Log in

No account? Create an account
.::.:::...... .::::: ..:..

August 2009
2 3 4 5 6 7 8
9 10 11 12 13 14 15
16 17 18 19 20 21 22
23 24 25 26 27 28 29
30 31

harlzen [userpic]
Champions online... How to make a mediocre MMO

The champions NDA has just come down a week before the game is released. I quite liked
City of Heroes, an older superhero MMO, so I applied and got into beta and have been
playing it for a while. Sadly

1. Have a weak, derivative or inappropriate foundation for the game.

In the case of champions online the original design seemed to be most strongly
formed by the console environment and existing material (Marvel ultimate alliance).
These are fine as considerations or influences but the design process also has
to address what the game will be adding that is new or has interesting gameplay
possibilities. Like any large production getting this answer clearly defined in
the early stages will save a lot of cost compared to trying to retro-fit a solution

The main gameplay influence was on a more action oriented environment, no doubt
encouraged by the console gameplay. Goals included no downtime, high character
mobility and more engaging combat in highly varied environments. In practice though
some of these things are very hard to do in an MMO environment. And some of them
have immediate costs. For example with no downtime how do you encourage a player
to manage energy rather than just constantly use their biggest attack? How do you
have engaging combat in an environment that has to deal with lag and the resulting
uncertainty in character positions. Meanwhile having mobile characters immediately
limits how interesting power animations can be. Solving these problems in an
interesting way would provide a firm foundation for a game. Or you could...

2. Build the world first, worry about fun later.

It was fairly clear that they focused first on getting the engine and world working.
When beta started the game world was in a reasonably advanced stage but the mechanics
were still quite basic. It felt like they'd been farmed out to different staff members
which really limited the cohesiveness of how the powers interacted to form good
gameplay. In addition I can assume each developer had a fair amount of grunt work
such as fleshing out powers or designing itemization. Faults in the game world, such
as the chronically bad UI or massively undocumented powers are more obviously broken,
and attract more fault reports, than global things like design weaknesses. This tends
to distract developer attention without a core designer tying it all together.

However the two lead designers for this project were noticeable only by their absence.
Major game mechanics remained unexplained and unclear even very late in the beta process.
Information on design was more likely to come from Bill Roper doing publicity seeking
interviews than any sort of interactions with the beta community. The "State of the
game" posts which were probably meant to fill this function were often missing, outdated
or more like annotated change notes than anything which would offer insight. It may
have been that there was deep design work going on but the feeling was very much of
a rudderless ship left to drift.

As an example the beta community was fairly unified (barring the inevitable fangirls
trying to become the developers best friends) in strongly disliking the limit of having
less than 7 powers hot keys. The precise number varied a little as some of the 7 slots were
at times used for passives or required utilities but the end result was a very small number
of powers. This led to heavily repetitive combat and the bizarre situation where you'd
get powers due to levelling but be unable to actually gain anything from them due to not
having any slots to put them in. This was almost certainly a result of the games console
goals despite some impressive smoke-screening from the developers where they suggested
it was due to university studies on the limits of memory. The valid argument that the
study related to memorizing abstract information (phone numbers I believe) and that more
importantly their customers are used to having larger numbers of active hot keys were
ignored for months. Some people were even banned for being too agressive in demanding
an answer. Eventually the developers promised to explain why a small number of hot keys
made for better gaming. This never occurred and shortly before release they doubled the
number of hotkeys in addition to moving some powers into passive slots. However all the
powers were designed on the basis of having a very small number of active powers so this
last minute change also had negative gameplay interactions.

There were quite a lot of bizarre gameplay decisions that really gave the feeling they
were making it up as they went. For example in order to meet their "no downtime" goal
they introduced an endurance pool that started largely empty and was filled by using
a trivial damage attack (different by power set) to fill it before you could use your
larger attacks. This added nothing to gameplay because you had no tactical options, if
you had power you would never use your power building attack since it was very weak. If
you did not have power you could do nothing but use the power building attack. The end
result is it did not add any tactical depth because there was no choice and it removed
the ability to do things like alpha strikes or resource management and destroyed the
flow of combat. This system was modified in the last weeks of beta so that the power
bar started full (enabling alpha strikes) but would empty faster.

