Gamasutra: The Art & Business of Making Gamesspacer
arrowPress Releases
July 26, 2014
PR Newswire
View All
View All     Submit Event

If you enjoy reading this site, you might also want to check out these UBM Tech sites:

DC Universe Online DLC 10 Post-Mortem, Part 1
by Mark Halash on 07/10/14 12:39:00 pm   Expert Blogs   Featured Blogs

The following blog post, unless otherwise noted, was written by a member of Gamasutra’s community.
The thoughts and opinions expressed are those of the writer and not Gamasutra or its parent company.


DC Universe Online has been a successful live MMO for more than three-and-a-half years on both PC and two PlayStation consoles. As with every game of its type, new playable content, new features, bug fixes, improvements, and balance of existing systems must happen regularly to provide a continually engaging experience for veteran players and to attract new customers.

With our 10th DLC Pack, Amazon Fury Part I, there were two challenges – our ongoing need to create fun, rewarding content that players will want to repeat, and a new feature – Weapon Mastery – that would expand and deepen our well-established action combat system.  These two projects had to run parallel to each other and launch simultaneously in order to meet our deadlines. With one major update split into two paths, we will tackle each individually to see where we succeeded and where we could have improved.

PART 1 - Weapon Mastery 

Weapon Mastery was an expansion of our action combat systems, and creating it was an opportunity to do some important things in DCUO. For one, it allowed us to improve the visual fidelity of combat. With its focus on hitting specific timing windows to make weapon attacks and powers most effective, Weapon Mastery got us a little closer to the original vision I had for DCUO's combat system during initial development.

How does Weapon Mastery work? Weapon Mastery allows players to spend skill points to achieve "Mastery" in the game’s many weapons. Certain weapons are linked with other weapons. Each weapon has its own set of weapon combo attacks, and when a player achieves multiple masteries in the original and linked weapons, they can perform cross-combos to transition fluidly from one weapon into combos of the linked weapons.

The feature itself has beaten expectations and been received well by the community, but there were some bumps along the way, including the need to address long-standing weapon and power balance discrepancies across the game. Time is always short, and there are always unforeseen technical hiccups. If that wasn't enough, Weapon Mastery launched a new, more official method of interacting with the community called Work In Progress, where we invite players into the development process and discussion while development is beginning or underway.

I'm going to explain some of the hurdles we faced during the development of Weapon Mastery, how those got resolved, and what we learned. Then I'll focus in on some of the improvements Weapon Mastery introduced to the game.

Hurdle #1: Major Re-Writes Ahead…Good Thing the Systems Team is Awesome!

I have an awesome team. And I'm not just paying lip service to that. It is actually true. These guys care a great deal about the game and about making it really great. They are experts at their jobs and masters of the arcane idiosyncrasies part and parcel to the systems of any live game. Just about anything that isn't directly content related filters through my team. We shoulder a large and varied workload with gusto.

Weapon Mastery required a fairly significant re-write of the way player weapons are displayed on the client. Until now, you only needed to worry about displaying one weapon per player on screen. With Weapon Mastery, the player may use as many as three. Keep in mind, everything has to fit within the PS3's memory budget.

Working in tandem with engineering to accomplish both the design and technical tasks needed to launch Weapon Mastery meant additional challenge. Yes, we didn't need to actually see the weapon to stub in the combos, balance the damage, and all of that. But the hooks for visual FX, sound, hit reactions, timing and pretty much all of what goes into "feel" requires the weapon to be there and work. So despite hundreds of new attacks, we couldn't look at these things until late in the process.

As should be expected with any major re-write of a system, there were bugs. Rather than concentrating on nailing the "feel" items mentioned above, Jesse Scott, the designer now responsible for a lot of DCUO’s combat work, spent most of that last sprint helping the engineers chase down bugs. This left a tight, compressed window of time at the end of development for us to polish the system.

Hurdle #2: OMG They Killed Jump Canceling, Clipping, and Kenny

As I mentioned above, no. 1 wasn't happening in a vacuum. For the last few years and as we look to the future, SOE as a whole is expanding how we listen to and utilize our players and their feedback. We believe being as open as possible about development leads to stronger communities and stronger games. Weapon Mastery was the first topic of a new initiative called Work In Progress, where we will invite players into the development process to gather feedback, ideas, and concerns on what we’re working on, while we are working on it.

For Weapon Mastery, we were letting players in closer to the end of development. Work was nearly complete, but not finished, not on the public test server, and certainly not live. We launched with a livestream, where we demonstrated the system on internal environments, answered questions, and showed the new combat in action. This was followed by explanatory, official Work In Progress forum posts to begin the discussion.

