Tallowmere 2
Work in progress. Subject to change. More fun than you can shake a kitten at.
Tallowmere 2

Lady Tallowmere has begun expanding her empire.

Tallowmere 2 is an upcoming 2D dungeon platformer game, currently in its early stages, being developed by Chris McFarland in Auckland, New Zealand.

Delve Further
Blog Post

Returning to Tallowmere

1 November 2016  •  Chris McFarland

“A sequel? No more. Just expand the original.”

Experimentation

I enjoy experimenting. It's what keeps development and prototyping refreshing. A lot of Tallowmere's features were the product of experimenting.

Experimenting with Tallowmere 2's art and animation style was also refreshing. It looked cool, felt cool. Was something new and different. I liked it and ran with it.

However, one thing I didn't grasp at the time, was how animation can change the speed of the gameplay.

Tallowmere 2's melee weapon attacks had to feel near-instant to replicate Tallowmere 1's instant attacks. I toyed with some values for character movement, bobbing, attack speed, jump height, and durations and positions for the hands if you were attacking, jumping, falling, getting hit, or even just standing idle.

While it felt nice, was it close enough to Tallowmere 1's gameplay style? It seemed good enough back in January and February, so I went off and spent a lot of the next few months tackling some menus and netcode. I figured I'd get the hard stuff in first.

After the netcode came the world map and item attribute systems. I hadn't really touched the gameplay at all for several months at this point - and in hindsight I recommend not doing this. Gameplay aside, the item systems felt good – structurally, anyway.

Showing off what isn't complete

I ended up talking about Tallowmere 2's development thus far in September 2016 at the New Zealand Game Developer's Conference.

My talk wasn't recorded, but my slides are available. (the HTML version contains a lot of animated GIFs; you can download the ZIP file of all the slides if desired, or just check out the PDF or JPGs for the static versions)

After the conference I worked more on the item system, mainly the UI, shown right.

But something started to bother me. The gameplay felt off.

An analysis

Things felt too quick. Enemies were too fast. They were running and jumping at me - which they had been doing for about 9 months! Why was this suddenly bothering me now?

Ultimately it came down to me trying to nail the gameplay for new weapon and enemy types, now that the other systems were in place. And compared to Tallowmere 1, it felt like everything in the dungeon was moving around at the speed of light.

Yet the speed and movement values I set at the start of the year were still the best bet for the visual style of the game. I tried decreasing the movement speed, but had to slow down the animation to match, which made you look like you were controlling a snail, which just looked and felt wrong. I even tried zooming the camera out further, but everything was still too fast.

Had I shot myself in the foot? Was my fluid animation imposing limits on how the gameplay had to feel? Was everything too lively?

Frustrated, I attempted to work on other systems and do other things, but it was no use; I had stumbled upon an elephant I couldn't ignore.

Seeking clarity

As some of you may know, I'm no stranger to depression. And let's just say, this anti-eureka moment, this candle-snuffing pique of enlightenment, sunk me right down for a time. I needed some time to wallow and think in angst. (well, I didn't need to, and would have preferred not to, but depression doesn't work like that for me!)

Steam Dev Days 2016

On the positive side, I had booked a ticket and flights to Seattle for Steam Dev Days for October 2016, which was only a few weeks away at the time, so perhaps the timing was perfect; something to take my mind off things, a chance to meet other game developers, an excuse to escape from the norm for a while, something to look forward to.

I attended as planned; Steam Dev Days was an excellent experience. I've uploaded a few photos. As cool as it was, I'm not sure I'll do it again anytime soon - flights are expensive!

But, back to my dilemma. Tweaking Tallowmere 2 was not getting me results. Playing Tallowmere 1 only went so far. I needed something different to try and scratch this gameplay issue. I decided I needed a barebones prototype of a Tallowmere-esque platformer. So a few days before my trip, I experimented further.

Sanity check

I had to be sure I wasn't going crazy. I needed to reaffirm my grasp with Tallowmere's game design decisions that have gotten me to where I am today.

In addition to gameplay speed, something that had been irking me for a while was, how important is art? How important is character animation? How much should one enemy be able to do? Do arms matter? Do legs matter?

To try and get my head around things, the result of a few days of coding from scratch before, during, and after my Steam Dev Days trip came out as:


While I never got around to adding AI to this prototype, I was able to confirm:

  • Art is important - there's only so much you can add to a small-bodied, armless, legless, boxy-looking silhouette:
    • How would you know if an enemy was going to cast a fireball, or pull you with its tongue, or throw a stun bomb at you?
    • A unique visual for each attack type would be required to ensure the player is aware of what to expect, given the nature of Tallowmere's crowd-controlling enemy attack types
  • Forward-flipping is awesome;
  • Comical blood is awesome;
  • There is something special about enemies standing still and mowing them down; and
  • Slower movement speed feels less daunting.

