out of high school, living at home, and didn’t have many friends or a solid life direction (or a job, sorry Dad). I was far from living as my authentic self, because like most 18-year old kids at the time I had no idea who that was, and Oprah had yet to explain it to me (she had not reached enlightenment yet).

One fateful night, I went over to a friend’s apartment (he had a job) where he was playing EverQuest and he explained the game to me. I did not believe for a second that all those characters were real people. So, he started a conversation with the classic “a/s/l” opening. His entire guild replied in bright green text and I was dumbfounded! Baffled!

I then did the next logical thing. I bought the game and started a Barbarian Shaman, got to level 22 to get my upgraded “Spirit of the Wolf” run-buff spell and started peddling it throughout the game for Platinum to fund my actual main, a Wood Elf Ranger who spent most of his time face down. I also joined a guild and even though I died all the time (Rangers, lol), I was the truest, happiest version of myself.

If you saw EverQuest’s 15th and 20th Anniversary videos or watched the EverQuest Show on YouTube, you will see my story of emotional connection to not just EverQuest, but its community, isn’t uncommon. EverQuest is a place where people can satisfy the top three needs in Maslow’s Hierarchy: Social Belonging, Self-Esteem, and Self-Actualization[2]. More than that, it’s a comfortable home for thousands of people who don’t NEED to be in combat every moment. They relish the human/avatar connections as they “live” and evolve in the structure of EverQuest’s fantasy world.

We played EverQuest because every time we needed to connect with someone or share news or needed to be acknowledged for being good at something, the game and its players satisfied those needs.

In EverQuest we found connection and achievement. We found competition and mastery.

And those are still the primary player motivators the game is fundamentally designed around today. Over the 20 years that EverQuest has thrived, the team has added countless systems, mechanics, features, and an astounding amount of content. Were there missteps? Sure.  But the team doesn’t add or remove anything that fundamentally changes the motivations to play the game or upend core mechanics that players have spent years learning. And that is the next lesson to retain players for 20 years: Stay true to the player’s core experience and motivations. Protect and double-down on what your game really is to those that are actively playing it. Gauge all your actions by theirs. 

Lesson Learned #3: Change with Engaged Players

This isn’t to say that the game hasn’t changed. It has and in very significant ways. A good example is affectionately called “corpse runs.” If you ask someone who stopped playing EverQuest 15 years ago what they remember, inevitably someone will mention a memory of dying to some mob, losing their body and all their loot, with a dreamy smile on their face. Instinct would say, “Corpse runs must remain forever, untouched!” But memories are a tricky thing. Nostalgia paints a picture that is far different from the reality of the players who remained engaged.

The Planes of Power expansion for EverQuest introduced graveyards which started to make corpse runs less punishing as corpses would land in the graveyard locations in a zone. Eventually the item loss component was removed and replaced with XP loss that could be recovered by going to your corpse. There are several reasons the team removed the item loss component from death, and the primary reason is not because a bunch of “care bear” designers thought corpse runs were too mean.

In part, it was because as items with different rules and behaviors were added to the game, like teleportation items, and the total number of items a player could carry increased, there started to be significant technical issues. For example, when you’d loot your corpse with tons of items there would be tremendous client lag, making what was already a harsh experience exponentially less fun with each lost frame per second. And those teleportation items started to create edge cases around which items get left on your corpse and which don’t. And, in rare cases, all the engineering around items on corpses created situations where corpses could be duped or even worse, lost all together. (Bonus lesson: Stability is king, bugs kill retention and every other important KPI faster than anything else.)

Just as importantly though, lost items from death as a mechanic was no longer serving player’s motivations. At this point in the game’s lifecycle, zones were bigger, the number of items a player carried was much larger, and the overall time investment to get those items was incredibly high, skewing the risk-reward ratio all the way towards all risk, no reward. As EverQuest grew, the game no longer needed the additional risk of losing items for players to feel a sense of mastery and achievement. Lost time creates enough fear and focus and is more palatable than losing items you may have worked months for. Losing items made many give up and quit. Lost XP is measurable as time and playing. In a player poll several years later, players agreed with developers by voting against the return of items on corpses – a question raised by those who were nostalgic and not necessarily playing daily. And that is the next lesson: Acknowledge and change with your actively engaged players. Take stock and reset your view to match the players’ who are actively playing today.

