Why Shelve a Game?

The Concept

Wolves of the Dull Moon (WotDM) was intended to be a real-time horror/comedy roguelike. You would be dumped in a procedurally generated forest a few minutes before the sun set. No instructions, no direction or indication of what to do. You could wander around, harvest berries, and pick up rocks or sticks. Curiously you can also urinate and defecate. You also have a fear meter.

Then your time was up. Night comes and werewolves start spawning. You can’t fight; you can only flee. Werewolves would track you by scent, sound, and sight. Cross streams to cover your scent. Piss on a rock then throw it away from you to lead them astray. Eat berries that make you less stinky. Eat berries that make your piss more stinky.

Werewolves coming too close to you would cause your fear meter to rise. Full fear meter would cause full blown panic and almost certainly cause nearby werewolves to rip you to shreds.

Try to survive till sun up.

 

Itchy Feelings

I made a prototype in Unity 3D and liked it – mostly. It was fun to sneak around increasingly large packs of werewolves all sniffing around, trying to suss out your location. You could cover yourself in mud and swim through streams. In one build full panic would cause you to piss yourself, allowing werewolves in a wide radius to track you via scent for quite some time.

WotDM AI Detection
WotDM AI Detection

Maybe it was good enough to pursue, but I ran into three big problems I couldn’t resolve to my satisfaction:

  1. Displaying scent and sound propagation clearly and succinctly to the player.
  2. Having close sight range be powerful (werewolf will sprint toward the player and melee) but still maintain close range tension (werewolf walks right by the player who is hiding in a bush.)
  3. Map design and balancing evasion tactics and resources so the player can’t just run around collecting a bunch of stuff and sit in a bush using them throughout the night.

So of course I prototyped some more.

 

Play it again!

People pretty much told me the same thing – make visual marker trails. Of course this was the first solution I tried. It’s really the most obvious:

WotDM Scent Trail
WotDM Scent Trail

But the problem is that the tension in WotDM comes from not knowing what the werewolves are doing. Is there a werewolf just behind that hill sniffing your footprints? Should you have swam just a little bit farther upstream before coming out?

The scent trail visual is okay. What’s not okay is that the player has no information about how werewolves are using that information. You might want them to pick up that trail and other times not. How would the player know and react accordingly?

It’s all an information problem. You want to give the player enough information to go on, so he/she can make smart decisions. You don’t want to give the player complete information because that sucks the tension from the gameplay.

It’s even worse for sound. If the player throws a rock or splashes in the water or trips over a tree stump, what information should the player get regarding werewolves hearing the noise?

You would think that sight would be the easiest. Just hide in a bush or dark shadows and it counts as being effectively hidden. I’ve found that this is a pretty bad solution for WotDM because the goal is to survive for X time, not to accomplish some other Y objective.

 Aside: I find many stealth games to be poorly designed in this aspect. Sitting in the dark for a minute while a guard with an arrow in his back thinks it’s all clear is a terrible system. It’s just a mini-instance of wait for X time within another Y objective. Not that I have a good alternate, game-balanced solution.

This dovetails with werewolf melee and permadeath. Werewolves should basically be able to instakill the player in melee, and that’s how I made it. Health bars have no place in WotDM. And yet this exacerbates the “sit in the dark and wait” issue. No means of direct confrontation means the player must evade, evade, evade.

Werewolves of a necessity must move faster than the player. It is possible to have them move slower and just ramp up the number of werewolves on the map, but this is an inelegant solution that simply drags out a foregone game if the player is forced to constantly run from a poor mistake, a trail of werewolves all sprinting slowly after their quarry.

So they move faster. This in turn means that in melee if the werewolf spots the player it’s game over. This means that situations where a player can sit in full cover is almost a necessity because distance and distraction is the player’s game – melee and tracking is the werewolves’.

Aside: I say almost a necessity because it could be that the player just has to maintain a farther distance from all werewolves in line-of-sight. “Melee” is a rather arbitrary term when describing distance between the intended objects. But melee meaning “very close” is a better descriptor than “sort of close” since players will more intrinsically understand being too close rather than “slightly too close to some arbitrary circle in the sand.”

The two general solutions are unfortunately the most complicated – competing human AI units and game changing milestones.

 

Shadows and Shouts

WotDM wasn’t really intended to be a stealth game. Sure you hide and evade, but the shadows aren’t intended as the primary method of evasion. Actively casting false trails and erasing true ones are supposed to be the primary tools of the trade.

Other human AI units can be a really good, or really bad, monkey wrench. Is there cooperative interaction? You help me and I help you? Maybe there needs to be two humans to perform a task? What about antagonistic AI that flings poo at you? Can they yell to attract other werewolves to your location?

