Discussion Forum


After my tech demos (DEMO.WAD, CHARDEMO.WAD) were published in 2015, it seemed like ZZT Ultra was well on the road to excelling at the task of capturing the next-generation ZZT launch platform. Capable of playing the old and hosting the new, it had little trouble playing most legacy adventures, and it supported many advanced features for future development.

Except...ZZT Ultra desperately needed a true next-gen game.

Smash ZZT is an example of what I call a Z-Make, or a low-fi tribute game. I decided to make it programmatically ambitious, but conceptually very simple. The basis for the game was one of my all-time favorite action titles, Smash T.V.

Simple Does It

The concept of Smash T.V. fits on the back of an envelope. Most successful arcade titles had the advantage of being self-explanatory. For my tribute game, nothing has changed. You move, you shoot, and you try not to get killed.

The ability to move and shoot at the same time, pioneered by Williams Electronic Games, appears with this game as well. Twin-stick shooters force you to pay attention and constantly monitor the environment, so that you can take advantage of the situation, no matter how dire it gets. The action is nonstop, riveting, and outrageously fun.

What might not be quite so obvious about twin-stick shooters, though, is this: their roots were in a game designer's handicap. Which then spawned an idea. Which ultimately spawned the game known as Robotron 2084.

Riding Motorcycles in the Rain

Eugene Jarvis, a man of many talents, met with an unfortunate auto accident in the early 1980s. With a damaged hand, he wasn't able to perform the button-mashing action necessary for playing and testing arcade games at the time. This is one occasion when disability can be almost poetically cruel.

But Jarvis was not one to be stopped by mere injury. An arcade control panel can be rigged in a lot of different ways by any dedicated electronics hobbyist.

Button inputs don't have to be represented by actual buttons. As Jarvis discovered, it is possible to rig a joystick to do the same thing. So he could fire without destroying his hand with downward press motions.

It was during this time that Jarvis conceived of the idea of using an ad-hoc joystick in different ways. What if, in Berserk, you could move while firing in any direction?

Whoa, did that set things in motion.

Difficulty Squared

Smash T.V. and its indirect sequel, Total Carnage, are basically extrapolations of the Robotron 2084 core concept. Few people expect action arcade titles to be fair (at least, beyond the first few levels). However, for those games whose odds are so ridiculously, and obviously, stacked against the player, there is a strange appeal that does not appear in other unfair situations.

You might have heard the expression, an audience is ready to believe the impossible, but not the improbable? Williams shooters are quite interesting specimens when it comes to credit-stealing tricks. It's mostly the case that gross "scripted cheating" events, demonstrated by many arcade games at the time, are not present in Williams shooters.

The reason for this? The games are so difficult that they don't need to cheat.

They're just really, really hard.

When a game wears its difficulty out on its sleeve like this, can you really complain?


ZZT is a game that has you rack up a lot of kills of standard enemies. The Smash T.V. kill count is comparable, depending on how far you're likely to get during a playthrough. There are lot of parallels here, although not very obvious at first.

It's also interesting that ZZT and Smash T.V, were released in the same year. It could be there was some influence, although only Tim Sweeney himself would be able to confirm it.

As I began to remix the objects for Smash ZZT, I noticed some subtle overlap of enemy behavioral concepts. This is not to say the behaviors in ZZT were outright clones, though. Centipedes act very different from the snake droids, blink walls are obstacles and orbs are enemies, and most enemy placement is ZZT is strategic, as opposed to favoring spawn-heavy situations found in Smash T.V.

But a decent idea is a decent idea--and I didn't have to extend my imagination all that far to come up with how I wanted to execute my remixed behavior.

The Fine Details

I like ZZT. I like Smash T.V. Just because I want to make a cross-over game doesn't mean this game will be any good.

So...how to make it good?

For starters, I had to make every attempt to avoid the poor practices of the past. The importance of fine details is a lesson which ZZT modders often learn far too late. I'm no exception. My first several dabblings in ZZT in the early 1990s were embarrassing, to say the least.

The main problem confounding myself and other modders was that we wanted to make something look like "X" and sound like "X", but we were not committed enough to make our final product into a polished, outstanding "X".

Mediocre work might be enough for an amateur, especially if one lacks the knowledge or tools necessary to build something on par with ZZT. One can simply take the limited engine we had and design something around it. This is exactly what modders have been doing with ZZT for over 25 years.

With the ability to tailor that much more, though, the requirements are much higher. Smash ZZT has to be functional. It has to play like Smash T.V., it has to look like ZZT, and it has to do its job right for what it claims to be.