Lesson #4: A Labor of Love

Understanding and respecting players’ motivations and expectations can only happen when you have a team that understands them. I’ve been a member and lead of multiple live development teams. Some were more successful than others at retaining players. The most successful team I’ve been lucky enough to be a part of is this EverQuest team.

Development of EverQuest is a labor of love from the Executive Producer to the newest Associate Programmer. Every member of this team has their own personal story about how EverQuest, as a game, has impacted their life. Many of the team members have worked on EverQuest for 10 or more years with most of the art team on the game since launch.

This intuitive understanding of what “feels EverQuest” in every aspect of the game is priceless. When some new team member comes along and throws out random ideas (yes, I mean me), the team can quickly filter them for what is good and bad through the lens of a current EverQuest player. Not only does this save time, but it keeps us on course. New skills can be taught and mentored, but a true love and understanding of the game is invaluable. That is the final lesson: Staff your live service team with developers who love the game.

In the end its simple. Respect your game. Respect your players.

Systems Pro's & Con's: The Division

All analysis is based on the following high-level assumptions:

  • Core loop = quest -> explore -> combat -> improve (stats)

  • Short-term engagement = consistent, measurable power rewards

  • Mid-term engagement = difficulty modes that push the player back into the short-term cycle

  • Long-term engagement = create multiple characters

  • Primary player behavior motivator: Achievement through power

Pro’s 

Mission delivery: The implementation of the map-based mission delivery system, in conjunction with unlocking missions by using boards in those areas keeps the sense of large scale while reducing the overwhelming feeling of a map exploding with things to do. The waypoint system leading through the open world to the objective is clear and useful, keeping the player from feeling lost in the streets of Manhattan.

Co-op roles: Each skill line provides clear and familiar role specialization into Tank, Support, and DPS, which makes co-op more rewarding and necessary.

Base: The base is a fantastic and motivating way to track the player's progression. They get to not only watch a meter fill, but also see the changes reflected in the world. This serves to immediately immerse the player in the narrative and provides a method of progression outside of power that is satisfying.

Dark Zone: MMO logic dictates that the most satisfying reward is showing off your progression to other players. The Dark Zone is a unique way to express and show off your power.

Cons 

Base: From what I've been able to gather, while a successful mechanism for expressing the player's progression, I think this feature was under-developed. Because it was mostly used as a location to spend skill points, it ended up being more of a hassle to have to return to then the centralized motivator it could've been. If you compare this base implementation to the X-Com HQ or Dragon Age: Inquisition's keep it is pretty shallow and superfluous.

New player difficulty: The Division is a pretty unforgiving tactical cover-based third person shooter. Based on this basic barrier to entry, I think they failed to provide enough starting options to players of various skill levels to feel successful.

Power curve: In its current implementation, with the type of gameplay the game has, which is highly skill-based, I would argue that power progression is a mismatch and should be removed. The moment to moment challenge focuses on managing positioning in relation to the AI, not the configuration of your gear. All the stats make the effect of you being good at the game better, but do not compensate for you being bad at the game. In my opinion, if you want your game to be about getting items, those items need to be a significant portion of the method of your success with skill taking very close second place. Borderlands does this expertly. This mismatch is one of the reasons the game gets complaints about the AI being bullet sponges.

Improvements

These suggestions assume that the design intended for power-based loot to be the core reward mechanism.

Power curve: For me, starting player power is skewed too low and scales too slowly. This causes the early pacing of the game to feel slow and a bit punishing. I'd skew in favor of the first 5-10 hours being relatively easy and ramp difficulty up by decreasing the power ramp over time, while introducing new mechanics or stats to make up for the decrease.

