In MMOs, on opposite ends of the spectrum there are two models for class design:
- Each class supports very high customization, where the player can invest in particular skill trees / talents / abilities to support their preferred playstyle and capabilities. E.g. WoW, RIFT, SWTOR, WildStar
- Each class allows zero functional customization, any customization is strictly cosmetic (skins, etc.). E.g. Overwatch
For years, the high customization model was the industry norm, starting with WoW and later reaching its zenith with RIFT in 2011. In RIFT each of the 4 classes – cleric, mage, rogue, warrior – had 9 (yes, nine!) talent trees, and you could mix-and-match any 3 trees and decide how to spend your talent points in your chosen 3 trees. I loved the RIFT’s class design – there were so many different combinations and I had a field day trying to create different viable specs and builds, including my M*A*S*H Melee Healer Cleric spec for PVP.
One obvious issue with the high customization model is balancing trees and classes, both within a class and across classes. This was very true in PVE for classes that could play the same role, e.g. tank or healer or DPS.
Exacerbating objectively real balance issue are two factors:
- The ability or inability of players to understand class mechanics and make sensible builds
- The popular misconceptions on class balance. There were times when the community incorrectly believed that specs were non-functional/underpowered or overpowered, when in reality players simply hadn’t sorted out yet how to play them correctly or counter them.
A good example of a misconception of balance was Prot Pally in early Cataclysm. When the Cataclysm class changes were announced, many WoTLK veteran paladins wrote off Prot as a viable PVP spec. I was able to sort out how to make Prot work based on the new mechanics and abilities, even though I had no prior endgame PVP experience with Prot Pally. Sometimes you just gotta think through the mechanics and experiment yourself.
However, there were definitely cases where particular classes and builds were stupidly OP, as was the case with DPS Spellslingers and DPS Warriors in WildStar. The balance was so pitiful that Spellslingers and Warriors were selling carries to 1800 rating in 2v2 Arena for 50 platinum, and it screwed up the economy and created a context of have-and-have-nots, not based on player skill primarily but rather heavily influenced by selection of character class and willingness to spend RL money. I managed to reach 1800 Arena rating with a non-FOTM spec, my “Melee Mage” Esper, but it was frustrating having to face players with faceroll specs, especially when they had already acquired a full set of faceroll Arena gear (gear scaling is a story for another day).
I believe that striving for 100% balance is a theoretical goal and not a practical one. In reality, if the 10% overpowered specs/classes and 10% underpowered specs/classes at any point in time are gradually tuned, a game can have reasonable class balance. In many cases it’s fairly obvious to highly-rated players which trees and abilities are problematic from a balance standpoint. And of course with telemetry and analytics, a developer can do the number-crunching to identify outliers in terms of measurable performance and which classes/specs are prevalent at higher ratings.
In recent years, especially with the mass popularity of MOBAs, we’re seeing more minimal-or-no customization games, where a given hero/class has predefined capabilities. Some MOBAs have the in-match customization where you spend points on things as you level up (which I’ve never liked, I find it tedious). Overwatch has taken no customization extreme – each class is 100% fixed in terms of abilities.
On the one hand, fixed classes makes it much easier for players to understand how a given class should be played, since they have zero control over how the class is designed – the issue then becomes playing it correctly. Whereas with the high customization model there are two hurdles: player understanding of what is a functional spec and playing it correctly.
I believe these two hurdles are less of an issue in our current world, now that many highly-skilled players stream high-rated PVP on Twitch/YouTube, talk about how to play classes, and discuss what balance issues exist. These players are the experts.
I would so love to see more games with class customization as rich as RIFT’s system. I love dialing in how a spec works, kind of like making cookies with the ingredients you like, versus having a fixed menu of cookies to choose from.
Which model of class customization do you prefer and why?
Taugrim's MMO Blog 
I think the first problem is you put World of Warcraft on one end of the spectrum. If anything its moved away from customization and more to homogenization. The funny thing about it is they did so because “choices” could be “bad” or “good” and the majority of people ended up following the leaders (and thus choice was irrelevant).