I also experimented having loot drop as colourful packages that you'd pick up, rather than having items be restricted to treasure chests, but without feeling like each piece of loot had to represent its actual shape; something abstract to hide its contents. Maybe it just contrasted with the visual style nicely, but I liked it.

And while you won't see it published, I tried adding Tallowmere 1's character sprites onto the bodies of these monsters, to get a feel for how the bobbing felt. In short, I learned that Tallowmere 1's character sprites bobbing up and down does not work. The heads don't change size, only the bodies, but there is not enough pixel data available to make it look satisfying, whereas Tallowmere 2's sprites can get away with it because there's enough visual fidelity to stretch and squeeze. Tallowmere 2's armless hand style also helps make it work; Tallowmere 1 doesn't have that.

Rediscovering the key ingredients

I feel like I've figured it out.

From coding the little prototype, playing Tallowmere 1 and 2 a lot, and taking a good, hard look at all this:

  • Tallowmere 1's gameplay works because:
    • Character movement is slow-paced
    • Everything around the player is slow, letting the player position themself with confidence
    • Because the characters look crude, seeing them non-animated doesn't break much immersion
    • The player can attack as fast as they want because it's as crude as everything else
  • Attempting to recreate Tallowmere 1's gameplay with Tallowmere 2 is unfortunately not working well:
    • Character movement has to be fast to ensure the animation looks smooth to match, but fast movement is too stressful
    • Slow-moving characters with this animation style looks mediocre
    • Animated, bobbing characters standing still (or even just not walking very much) makes them look too lively considering they're trying to be dumb
    • Characters running at the player fast leads to high stress past the fun level
    • Art is too high-res to make flat-style gibbing satisfying

So from 10 months of dabbling with Tallowmere 2 to come to these conclusions, it feels like I've shot myself in the foot in some ways, but such is life.

Where to from here?

  • Tallowmere 2's art: I like Tallowmere 2's art, animation style, and fast-paced gameplay, but no longer feel it is the right fit for a "Tallowmere 2". Perhaps the fast-spaced style will work better for a future, different game.

  • From T2 to T1: I want to bring the overworld map, the story I have planned, the new weapon and enemy mechanics, and the networking to Tallowmere 1, probably as an update. I'd have a hard time justifying a sequel that uses the same art as the original, so an update or expansion sounds better, but time will tell.

  • Adding new things: I purposely built Tallowmere for its dungeons to be changed. The start of the game is the home room, and the end of the game is when you die. Everything else in between can change. Revisiting this simple game design philosophy is an important step.

  • Unlockables: I want to try adding unlockables, but ultimately need to create a lot of stuff to unlock first. Given the nature of Tallowmere and my development style, I am more likely to trickle out multiple new updates rather than squirrelling away everything for months for one giant update. Same thing goes for story and bosses. Wouldn't surprise me if I add new weapons, enemies, and bosses to Tallowmere 1's current game design implementation, and then carve out the unlocking system and progressive world map later as a last step. We'll see what I hack together.

Takeaways

  • Code gameplay first. Get as many different gameplay mechanics in as you can early. Systems can come later. While I've learned a lot from doing network code and these item systems, ultimately the gameplay for Tallowmere 2 was not the right fit for the style I was trying to achieve, and this could have been realised earlier than later.

  • Games as a service. Are you prepared to make a game and then stop working on it? If you make a second game, are you prepared to maintain two codebases? What defines a game as "done"? There's a saying that an artist's work is never done, it is merely abandoned. Yet when you create a high-score based game with procedurally-generated rooms with potentially limitless amounts of filler, why stop? (at times I need breaks - motivation comes and goes - but it seems I can't pull myself away completely!)

Network multiplayer? Mod support? Language support? New weapons? New enemies? New game modes? New difficulty options? I've been saying "no" to myself and other people for a long time. "I'll do it for the sequel," I'd say. New features are hard. Game development is tough. But making a new game based on an old game is perhaps even tougher in some regards. Some people dislike change, especially if certain things might make the game harder, but this is Tallowmere: tough to play and tough to develop.

But code can be refactored. New classes can be created. New forms of fun can be built upon existing foundations.

So hello again, Lady Tallowmere, my old friend. We shall see about expanding your original dungeons, crude art and all.

– Chris