Some Reflections After Two Weeks Of A Valley Without Wind's Beta

If you've been following the main Arcen News Blog, or the AVWW Forum, or any of our social media, then I'm sure you've seen that we've put out a few metric tons of releases for A Valley Without wind in the last two weeks.  We're presently working on update #20, actually.

So, how have things been going?  Sales-wise, we started out with a real bang, and then it's gotten a bit quieter as the initial wave of press passed and all the players who had been most anxiously awaiting the game already now have it.  This is already our second-best-selling beta yet, though, and it's only been two weeks -- within the first month, this will have surpassed what Light of the Spire sold in three months of beta.


That's great, but of course sales aren't the core reason that we do these public betas in the first place -- what are we learning from all this exposure to players?  Well, quite a lot actually.  The difficulty curve was all out of whack, the lack of loot drops and the unkillable monster spawners really bugged people, and there were a number of mechanics that were found to be too complex even by our hardcore forum fans, and so on.

Difficulty
The difficulty curve is now consistent throughout the game, which is great, but in the process we've had to rebalance everything, and that's led to some wild oscillations in some parts of difficulty -- mostly toward the harder end.  Still working on that!  This 20th release, to be available later today, should address most if not all of the remaining large issues with difficulty being too high.

I was actually really pleased with the difficulty flow of the game right at the start of beta, but after about 10 levels of play that really started breaking down and the game got progressively easier.  Having a growing pool of players in the level 30-40 range or even higher has been really invaluable for us -- testers invaluable to any game, especially at this stage in the life of a game.

Complexity
Ever since A Valley Without Wind was first conceived, it's been our goal to have it be so simple that we didn't even need a tutorial to get new players started with it.  I loved how Minecraft or similar just plops you down and you're left to forage and expand, etc (though I understand a tutorial was added there more recently than I last played).

On the other hand, we wanted our usual incredible amount of depth to this game -- just gating most of that depth and complexity into a natural game-like progression.  Games that are easy to pick up are awesome, but once players have their feet wet they shouldn't be asking "is this all?"

So that's been an interesting balancing challenge for us, and I think overall we've done a pretty good job of getting into the right ballpark.  We were taken by surprise both in alpha and beta by how complex certain parts of the game were perceived to be, but again that's why we get the players involved early -- it's very hard to objectively judge the complexity of something you thought of.  And if you let the game get too far along without getting that critical external feedback, there are some kinds of complexity that just get too interwoven into the game that you can't later take out.

We've made a lot of strides on the complexity front already: adding more tooltips, simplifying the crafting interface, level-gating the strategic and citybuilding aspects of the game so that you don't have to deal with them immediately, and most recently taking out the "profession books" for the main three kinds of crafting.  Those added complexity without fun, which I wrote about in more length in the release notes.

So all of this helps a lot, but there is still more we need to do.  We're going to be adding a little introductory "mission" that players have to go through in new worlds.  It's not really a tutorial -- the linear mission will give you the basic tools you'd need to find for yourself anyhow, and make you use them as you would anyhow, while giving you onscreen tips just in case you feel stuck.  This should take all of 5 minutes to play through, or even less for players who already know how to play, and that should serve as a much more fun and game-like introduction to the game rather than just reading lots of stuff from the adviser guardian.

"Signal To Noise"
One of the interesting things about making a really huge infinite world is that one of two things happen: on one end of the spectrum you're constantly tripping over awesome stuff all the time, and the world feels small because you don't really have to go anywhere to accomplish anything; on the other end of the spectrum the world is incredibly vast and sparsely populated, and you wind up having to really trawl immense tracts of land to get anything done.

Right now AVWW is somewhere in the middle, which is the overall best place to be, but "the middle" is a broad place.  AVWW definitely leans toward the side of being too large and sparsely populated.  There's a ton of awesome stuff in here, and players who bother to learn how to use the dungeon map and world map, and who read the tooltips, can find those things with ease.