I’d suggest taking a look at https://stealthed.blog/over-a-quattuorvigintillion-problems-and-the-skill-tree-aint-one-of-em/ to see what the real customization -> no customization spectrum should actually be. The titles you indicated are more centric then on one end or the other.
The fact that you even made this case shows how far the MMO genre has fallen in recent years.
Easier = Better = More Casual Players = More Money.
Follow the money and you’ll see why. Most of the titles you listed had Publishers. Hmm. Coincidence?
First of all, GREAT article. Really enjoyed reading it, especially as I haven’t been following Crowfall that closely.
I think you’re absolutely right.
Games are getting simplified in terms of mechanics, which I think is a shame. It’s the rich complexity that makes a game interesting.
Thanks mate, I know it was a bit in depth and slightly off topic, but its a good step to understanding why so many people live in this world where WoW is the extreme end of anything.
The problem is WoW does such a fantastic job of hugging the centric model when it comes to game design. It does everything to appeal (or at least to not offend) the majority of people.
Hate when I see it cited as anything but that.
Thanks for the comment though!
I definitely prefer the rich customization that MMOs of old offered, over the more cookie cutter style. It’s kind of the game within a game that I really enjoy. I used to play a lot of WoW, and while some parts of that game have gotten better over time, I feel like it kept getting less interesting for me as the talent trees and ability reworks kept making things simpler.
I wonder what motivated Blizzard to streamline talent trees.
Was it to simplify things for players? For Blizzard as the developer? Both?
Part of it has to be the added strain on the designers. Back when Blizzard introduced Death Knights, each of the three talent trees was capable of tanking or DPSing. I am always one to go with the oddball spec, so I ended up doing “Unholy” tanking. There was so much variety though, that I could approach the same job in different ways. They called this out as the reason they changed Blood to the tanking spec, while Frost and Unholy became the DPS specs; Death Knights essentially had 6 different specs they had to worry about.
I’m sure there is pressure on the financial side too, to make things simpler for new players, and trying to increase those sub numbers. Before WoW, MMOs were relatively small niche products, and perhaps that niche was really tolerant of, and actually desired the complexity; where the masses instead want more easily digestible systems.
Off topic, but that article you wrote years ago about the purely horizontally scaleable MMO has always stuck with me, and if feels as if things are trending that way, slowly but surely: the max level boosts in MMOs, zones that scale to your character’s level, like in the Legion expansion. Just take away the levels, and you’re almost there.
That particular case sounds like poor design. It would have been better for each tree to have a distinct role. Having 3 trees with the same roles (tanking or DPS) sounds cumbersome.
This is very insightful.
Many players just want to play and have fun. They don’t want to have to study and or research on how to build a class.
You could say it’s no different than car models. It used to be back in the 80s for American car manufacturers, you could pick from a zillion options. The Japanese auto makers (Honda and Toyota) instead offered 3 packages (economy, nice, nicest) for a given car model, e.g. DX/LX/EX or whatever it was. The latter seemed to be much easier for the consumer to deal with, and it streamlined operations for the auto supply chain as well.
Horizontal scaling via incomparables has always been my dream. Give us lots of reasonably balanced options, have the game be about unlocking/earning them.
I personally think it’s a bit of both.
In my opinion, the prevalence of the “internet community”, creates problems for complex class design. Information is so readily available, and since WoW, players have become accustomed to relying on guides, especially when it comes to MMOs.
Even outside of MMOs, citing one of my favourite games, Path of Exile, which gives the players a lot of options for how to build a character, most follow build guides, a new player asking for advice on the forums or reddit etc is usually met with “Find a build guide for the skill you want to use”.
This creates a conundrum for the developers. Anything that isn’t as shown in the guides is seen as “trash” and goes unused by the vast majority of the playerbase. Which gives them a reasonable standpoint of “Why should we bother”? Why bother giving people 100 options, when a handful of people will work out 2-3 optimal builds by running simulations, and then publish that online?
