Things can still go terribly wrong for
developers that choose to expose
their
games to the masses at a premature stage in the game’s
development. Although Scott Hartsman argued that big betas are still
necessary for the AAA titles with the multi-million dollar budgets,
developers need to be wary of allowing the public to see their vision
in a state that isn’t worthy of their time. Like a leak of a
blockbuster movie or a pirated draft of an upcoming sequel to a
bestselling book, the audience may ultimately decide that what they see
in the early version of the product is what the final experience may be
like.
If we turn an eye back to NetDevil and
Auto Assault, at their
post-mortem talk on the game at OGDC 2007,
Hermann Peterscheck and Scott
Brown both stated that beta tests should simply be used for marketing
and stress testing. A development team shouldn’t rely on beta
testing to point out all the flaws in a game’s design,
because that simply won’t happen. Technical problems can be
ascertained through.
In a later interview, Scott
reiterated this thought again saying, “I no longer
believe that beta’s purpose is to find bugs. I think
beta’s purpose is to market your game. I mean it is the only
way you can find things like huge balance issues and what happens when
a bunch of players do something you never expected. It’s
certainly there to find those things, but if you’re running
beta for a game and you have crashes and bad frame rates,
it’s not good.”
So while late game testing remains an important part of the development
process, game companies need to insure that there games are in a
“finished” or “near finished”
state before they ever unleash their products to the public. And the
proof is easily visible in two prominent (and cancelled) NCsoft titles:
Tabula Rasa
and
Auto Assault.
As the numbers slowly revealed a
lackluster launch for
Richard
Garriott’s Tabula Rasa,
the legendary developer announced
that
inviting too many individuals into his game’s beta test
caused a large amount of fallout and lost sales. Here’s what
he told Gamasutra in a late 2007 article:
“I actually think the biggest mistake was made not by the
marketing department, but by the development team. We invited too many
people into the beta when the game was still too broken.”
“We burned out some quantity of our beta-testers when the
game wasn’t yet fun," he said, adding, "As we’ve
begun to sell the game, the people who hadn’t participated in
the beta became our fast early-adopters.”
Scott Brown
echoed this sentiment in our late
2008 article with him:
We have a bunch of meters that
we use to monitor buzz and interest in
the game, and when we went out with our first Auto Assault beta and
realized that we had a ton more work to do, we shut the beta down for
awhile. We did a great big polish patch, and everything played way
better, but when we reopened the game again we never had the same
number of people visiting the web page or downloading the client ever
again.
Ten
Ton Hammer: Really?
Nope. Never again. Never
even came close. So many things about the game
played so much better after that polish patch, but the numbers were
never even close. That fact killed that game.
But it wasn’t closing
the beta that caused the problem, it
was starting the beta too soon. We fooled ourselves into thinking it
was ready to go.
And these thoughts are continued by our own Ten Ton Hammer
premium members. When asked about their worst beta experiences, most of
them didn’t cite “unpolished content” as
their major concerns. Instead, concerns with “being lied
to” and “client not running” were top
among the responders:
“Worst beta experience, for me at least, was with
AoC,” Condar replied. “Seeing that Funcom had
basically lied to us about what was in the game. Even those of us in
the closed betas weren't given the correct information. That and trying
to get the client to run some days was like having a second
job.”
“Vanguard,” Thansal stated. “It didn't
run.”