Of course some of the traversal is still a bit too longwinded to be as fun as we'd like (maze rooms have come up a lot in particular as needing tuning).  And even more core is the issue that players tend not to want to read.  I'm as guilty of it as anyone -- just let me play.  If the game looks familiar enough that I think I should know what I'm doing, I figure I'll figure all the stuff out as I go.

That's a challenging problem to combat, because one of the things that I like to do in my games -- as you see in AI War, for instance -- is to give the player a lot of choices and let them figure out what matters to them.  In AI War, if you try to take every planet you will usually wind up losing because of the AI's rising aggression.  In AVWW there is no penalty for trying to explore every room in every building, but it sure is boring to do; we intended for people to scout the buildings, find the rooms of interest, grab the loot, and get out.

And a number of players do just this; but still others don't.  So that's still something we're trying to find the sweet spot on, and again the beta testers are being absolutely invaluable for this.  I think we're not that far off at this point, but there's still some more condensing and such that needs to be done.

New Content
With all the flood of feedback about the existing content, and all the revisions to balance and so forth, there's been limited time to work on new content thus far.  But we have managed to add in a couple of new enemies, and our update #20 includes the first batch of new spells.  We've also added a lot of macro-game type additions all throughout -- the citybuilding bits have really grown, and we got things like consciousness nodes, pylons, vortexes, and resource deposits added on the world map.  And there's been a large number of new boss room templates, too -- things like that.

Thankfully a lot of the revision-type work looks like it is overall winding down, so I expect to see the focus really shifting soon into blasting out tons of more content.  I know that's something that a number of players have been really waiting for most eagerly, and so have we, but I think the last two weeks have been incredibly valuable to the game.

Multiplayer
Multiplayer is something we've decided to make a focus sooner than later.  Still no ETA on that, but Keith is going to be sinking most of his time into that starting today.  It's going to be rough and in need of a lot of beta testing right at the start once we first publicly release it, but hopefully within a couple of weeks of the first release into the hands of the core beta testers, we'll have it ready for wider dissemination.

Why Release This Early?
A comment that comes up occasionally from people who don't really "get" the game yet, or who just aren't that interested in it at all (hey, that happens) is "why release the game in such an early state?"  Usually with added snark or nastiness.

That really gets under my skin a bit, I have to say.  We've got loads of players who are clearly having a blast with the beta, and the feedback we're getting from the dozens who are choosing to give feedback has been nothing short of amazing.  This is how we make games at Arcen -- we actually listen to players and value what they have to say.

But you can't listen to players if they can't play the game.  And if the game is already finished, then there's not a lot that player feedback that the players can really give.  Ergo: if you want player feedback, you have to get them involved once the core of the game is solid, but before the full shape of the game is even remotely finished.

I think that when a lot of people ask for feedback, they're really saying "tell me you like it and think it's perfect in every way."  I actually want to know when there's something up with the game that makes it less fun, or more tedious, or too complex, or whatever.  When we know about those things, we can do something about it.

To some extent, that means that the players who get in early get to see "how the hotdog is made" a bit.  And if somebody really doesn't want to know that sort of information and gets turned off the game by it... well, that's a cost of being open, I guess.  But I think we and the game gain a lot more in the long-run by getting players involved in a real, early, and meaningful way.  Next time you complain about a released game having some fundamental flaws that you just can't understand, remember -- that's what can happen when games are developed in a vacuum.

The Bottom Line?
Things are going really well, I think.  We have a growing contingent of players who seem really happy with the game, and the early press has been largely quite positive.  There are still a few fundamental things we're trying to get tuned exactly right, and the new player experience still needs a lot of work in particular, but in another week it's going to be in a whole different place than it is even now.

In short: we're one of the few game developers, indie or otherwise, to experiment with a development process remotely this open.  And I think that the process is working quite well.  Only time will really tell, of course, but based on the first two weeks of beta I'm strongly encouraged.