Is it because it’s easier for the developer? I personally dislike the “devs are lazy” argument. More of a case that they have limited time. Why waste man hours on a complex skill/talent/gear system that goes unexplored by the majority of your players? Might as well give them 2-3 options as opposed to 100 options if they are going to use the one that the guide tells them to use anyway. At that point your developers are better off working on something else than creating a complex character building system, and I think that is the point. It’s a trade-off, they would rather their devs be doing something else with their time than creating a great character system that will go unappreciated by the majority of players.
Mass appeal is a closely linked but also important part of their considerations too. Their aim is to make money, they don’t necessarily want a system too daunting that it turns people away.
Call it “dumbing down” if you will, but most people aren’t interested in crunching numbers, tinkering with character builds etc etc. As much as I dislike it, that is the case, most people just want to log on and play some video games without feeling like they need a degree in MMOs to understand what to do.
I prefer the overwatch route when playing PvP. I then know i wont be rekt because of subpar talents, unbalanced classes or lack of time to grind gear. If I loose it’s because I have less skill, lack of team communication or don’t know when to change to a different hero.
Choosing a different hero replaces the choice of talent in this case.
When i play Heroes of the storm I can’t swap heroes midmatch but customize my char with talents. I choose a character based on the map, the other teams comp and my teams comp. I would call these games casualfriendly with a high skillcap.
In PvE I’m all for customisable characters. I like the way Path of Exiles works in that regard. I have not tried it’s pvp so I cant comment on that but i can’t imagen it’s very balanced.
Thanks for the interresting topic!
Oh and in Heroes the multitude of characters to choose from are my cookies, not the talents.
I’m OK with gear customization, so long as the gear is balanced.
E.g. you can choose to improve your bow’s accuracy, or your damage per shot. But you can’t improve those in a way where you have more “points” to spend on improving your bow than other players.
I’m super against gear tiers that provide a faceroll advantage.
It does to an extent, but you can only choose from their cookie cutter heroes.
Some games, like LoL, do regularly release new heroes, so there is variety. You just can’t create your own unique snowflake, which is something I greatly enjoy.
I prefer MMO’s that encourage theory crafting spec’s for talents and stats for gear. When RIFT came out it was awesome. I had a Sentinel, Shaman, Warden spec that was perfect for my play style. I recently started playing WoW cause my guildies from other MMO’s are still playing it and I find it disturbing that dev’s are basically spec’ing my toons gear stats for PvP. For BG’s all you need is iLVL and they determine your stats. World PvP is a joke unless you’re a tank.
I agree to an extent.
Stats IMO should have a fixed pool that can be tweaked, e.g. pick whether you want Strength or Dexterity, with downsides of having low scores as well as upsides for having high scores. E.g. if you go high Str and low Dex, your ability to dodge and dual wield is weaker.
Wow, no pun intended. I wonder how the WoW veterans feel about that extent of simplification.
This has been kind of true for many MMORPGs unless the game was designed with World PVP in mind. E.g. WAR was designed for World PVP / mass RVR, so you had specs that could flourish depending on the context: keep fighting, open map, skirmishing, etc.
At any rate, World PVP with large-scale fighting has typically favored tanks for melee and ranged for DPS – squishy melee DPS can have a hard time due to the high amounts of damage and AOE.
Great post Taugrim! For me the question of what do I prefer I guess depends in a sense on the type of gameplay. sPvP, Open PvP, PvE, sPvE
There are a couple of things that come to mind in your question and blog title:
Balance is a major principle especially if a game company is looking for competitive and marketing leagues and prize winnings.
They want classes that can be AI inspected
They want to control for cheats
They want low cost of entry
They want players that can understand the world-arena in which they come against: what can x do to a y attack in a z scenario
2.Where in one sense MOBAs seem to have low customization, the question is do they? While Overwatch as leveraged for your example has no trees, it has a Holy Quadranity and over 30 variations of fixed classes. That is in one sense customization.