Beyond overall impressions and concerns, we were specifically interested in how the community would respond to how Damage-centric Weapon Mastery was at the time. (Players in DCUO can choose to play in different roles – Tank, Healer, Controller, or Damage.) There was ongoing internal debate about whether or not Weapon Mastery should have role-specific benefits beyond simply more and better damage potential. The community responded with an overwhelming YES, and we were able to implement role-specific bonuses prior to launch with their input. We believe this is a large part of why the system was so well received in the end, as the final version would apply to all players.

To get to that point, though, we knew as we were creating Weapon Mastery that we needed to address long-standing and complicated balance problems, and saw an opportunity to correct these and level the playing field in one fell swoop.

Our combat is highly visceral and fluid. To achieve this, many channeled abilities prevent movement but can be canceled by jumping. This is "jump canceling". It allows you to end the ability so you can move away from danger and continue to react fluidly to the moment. You can attack or use another ability instantly after you cancel. So, if you cancel instantly you can use that ability as a bridge to perform another weapon attack and then of course clip that to do another super power, one on top of another on top of another.

Still getting the full effect of the canceled ability, plus being able to do more attacks and abilities right away led to unintentionally high damage output. This was unintentional, and there was significant variance between how different powers and weapon combinations could perform. Because we truly wanted balance, we would either need to bring everything to this level or remove the exploit. In the end, we decided bringing all powers up to that level would be problematic for several reasons. Everyone doing that much damage would trivialize the game, to start, and there was also a perceived difference of how much skill was required to achieve maximum potential with different powers. To players, it wouldn’t seem fair to have some powers be more difficult than others without a resulting pay off in potential. In the end, we went with the fix, which was simply enforcing a minimum time before you could attack again after jump canceling, while still allowing instant movement.

Further complicating the situation was that DCUO was a live game in its fourth year, successful largely due to its visceral, action combat, and jump canceling had never been publicly referred to as an exploit in those years. The forums exploded after this news. We were changing what players saw as core, fundamental mechanics, or why they play the game, and it was a challenge to explain the reasoning and our vision for the future.

Between the initial announcement and launch, we made a number of related systems changes, new Work In Progress threads on more specific elements of Weapon Mastery and combat overall, and outlined our plans to fully balance the game’s powers, both in terms of actual damage potential and equal presence of advanced, skill-based mechanics over the coming months. We even met with hand-selected, constructive members of the community in person on the test server to discuss specific concerns and potential right-now changes to address the removal of jump canceling on the most impacted powers.

We were able to get both the Weapon Mastery and related power changes implemented in time for launch. Because Weapon Mastery makes the game so much better for everyone, we decided we didn’t want it blocked behind a DLC Pack pay wall, and launched it as part of a Game Update. While the overall changes are to this day controversial, Weapon Mastery itself was well received, and did bring new, fun mechanics to our ever expanding action combat system.

Win #1: Improving DCUO's Core Combat

It's difficult to understand the impact of Weapon Mastery without at least a top level understanding of DCUO's combat system. Some of the most controversial aspects of DCUO's combat system have the idea of animation canceling or "clipping" at their root. I'm going to take a minute to explain what this means so you have an understanding about why it's both important and controversial.

In order to keep things fast and fluid and to support our reactive combat model we have a hierarchical priority system in place allowing abilities of certain categories to interrupt the animations of other abilities. At its simplest, the rule says that super powers can interrupt weapon attacks.

For example, if you're in the middle of a weapon attack and your tank suddenly gets rocked by an unexpectedly huge hit, you can fire off a shielding super power instantly in the middle of your weapon attack. Your weapon attack still goes off, as does the shielding power. But your weapon attack animation stops and is replaced by the animation for the shielding ability. Players have dubbed this "clipping".

My original design ideas for DCUO's combat system had much heavier restrictions on "clipping". Given the speed of the game, you must be able to quickly react to save yourself or others with beneficial super powers. But for offensive powers, I wanted a timing mechanic that only allowed activation at the moment a weapon attack hit.

The one-two punch is a feature I've identified as being present in pretty much every combat system I think feels great and as missing from those that feel "meh". The one-two punch tickles the part of your brain that cackles gleefully when you pull off a fierce punch/shoryuken in Street Fighter or do a Fireball/Fireblast in World of Warcraft. It results in a visceral feeling that makes you want to shout!

Unfortunately, we lacked the engineering bandwidth to implement a system that supported these elaborate timing rules. So faced with the decision of whether or not to allow harmful super powers to interrupt the animations of weapon combos, and I erred on the side of fluidity. There was no comparison about which felt better to play.