Resists and damage types: One of the mechanics I'd add over time is damage types and enemy resists. These don't necessarily need to be elemental and fantastic. They could put in a simple system, like Mass Effect 2 & 3, where there are 3 resist types and weapons and abilities counter those. This would allow some forgiveness for the bullet sponge feeling since there would be context and a counter for it.

Enemy durability expectations: With a game that feels as skill-centric as this one with human enemies, I think they needed to reinforce the time to kill through the visuals of the enemies. When I see an enemy wearing a hoodie and jeans, I expect two or three bullets and definitely a headshot to take him out. If it takes multiple headshots, he should at least be wearing a helmet.

Increase weapon types & mechanical variety: Their weapon selection seems limited. I'd add another category of weapons, like rocket launchers, crossbows, or energy weapons to give more playstyle options, feel variety, and spectacle. I'd also move out of the realistic and into weapon mechanics that are fun and interesting, not as far as Borderlands, but maybe a sniper rifle that charges up.

Perk, Talent, and Ability upgrading: To increase the tail on progression, I'd add upgrade levels to the player progression systems. Upgrading these would require the same currencies used to unlock them in the base.

Base: I think the base could've been made more integral to the game by including it in content. If, for example, the base was under threat of attack, then you could add defensive and offensive progression leveled up by items and currency and have events where enemies attack it either synchronous or asynchronous. They could also add more flavor touches and customization options that sink currency. Basically, extend the functionality to tie it deeper into the reward loop. I'd also evaluate if there should be duplicate functionality outside of the base, like merchants in other places in the world, to make going back feel useful.

Systems Pro's & Con's: Borderlands 2

All analysis is based on the following high-level assumptions:

  • Core loop = quest -> explore -> combat -> improve (stats)

  • Short-term engagement = consistent, measurable power rewards

  • Mid-term engagement = difficulty modes that push the player back into the short-term cycle

  • Long-term engagement = create multiple characters

  • Primary player behavior motivator: Achievement through power

Pro’s 

Procedural item generation & manufacturers: Like Diablo 3, Borderlands 2's primary reward loop is driven through the itemization. Which again makes procedural item generation key for the same reasons. The manufacturers concept to bucket the feel and playstyle of weapons makes the system still feel custom and rooted in the narrative. In addition, the wide variety of weapon mechanics, even the hokey ones, are necessary to make sure the loot matters.

Elemental Damage Types: When combining power with a skill-based game, the biggest challenge is to minimize the 'bullet sponge' feeling when fighting humanoid enemies while still providing a challenge. Elemental damage types and enemy resists reframe the 'bullet sponge' to instead be communication to the player that they are not using the best possible strategy.

Whenever you are shooting an enemy forever in Borderlands you know you should try a different weapon. This of course also drives the player back into the itemization loop to get the right weapons to face the challenge at hand.

Class skill trees: The secondary reward loop of leveling up to unlock skills is very successful in Borderlands, because shooter game players generally have playstyle preferences/needs, which are supported in not just the base classes, but in the how they specialize over time. These also make it interesting to try new ways to play the same class and try out other classes. Each also assists in co-op play to create roles that allow for synergistic teams.

Cons

Overall, Borderlands 2 was a really tight game. The team did a great job editing and only including systems they needed to support the game loop.

Vehicles: Although useful for co-op play, vehicles overall were more of a pain and interruption to the flow of the game when required for specific missions.

Improvements

Weapon Attachment/Mods Crafting: Borderlands embraces randomization completely, but this is somewhat dangerous for engagement in a shooter, where skill and pre-determined playstyle preferences are big factors to feeling successful and smart. Adding an aspect of itemization agency for the player through crafting weapon customizations could help give players an avenue to get exactly what they want and reduce grind fatigue when the RNG gods aren't on your side. I'd allow for crafting of scopes, silencers, mag extenders, etc. I'd also consider adding mod slots to weapons or the character that compliment their chosen playstyle. Mods would give passive effects, like fully reload a sniper rifle after N consecutive headshots. Or a damage modifier when attacking from above an enemy. This would give even more items to collect and methods to further specialize your playstyle. They could also help double down on a role for co-op.