Newcomers pick bad builds. In team games, this is important especially in random queues where the option to invest in another player is limited and it has a downside effect on you.
Many games with customization have “meta builds.” If you are not running the meta- you are shamed, you are called out, and likely…very likely you are actually hurting your team.
Open PvP or a world scenario may be more appropriate for customer builds.
PvE is the best place for custom. Content and scenarios (by devs) are a great place for diverse runs…but even these usually amount to an instance/encounter that is more ideal for build A versus B and a death wish for Build C, but then encounter 2 flips it….this is awesome.
PvE and PvP cannot…probably in the truest sense…co-exist anymore
Finally, yes I will say it….PvPers actually do not want customization and unpredictable encounters.
They once hailed argument that PvE was for noobs and boobs and easy mode players bc it was SCRIPTED play (was always moronic), but has become actually what competitive PvPers want, expect, and get raged over.
Player A does not have Meta Build – = flamefest
Player B does not go to point X in scenario Y = Noob
Player C expects that all other teammates know to go mid, then far after a face-roll
Player D pisses on the floor of the double class queued players who will not switch to
make better group comp
Player E blames any other player if the match result difference is more than marginal
Player F tells Player on team with Class X to switch to get better match make-up against opponent.
I know this is such a shock and I should not say it…but PvP has become scripted. It has become the “get on the A train” or get off. If you do not play meta…if you do not know where to be…if you do not know the correct counter skill…the correct series of steps in a particular map….GET OUT. I hear it all day long in multiple games.
Seriously. PvP is in a great financial and player base position (like MMO for PvE was once upon a time)…it got there by players who wanted predictable outcomes…players who knew the map…knew the classes…knew and followed each step….THE DAYS OF NON-SCRIPTED, PLAYER UNPREDICTABLE< ROLL WITH IT MENTALITY = nearly gone.
What does customization look like? It looks like players (pvp ones) that can craft their own skills, track those skills merits, use armour they want and weapons they want to cast/use skills.
Wise
@wisegameinsight
It’s good that you pointed both of these out, especially the latter.
The latter has been very true for PVE in GW2 – the meta builds tend to be glass cannon as the boss mechanics are predictable and with dodge rolling and such, players can stay alive and heavy DPS makes runs faster and more efficient.
I wonder if some sort of hybrid model would be workable. Meaning there are some cookie cutter builds that the player can choose from – and this would help newbies and less-knowledgeable players pick bad builds – and at a certain experience point threshold, the player can unlock the ability to customize their class from the cookie-cutter models.
The other thing, as I wrote about in my post on Horizontal Scaling several years ago (https://taugrim.com/2012/04/19/why-games-should-scale-horizontally-instead-of-vertically/), would be to offer choices that are incomparables, so that there are no “bad” choices or “overpowered” choices per se, just different choices depending on how you want your character to function. E.g. a bow is not inherently better than a sword, and vice versa – they offer different capabilities. Or to talk about a more specific example, a Cinderburn staff to maximize DoT is not necessarily better than an Earthbind staff to maximize CC – they just offer different capabilities.
Very true. This has been an issue since WoT popularized MMORGPs.
The playerbases want different things, and it’s hell for the developer to appease both communities simultaneously by offering a well-designed, balanced game for both modes.
Even if a game were theoretically balanced, it may not matter – players would still cry that “PVE changes are screwing up PVP” or “PVP changes are screwing up PVE”.
I think PVP players don’t want RNG – they want skill-based outcomes. RNG in PVP is a big issue, as we see in World of Tanks where there is high RNG on where your shell goes, +/- 25% on the penetration roll, and +/- 25% on the damage roll assuming the shell penetrated.
I think most PVPers would be OK with customization, assuming the choices are all reasonably valid – this takes us back to incomparables.