The powers themselves were another issue. Because of the small number of hot keys each
power set (eg. fire, single blade) had to cover all the basics but gained little from
duplication. Thus each power set would have an end builder, a ranged single target
attack, an AoE and then maybe some minor variations such as a charge up single
target attack or a cone AoE. Almost all of the powers were about doing damage probably
to support the "action-RPG" design goal. The end result however is that a lot of the
power sets felt very similar in play. Zapping things with lightning, fire, force, bullets
and such was ultimately all about a difference in special effects and minor game mechanics.
And given their goal of total customisation people were free to cherry pick the best powers
for each need reducing gameplay variety even more. The first problem was never fixed, the
power list is actually far more limited than the number of powers might lead you to believe.
The second was solved by making powers interact such that you were heavily encouraged to
focus on one power set. For example a gun use power that made all gun use powers half energy
cost but all other powers double cost.

The small number of active powers, similarity in the powers (and strongly DPS focused), holds
and heals being nerfed for PvP balance and mob hitpoints being steadily boosted to slow
progression and increase challenge led to some very repetitive combat. You'll spend a lot
of time alternating between end builder and either AoE or single target damage to grind
the opponent down. Mob AI and powers are likewise fairly basic. The much vaunted "run & gun"
is largely useless because ducking out of sight simply stops your power regeneration due
to the need to have constant line of sight for the attack. Strangely using cover worked
better in City of Heroes where you could make the tactical decision to gain some endurance
and let powers refresh by running and hiding.

3. Consider the beta a promotional tool

The Champions online beta was frequently labelled as a product "preview" and this felt
about right. Despite the importance of balance in making the game enjoyable the beta
was run in a very casual fashion. The game was up for very limited periods of time, even
in the last weeks of beta only running for 2 sessions a week. Testing was rarely focused
to any useful extent. Testing tools like being able to re-pick powers or level up in order
to test power builds were absent for the majority of beta and then quickly removed or
weakened when introduced. Things like introducing end-game content and then doing a
character wipe 2 weeks later ensured that testing was much less useful than it could have

In addition the game mechanics were so clearly in flux, and developer communication so
poor, that it was very hard to have a baseline to give bugs against. For example the
might powerset was felt to be very weak, the passive defences too strong and a huge
variety of other obvious imbalances. But without some idea of what the balance goal
is meaningful feedback is impossible. It was further discouraged by the developers
putting "powers are imbalanced" as a known issue that was in place up throughout beta.
In many cases the beta testers could barely determine what the power was supposed to
do since the only documentation was algorithmically generated from the power mechanics
and generally incomprehensible. Heavy balance changes were put in at the last moment
with no opportunity to get feedback or iterate on it. This process of power balancing
will almost certainly continue into live.

4. Launch content light and expect to generate it live

This is probably the biggest problem with the game. The game has effectively 5 zones
consisting of one city zone and four outdoor areas (desert, snow, forest, underwater).
Levelling is done in a wowlike fashion with each zone having a sprinking of points of
interest which often badly conflict with one another. Having an alien invasion, snow
demons, a canadian uprising and an air disaster all within a kilometer of one another
makes the environments feel more like a theme park than real places. Each location will
have a number of quests of the traditional kill this, collect those and escort him type.
These quests are the only meaningful way to progress as mob XP is very low. There are
Warhammer style open quests but these are often imbalanced and the reward for doing them
poor which combined with the bad grouping mechanics and shard design (no servers, multiple
instances of the zone with quite low populations) means they're frequently laying idle.

In general to get to the level cap you will need to do pretty much all the quests in
all the zones. There is some switching between zones, for example the city fills a
small segment of levelling between upper and lower desert quests, but you are going to
be spending a lot of time in these areas. Any characters after the first can expect to
follow pretty much exactly the same path with minimal variation. Nor are the quests
interesting enough that they avoid blurring into each other and people just batch
processing them.

