Bad Game + Bad Game = Good Learning Experience

Combining a game that half the world
hates with a game the whole world hates… What could possibly go
wrong?

The Plan

One of the biggest issues players have
with Hydlide is the combat system. Long before Zelda’s exciting blend
of action and exploration, Hydlide was all about walking into
things.

No, I’m not kidding.

To attack an enemy in Hydlide, you walk
into him. You and the enemy each trade blows every time you step onto
his tile or vice-versa. The deepest strategy you can apply is to hold the A
button to enter Attack Mode. While in Attack Mode, the player
inflicts more damage, but also suffers more. If it looks like you
might lose the battle, you can just move away and wait for your HP to
recover.

Fortunately, we have a genre of games
that is beloved despite its similar lack of gameplay: the MMO. 

In
classic MMO format, the player chooses a target enemy and trades blows automatically, until either the player dies or the enemy dies. This was actually the core concept I used when I started this remake, because it solved many issues:

• If I made this first person, I wouldn’t need custom attack animations and could tie into the Virtual Hydlide style of gameplay.
• Because MMO combat-style trading of blows is so similar to the old style, I could keep the stats exactly the same and maintain the original balance.
• I could even keep the Attack Mode and Defense Mode distinction, and have it function the same.

The other major advantage is that I’d been working on a 3D sprite engine the previous month, and this would allow me to use the original NES sprites in a 3D
world!

Note: The gameplay images in this article use custom sprites to avoid copyright issues. Virtually Super Hydlide Special uses the original NES sprites.

Finally, the sprite engine allowed me
to import maps from a directly from a bitmap. In order to create my
game’s maps, all I needed to do was trace the original maps with
solid colors and shrink the image to a reasonable size. The engine
does the rest.

Looking back, it really was actually a
pretty good plan. But, you know what they say about the best laid plans…

What Went Horribly Wrong

I Vastly Overestimated My Engine

I’d built my engine for
old-school dungeon crawls. The largest map that I had built was about
30×30. The smallest I could make Hydlide’s world map while keeping
that expansive feel was 83×83. 

At almost triple the tested size, this lead to a long loading
sequence. This is problematic, because the dungeons are not large and
you’ll be loading the world map often. I ended up spending a few days
implementing a streaming map, only to cut the feature later because I
realized the loading screens were a good way to give the player information they needed.

The days I spent making the map stream
were wasted. If I had taken the time to think about things prior to attempting to work through the problem, I’d have saved time and
stress.

The other set of issues came from
modifying the engine to use a looping map. The tiles that I used were very
small and I needed to draw a lot of these to fill a reasonable draw
distance. This lead to visible slowdown as the map tiles looped around.

My solution? I loaded more tiles!

During the loading screen, I loaded
about twice the tiles that I needed to fill the draw distance. This
means that, as you walk, the map is already fully loaded and in
place, but tiles in the distance are loading in real time.

At this point, I modified the system
to load only a few tiles each frame. To the player, it looks like the
map is streaming around seamlessly, but beyond the camera there are
holes in the world where it’s actually being built a little bit at a time.

Overall, the engine worked, but it sure as heck wasn’t as smooth as I had intended. I had wanted my core gameplay to be implemented by Day Seven, but didn’t actually finish until almost a week later.

I Underestimated the Conversion to 3D

Because I put so much time into making the world map work, I didn’t foresee this second major issue until I was well into the project. I have enough experience that I should have seen it coming a mile away.

The dungeons in the original Hydlide are mazes: The Original Dungeon Maps

It was forgivable in the original game, because a bird’s eye view allows you to see the correct paths. But, as a first person game, directly using the dungeons’ maps as I had planned would make things frustrating.

I couldn’t import the levels as I’d planned, and so I needed to hand-make new ones. While creating short and functional maps was not a major problem, it was an unplanned expense.

I Was Overconfident and This Made Me Stupid

Two months earlier, my previous remake project, Ghostbusters Inc., had been amazing from a work standpoint. Everything fell into place, and I ended up implementing nearly every feature I had hoped for, while still working a reasonable schedule.

My last #1GAM project had gone extremely smoothly as well. I completely designed and implemented Tactical Strike Micro in a week. It had used the same engine as Hydlide, and everything worked out well.

I went into this month with good projects on the brain, and I ignored one of my most important tools: planning!

I’m usually a fanatic about taking the time to write down a schedule, plan features and dependencies, and correctly prioritize everything. I’ve seen and often taken part in the stupidity that comes from “winging it”.

On the 13th>, I was still implementing core gameplay features that should have been either implemented or cut by now. Stress was beginning to get to me, and I could feel the developer tunnel vision coming over me. I was working hard, despite having lost sight of the big picture.

Honestly, I wasn’t sure I’d finish on time. Half a month on a small project seems like a long time, but there was a lot that needed to be done, and it wasn’t going well.

I often warn people about the dangers of blind passion, yet this project was a fine reminder that I’m not above it.

The Good

I Didn’t Panic

I’ve worked with myself enough to know that this is an accomplishment. When things go south I usually enter a purely reactive emotional state—highly effective, but misguided. But I didn’t do that here.

On the 13th, I realized that I was in trouble, and I stepped back to review my process. It didn’t take long to see that I was so focused on controlling the steering wheel that I wasn’t watching the road. I was fixing issues without any consideration for how they fit into the final product.

I took the entire day of the 14th off from development to outline a proper plan:

• Dates by which each part should be ready.
• A prioritized list of features and bug fixes.
• Questions that I needed to think about, such as “How will I help guide the player without NPCs?” With this list, I knew what to think about when I wasn’t working.

When the plan was ready, I sat down and relaxed for the rest of the evening. I knew that stress had started to affect my thinking and I needed to clear my head.

Starting on the 15th, my productivity skyrocketed. I accomplished more in the next three days than I had in the previous week.

The Style

I was amazed how good the 2D NES sprites looked in the 3D world. All the enemies, dungeon walls, and terrain tiles are straight from the NES game, yet they stand out and make the game look unique.