In GW2, the meta for PvP is even more important than PvE. The crazy thing is that the sPvP actually enables build suggestion by class for new players. They are the sort of builds that will quickly make the newcomer quit…they are not meta…even as a skilled player (Gold T3) after 3 years, I can rarely survive using them. For those unaware, GW2 requires no gear no stat grinding for PvP – it’s all flat..All that matters is your build and your skill….player skill has brought that down to 1 (maybe 2) builds by class that work.
We can blame the devs, or we can start to accept that sPvP is destined for scripted play.
I am unclear about things like NBA championship, Superbowl, or World Series, but if there is a workable analysis there…maybe there is customization possible based on winning teams or routine winning teams YoY.
I LOVE customization, I am just jaded now by PvP, which has become increasingly more scripted and meta than PvE raids/dungeons.
I love this thread. I also agree that most PvPers want customization…I just am unsure if there is actually a reasonable space for it with competitive play, sPvP, and money/rating consequences…I think open or server vs. server type PvP is different or more viable for custom.
Final thought: AGS is obsessed with Twitch. Perhaps that invisible “extra” player from the outside throwing a random thing into the map/environment…will change the dynamic enough to be on the ready for “things unplanned” and “unscripted.” THAT intrigues me….but it does so in PvE, too.
even gambits would be a welcome introduction to “customization” with flat skills and meta builds.
I think that for customizable hero pvp, developers in general have given up. They seem to fall victim to the 2 major mistakes, and I will beat a very dead horse twice more. They don’t balance their game mechanics and they don’t balance their classes.
And I think this is caused by the development process for three reasons. First, PVP isn’t top priority. Second, group think, their decision making lacks a singular vision. And thirdly, unwillingness to finding elegant solutions to problems.
Not really going to comment further on the first two, but regarding the third point, the Resolve system in SWTOR was one of the most elegant solutions, I have ever seen in MMO PVP systems. To this day, WoW and a lot of other MMO games have never copied it or created their own version. And I don’t get why, when chain CCing has remained a problem in their games as well as class composition, exclusionary problems entirely the fault of overlapping diminishing return, spell anti-synergies. “Too many stun effects, we need a whirlwind” Paladin kicked from group
To me, it shows a complete lack of willingness to tackle the mechanics of their own game. Unbelievable complacency. In the end it’s fine, I’m just waiting for the Path of Exile of MMO PVP to come along.
On a side note, I was thinking about AOE buffs – they’re usually problematic because they scale too well, but they suck in that buff builds tend to be weak when soloing.
Here’s a simple solution: have the buffs scale down based on # of affected targets. That way buff builds still work solo too, and they don’t become too powerful with lots of affected targets.
Yeah AOE is a problem in general for mass pvp. Myself, I’m a big fan of elegant solutions to pvp problems, like SWTOR’s resolve system solution to the problem of chain CC. And to this day, no one else uses it or a similar system because … well I think they are prats.
With AOE buffs, there are some solutions currently out there. First, diminishing returns the further away characters are from the party member with the aura. Limiting it to your party and not huge groups. The MOBA version, I’ve seen, AOE buffs tend to be really weak, but ties it in to an actual ability. There is also the Lucio model, I’m sure you are familiar with, of giving the AOE a temporary boost to its effect.
The other thing I considered thinking of this, was the AOE realm vs realm battles, that make single target dueling impossible. One solution I’ve seen is they take an AOE damaging ability, give it a hard cap damage potential, like this ability does 10,000 damage spread across all targets in the radius. Unfortunetly I think it makes them easy to heal through.
Myself, I think a better solution would be to make it so AOE abilities override each other if cast on the same area. You can’t have a blizzard and a flamestrike dealing damage to a group of players at the same time. The second causes the first to disappear. Of course part of the reason I like this idea, is it would help with the lag that comes from AOE spamming.
IDK, there are a lot of solutions to problems I’ve come up with over the last few years to these things, but no one is willing to pay me for it so. #ArmChairDeveloper