More seriously the game is missing 20% of the content it was meant to launch with. In
theory the max level is actually 50 but the game will launch with the level cap at 40
and release new content soon after release. However this line was used before the
development process was extended by 3-4 months which should have been enough time for
it to be included. It is more reasonable to assume this content is only in its very
early stages and not close to being release ready.

The end-game (well, not really since it's level 40 content) which might bridge this gap
and stop bored people cancelling is at a primitive state. It was only introduced in the
last weeks of beta and consists of 5 daily solo instances. These instances are featureless
maps with a sprinkling of mobs which have clearly been rushed out. Doing 5 of these gives
you one additional mission. The rewards from these missions can be used to buy access to
one of two group instances or gear / costume rewards. It's almost completely untested and
the solo instances are dull, involve a lot of travelling and have minimal challenge.

In addition there's no real reason to bother. In general you'll have all your core powers
long before you reach maximum level. These powers can be enhanced with upto 5 points you
also get from levelling. So your character is already complete. The end game content offers
you gear which due to poor itemisation and stat mechanics is of minimal interest, especially
since gear from questing is sufficient. You can also buy new weapon models and vanity cloaks
which many players will have very limited interest in (how many cloaks do you need? most people
will only want a single weapon model that fits their character image).

Itemisation is weak for a couple of reasons. One of them is the use of algorithmically
generated names for both crafted and found drops. A piece of eye-wear (by icon, gear
slots are actually generic) called "Energized Torpedoes" does not make you care much
about the item. Most items will have a 1-3 stats on them but the influence of stats on
the game are so indirect and obscure (and in some cases known to be minimal) its quite
hard to care. This pretty much destroys crafting as well other than the ability to make

In short people will game the quest system for rapid progression. They can't slow it down
much because there's no content to support a longer levelling curve. This will quickly
lead them into an end-game which has little interest, challenge or reward.

5. Launch with bad design decisions because you were rushed.

This is sort of the catch-all section. But it also reflects that beta feedback is pretty
useless if the developers barely have time to glue the bits together and make the game
minimally saleable. There has to be time for the "is this fun?" test before you release it
and it seems a lot of games companies just don't feel they have the time for that. Anyway,
some smaller results of that.

a. Trying to balance PvP and PvE with the same ruleset. That just doesn't work. Blizzard
are still failing for this reason and most companies can't afford to waste that much energy.
For that matter don't expect PvP to make up for having no end game content unless your game
does something really special in the PvP context.

b. Remove release content to stock an in-game store when your game is already content-lite.
Buying the box should be the price payed for the content developed at that point.

c. Make a character creator which uses 3D objects extensively but has very poor texture usage.
So the furries and robot fetishists like it (especially since everything can be made shiny) but
trying to make a traditional super hero is very limited. And the 3D models are pretty crude.

d. Low detail models. Many of the objects in this game are really basic. For example the jet
used for transport or ships in the harbor are just simple geometric shapes. The wolves in
alaska are almost painfully bad with their fur chipped out of fresh plastic. This combines
with generally bad animation (watching an NPC talk looks like someone in the throes of a facial

e. Highly derivative environment. From monster island being directly ripped off from the island
of Dr. Moreau, underwater environment from various atlantis stories and city gangs ripped off
from clockwork orange. It just feels really lazy and excessive. The champions lore clearly only
gave them a bunch of uninteresting heroes to use as quest hubs... though foxbat comes close to
having an actual presence, even if it is hugely geeky and comic-relief.

f. An entirely underwater zone. It always sounds like a good idea, it inevitably isn't.

g. The much vaunted nemesis system is extremely basic. Design the appearance of the boss, a
text and power set and then have random ambushes by his minions and a couple of missions with
that design applied to the mission boss. End result is not much different from running normal

Current Mood: stressedstressed