Vehicles: I'd either remove the driving of them and make them AI controlled making the player a gunner or spend more time tuning to make them super fun. If time is spent on them, they could become a more interesting component in the game by giving them progression and item slots of their own.

Hats!: I'd add far more customization options and unlocks for the character's looks. Overwatch style skins that unlock through progression, achievements, and Badass Rank would give another motivator to engage in those systems, especially when the player has collected their near ideal playstyle loadout.

Systems Pro's & Con's: Diablo III PS4

All analysis is based on the following high-level assumptions:

  • Core loop = quest -> explore -> combat -> improve (stats)

  • Short-term engagement = consistent, measurable power rewards

  • Mid-term engagement = difficulty modes that push the player back into the short-term cycle

  • Long-term engagement = create multiple characters

  • Primary player behavior motivator: Achievement through power

Pro’s 

What makes Diablo's loot and progression successful is variety and build customization in combination with the frequent pacing of rewards, which are best achieved through the following systems.

Procedural item generation: By bucketing stats (of the bear, hawk, lion, etc.) and then randomizing additional attribute additions they create thousands of items that (mostly) provide meaningful choice for the player. In addition, level scaling and rarity creates even more variety, by extending the stat and attribute budget of each item. Without this method of itemization, reward frequency would end up being reduced significantly due to lack of meaningful items.

Class ability runes: Abilities and passives are standard in RPGs, but Diablo 3's ability runes extend the system by allowing the player to specialize even deeper into a role and playstyle and change up the pacing of combat without having to relearn an entirely new ability. Because of this, unlocking a rune on level up is nearly as rewarding as unlocking a new ability, but without the overhead for both the player and developer.

Paragon levels: Diablo's relatively long player retention is based on reengaging the player into the power acquisition loop by making new characters. The Paragon system incentivizes this by granting Paragon points across all characters, giving players credit for the time they've already invested and a nice leg up in power.

Kanai's cube: Item attachment is Diablo's enemy, if the player becomes too attached to a specific item they are disincentived from participating in the itemization loop and are likely to feel that power is being taken away. The addition of Kanai's cube, which allows a player to equip the defining ability of a legendary item directly on their character, turns that around by making all legendary items always usable and even more desirable to collect.

All the variety would be meaningless without appropriate challenges to test and feel it. The following systems provide that challenge:

Difficulty modes & Enemy scaling: In addition to scaling enemy health, damage, and resists by level and difficulty mode, beefier elite versions of enemies with new or more elite modifiers are added based on the selected difficulty mode. This addition not only provides more chances to feel their power but requires the player to analyze their build and alter their playstyle providing new opportunities for the player feel smart all over again from mode to mode.

Greater Rifts: Advancing difficulty and rewards based on a player's speed through content is the ultimate test of the player's power and build. By tying access to legendary gems to this content the player is highly incentivized to return to the power loop to achieve ultimate success. By requiring greater rift keys, the player sees this content as a reward of its own. 

Cons

Crafting: While I understand why salvaging and crafting are in the game (so all items always have a purpose, promote downtime, give agency in a random system) I think the implementation of weapon and armor crafting didn't contribute as much as it could have. With so many and so frequent drops, armor and weapon crafting feels mostly irrelevant.

Improvements

Paragon System & Difficulty Modes: With the current implementation, the developers must keep adding new difficulty ranks (up to Torment XIII now!) because there is no mechanism to reset a player's power progression. They have attempted to address this with Seasons, which drives re-engagement with the game, but the axiom of 'feed the beast' will always apply in this paradigm, as they must add increasingly powerful gear and enemies to keep the game loop going. I'd try redesigning the Paragon system to function more like a Prestige system in Call of Duty, where at max level the player can opt-in to reset their progression to level 1. Upon reset, they would receive a Paragon point and something cosmetic. The bonuses from the Paragon points would become larger to offset the time investment. Ideally, this would level scale currently equipped gear, so the player doesn't feel like too much has been taken away and that would be a fun benefit if you have really cool gear equipped. The reset would prevent the need to scale into crazy high numbers (like 2108607% Health) through the lifetime of the game and push the challenge in difficulty modes away from just stat modifiers and bonuses and into adding more interesting mechanics or new enemies that challenge the player in different ways.