Aside: Can you tell I like the scat mechanics?

WotDM Pants Pissing
WotDM Pants Pissing

It does add a lot more dimensionality to gameplay, enough to make up for information shortcomings regarding werewolves. And humans can’t attack, so maybe the player gets so involved dodging an antagonistic human AI unit that a werewolf gets them both.

Game changing milestones are the other solution. For example the moon could cast shadows and the player might be forced to move, at least a little bit, to stay within an object’s shadow which might in turn cause nearby werewolves to hear him. Werewolves could spawn faster or slower. Werewolves without a scent trail might start congregating on a random position, or others would call them to hunt a human AI unit that was more exposed and actively being tracked.

You get the idea – mix up the gameplay by adjusting werewolf and human AI behaviors. And yet when I started adding all this the gameplay got more complex. It wasn’t just you in the woods being hunted. It was this torrent of information about other units that couldn’t directly hurt you, werewolves with unpredictable and non-deterministic behavior – I felt it lost the essence of tension, even comedy, and turned into a slog.

 

What now?

Well, I shelved it. I’ve repurposed the name for my next game but essentially I’m not touching WotDM again until I come up with a better solution set to the problems I’ve described.

There are, of course, other solutions. And other problems. Multiplayer is a two-for-one sale. But for now I’ll probably just focus on wrapping up development of Steam Marines and designing its sequel.

 

Thanks for reading,
Mister Bums

You can contact me at yjseow@worthlessbums.comTwitter, or leave a comment below!

 

Steam Sale – Community’s Choice

It’d be great to be a daily deal! Front page, on sale, and, and, and!

Steam Marines was on sale for 50% off on Steam during the Holiday sale. On December 27th it was picked as the Community’s Choice and went on sale for eight hours at 75% off for $2.49 USD.

The impact was enormous:

Communitys-Choice-Graph
Steam Marines traffic on Steam

The graph is for unique visitors to the Steam Marines Steam Store page. When Steam Marines first launched on Steam I did not have Google Analytics set up properly, so the first portion of the graph is missing data. But that doesn’t really matter. What matters is the spike on the right side, because you can barely tell where the launch and Autumn Sale dates are if I didn’t mark them.

 

Community’s Choice > Passing Go.

I don’t think you can plan on being selected for Community’s Choice. I don’t know how they picked the games I was put up against. I don’t know by how much I won the vote or how many people voted. But of course I voted for myself like the scumbag dev that I am:

Community's Choice Options
Community’s Choice Options

Aside: The developers of Kinetic Void (one of the other games in the Community’s Choice pool) were kind enough to post congratulations on the Steam Marines discussion forums. Very groovy of them (i.e. not scumbag devs.)

 

Getting on the front page in this manner also drove traffic to the official Steam Marines website:

Community-Choice-Steam-Marines-Website-Traffic
Steam Marines Official Website Traffic

 

Armchair Analysis

As you can imagine this also drove sales on non-Steam platforms, although the numbers were nowhere near as large as on Steam itself despite the fact that all purchases on non-Steam platforms also grant Steam keys. I probably could have made that fact more obvious, although I suspect it would not have made an enormous difference. I am certain that many of the sales were impulse buys.

When Steam Marines launched on Steam it did so at a price of $7.99 with a launch discount of 10% off. During the Autumn Sale it was listed at $7.99 and on sale for 50% off. During the Holiday sale it was listed at $9.99 and on sale for 50% off; also 75% off during the Community’s Choice eight hour period.

It may have helped that Steam Marines was the lowest cost game for that voting pool at $2.49 versus $5.00 and $6.80. Anecdotal observation of other Community’s Choice picks suggests that people go for low cost versus cost saved compared to base sale price, presumably in part due to Steam not readily providing that information on the voting panel. Of course even if I had the actual voting data it’d be difficult to draw that conclusion. Lots of other variables to consider!

 

Closing Comments

Being on Steam’s front page with a steep discount, if only for eight hours, was amazing. Lots of sales, lots of discussion, and an avalanche of bug reports I’m still wading through.

I apologize for not being able to hand out solid sales numbers. However the site traffic should give you a general idea of the relative power of sales and the Steam front page on the data point that is my game.

If you’d like a few more words on this general subject matter, I hope you take a look at my older blog post, Commercial Indie Games & Risk.

“I didn’t get to craft my Snowglobe thingy because the Steam servers were all wobbly whacked.” – Unknown Scumbag Dev

 

Thanks for reading,
Mister Bums

You can contact me at yjseow@worthlessbums.comTwitter, or leave a comment below!