Clipping introduced some side effects. One side effect is a loss of visual fidelity. Obviously, when you clip you're skipping some portion of an animation for that combo, blending into another attack entirely. Weapon Mastery rewards players for not clipping their weapon attacks with super powers. If they wait until the moment their weapon attack hits and then launch a super power, it receives a significant power increase. So it's an opt-in system, but with a beefy reward you don't want to leave on the table. Because you're going out of your way to fully animate, the game looks a whole lot better. This type of gameplay is much closer to what I originally envisioned for DCUO's combat. It's a very satisfying way to play.

Win #2 Addressing Old Balance Issues

DCUO is a pretty easy game to pick up and play, which is part of what makes it a success. But the difference between players of mediocre skill and those of high skill is pretty dramatic. Not only do we have the standard MMO "knowledge" based mechanics like how to spec, what to mod, which buffs to use, when to use your trinkets, and what's the best rotation, but we also have a full-on skill-based action game on top of that. You need actual manual dexterity to consistently pull off combos, score counter-attacks, and avoid being countered yourself, and all while trying your hardest not to get flattened by super villains, demi-gods, hyper-intelligent alien beings, and the random thrown bus.

Another side effect of "clipping" is an increase in potential damage output. If you are skilled enough, you can effectively reduce the animation time of the last hit of your weapon combo to near zero by ending it with a super power. Offensive power clipping represents about a 30 percent average increase in damage potential over not clipping. Basically you can do about 4 seconds worth of damage in 3 seconds depending on how skillful you are at it. It's sort of like haste. It gives players another skill-based axis to master and improve their performance, so I honestly didn't mind much.

Just prior to releasing DCUO, we had actually begun working on our first DLC which focused on Green and Yellow Lantern content. I was working on a Green/Yellow Light power set that would allow players to get a ring and play as Deputized Lanterns joining the ranks of such heroes like Hal Jordan or villains like Sinestro.

The mechanic for this power was something we called "Construct Combos". Basically players were able to use a power from their tray and then launch into combo versions of other light powers they had access too. So this was the first introduction of "combo powers" into the game.

Now if you've been through college you surely remember nights where you said, "This is totally awesome. I hope I don't regret this tomorrow!" Light powers were too fun to play. The genie was out of the bottle and there was no stopping it from being the life of the party. I didn't anticipate the impact that combo powers would have when combined with clipping until it was too late.

Combo powers start with a super power from the tray and then transition into a combo. That combo is considered a weapon attack. Remember how weapon attacks can be clipped by super powers? This means combo powers can be clipped with combo powers. The lead up power from the tray is perhaps 1.3 seconds in length. So, instead of causing four seconds of damage in three seconds and getting a smallish bonus, you are instead doing 2.3 seconds of damage in 1.3 seconds, which is a much bigger bonus. That left us in the position of combo powers being simply better than the rest.

Now over time we have been and will be bringing the other powers up to the capabilities of combo powers, by introducing their own advanced mechanics. But, these systems take time to create, and a future fix doesn't help those players who are waiting right now. One of major things Weapon Mastery did was give everyone at least one way to play that results in maximum damage effectiveness. The difference was pretty dramatic. We went from only seeing the "upgraded' powers producing top damage numbers, to seeing no clear winners day to day and week to week in our DPS logs.


I think we've learned some things from this experience. Implementing the Weapon Mastery feature was at times trying but ultimately quite rewarding. It was a noticeable improvement to the game as a whole. It improved the visual fidelity of the game and helped improve combat. It gave players an exciting and effective new way to play, and it bridged the gap between the previously "alpha" powers and those lagging behind.

Changing core concepts like combat in a live game is never easy and must be done with great care, and at SOE we know we cannot do that in a vacuum, without our players’ input. At the end of the day, we were able to open this necessary development process to the community and then truly listen and take action, and the result was well worth it in a better game.

Check out part two of the post-mortem to learn more about what we discovered with Amazon Fury.

Mark Halash is an industry veteran with fifteen years of experience credited on over a dozen titles and expansions for genre defining online games such as EverQuest, Star Wars Galaxies and DC Universe Online. Mark served as Combat Design Lead during initial development of DC Universe Online and was promoted to Lead Designer, Systems after the game's release.

Related Jobs

24 Seven Inc
24 Seven Inc — Los Angeles, California, United States

Game Programmer
Galxyz Inc.
Galxyz Inc. — Mountain View, California, United States

Narrative Writer for Interactive Media
American Girl
American Girl — Middleton, Wisconsin, United States

Game Developer
InnoGames GmbH
InnoGames GmbH — Hamburg, Germany

Quest Writer (m/f) for The West


Trenton Kennedy
profile image
Good stuff, Mark! Sounds like you guys are kicking butt as usual.