Resists & Elite Modifiers: While there is some amount counter itemization required to take on specific enemies at harder difficulties, it's a minimal requirement. Getting the player to reconfigure their gear to counter specific challenges more explicitly could driver deeper engagement into their itemization loop. I'd experiment with making more specific damage type resists a part of the elite modifiers and boss fights. Then maybe compliment that with a new type of gems that apply damage types to weapons.

Crafting: To give crafting a clearer role, I'd try out bucketing all gem items into solely being acquired from crafting and remove crafting of non-set weapons and armor. So, enemies would not drop gems instead components salvaged from items would be used to craft and upgrade gems. This would clean up the recipe lists and make the role of crafting more important.

Systems Compare and Contrast: Diablo III & Fallout 4

The differences in systems between the two games stem from the high-level choices relating to how players are motivated to engage with the games. Both are great games that have diverse players, because they each satisfy all the primary motivators. So, you must examine the secondary motivators. Using the Quantic Foundry player behavior motivation model, you can map each game's secondary motivators out like this:

motivation.PNG

Of course, this analysis is a bit black and white, as there are definitely shades of grey here. Especially, when analyzing across the achievement and mastery motivations, since both games do rely heavily on feeling power growth and build strategy.

But because each game emphasizes different secondary motivators, you see many of the same systems and loops with significantly different implementations. The below chart lays out a few examples of the differences across all the motivators:

motivation2.PNG

Systems that exemplify the similarities and differences

Stats & Power Growth: The application of stats in each game increases power, but because of Fallout's shooter combat style, they chose to make the amount stats effect the outcome of an encounter far less than Diablo where every fight can be overpowered through gear.

For example, the highest base DPS of a weapon in Fallout 4 is 726, but you still must aim accurately to achieve that DPS. In Diablo 3, I've seen weapons that are 3,000+.

Fallout's stats also increase secondary attributes, increasing their utility value in addition to their power value. For example, strength increases carry weight and melee damage. Whereas, Diablo 3's strength affects 2 combat-only stats, damage and armor for Barbarians and Crusaders.

Leveling: Levels in both games are attained through earning XP. On leveling in Fallout you receive a perk point which is spent in a tree. Each unlock in the tree gives bonuses that tailor your character to a playstyle. Points can be spent in multiple areas to create a character that plays exactly how you prefer. 

In Diablo, levels unlock class abilities, passives, and ability runes. These unlock in a predefined order. Because of this the player's agency is centered around which they choose to equip to satisfy the specific role or ability combo/synergy they want to play.

So, in Fallout, I can be a stealthy melee guy or a stealthy ranged guy by swapping my weapon and spending points in perks. In Diablo, I can be a defensive wizard or glass canon wizard, depending on the abilities I've selected.

While both systems award leveling with giving the player more tools to play with, they achieve it in different ways that supports their individual game's needs.

Weapon customization: In both games weapons can be customized. In Diablo, the player equips gems on their weapons or rerolls their stats to generally increase the amount of damage the weapon does, but because weapons do not scale damage to match the challenge being presented that weapon will eventually need to be replaced. Because of this, Diablo avoids any systems that drive attachment to weapons, otherwise they would risk players feeling unsatisfied with the reward loop.

Fallout encourages weapon attachment by allowing players to craft different mods that can be equipped on an item that not only increase DPS but customize the weapon to suit the player's personal playstyle and needs. In a game that requires shooter skill to be successful this is highly important, otherwise a player feels they cannot be successful because the game isn't giving them the option to play how they want.