And simply being "functional" is never enough. Fine details, including many subtle quality touches, were necessary all throughout play-testing.

You want details? You got 'em

I already wrote an article called "Sharp Edges and Sharper Difficulty." I am fully prepared to follow my own recommendations, so that character cell-locking will work in favor of the gameplay, rather than against it.

  • Actors must always be drawn over stepmines, pickups, and shield barrier graphics.
  • Pickup distance does not need to be spot-on for character cells. If the player gets within one square of a pickup, it still counts. Held-down direction keys therefore do not work against the player's "aggregate" movement behavior.
  • Movement rate of everything should reflect real velocity equivalents along diagonals, such that a 45-degree vector moves at the rate of 1/SQRT(2) instead of SQRT(2).
  • The arena background should be free of incidental floor graphics. This heads off confusion about what is object and what is background. For the same reason, the scoreboard and ammo bars must be in the GUI area as well, even though they were part of the background in Smash T.V.
  • Player, actors, and projectiles are not allowed to exit through doors even if they are open (level exit contexts being an exception).
  • Nearly all actors must be pushable as a way to prevent doors from getting blocked when funneling in new actors.
  • Unpushable actors must be dealt with in some way (tigers) when a door opens nearby.
  • Ghosting of objects must be synchronized with #DRAWCHAR calls such that #DRAWCHAR is consistently followed by an idle cycle--not acceptable to move immediately after drawing.
  • Door spawn events are contingent on which door opened previously, which doors are already open (can't spawn for door open momentarily), and type of enemy funneled in.
  • The player and other actors must never be overwritten casually, by projectiles or other types of spawned objects. Point-blank firing behavior must be handled in a special way to prevent this from happening, but ignoring point-blank cases is not acceptable either.
  • Standard lion and ruffian enemies should be just "barely" escapable at non-augmented walking speed on normal difficulty. Sheer numbers (and being surrounded) should overwhelm the player; mere speed alone should not be enough.
  • A maximum actor count must be put in place to prevent the arena from spawning out of control. Doors will not open for more spawns unless the count is lowered (as enemies are eliminated).
  • There should be no arc-over player shots. The mortar launcher weapon proved to be unworkable except in boss rooms in Smash T.V., so its usage was quite sparing. All weapons in Smash ZZT, conversely, should be both available and balanced in all rooms.
  • Hit detection for projectiles cannot "miss" targets simply because target and/or projectile is traveling fast. Multiple tests per frame may be necessary.
  • Flame and N2 weapons should have much more ammunition than other weapons. The logic here is that users have a different perception of continuous splash-damage projectile ammunition counters and long-fired, pellet-like ammunition counters.
  • Flame and N2 weapons are not shot-duped; this is rather pointless.
  • Projectiles are always ghosted; they are more manageable when non-solid.
  • All enemy projectiles should flash, identifying them as both projectiles and as potential dangers to the player.
  • Stepmines should flash, identifying them as dangerous.
  • Frozen enemies are shattered by projectiles (including enemy projectiles).

The Writing

It might seem kind of pointless to talk about the importance of writing in a thoroughly action-based game. But it is still important.

A year ago, I delivered a speech titled "Williams Electronic Games: Training People for Mass-Murder." In recounting the course of the Williams action games, I insisted that the writing, sparse that it was, managed to give the games an "edge" that you didn't see anywhere else.

You can observe silly and dark comedy in Sinistar, Robotron 2084, NARC, Smash T.V., Total Carnage, and of course the incomparable Mortal Kombat. Even Midway's 1990s sports titles bought into this perverse sense of humor, and the results were some of the most successful arcade titles ever made.

Writing is important. However, any tribute game exists in the shadow of the original. So why break your back trying to deliver any writing at all?

Well...you don't have to break anything to write something passable. Only a few parts of scripting, dialogue, or visual design of any game need to be what the average person would call "clever." Most writing decisions serve only to connect the dots, drawing from points A, B, and C. The focal points need to really shine; not much else needs to.

I stuck to a familiar formula: solid action, rapid pacing, little barrier to entry, silly and irreverent dialogue.

You're here to do a lot of shooting. After enough shooting, you'll find out why you're doing all this shooting in the first place.

There is an underlying theme, as well. Because the ZZT ship has sailed, it is undeniably retro. Therefore, it should be unabashedly retro, making fun of its own outdated graphics even as it proudly showcases them.

Enough, already

That's just about all I have to say about Smash ZZT.

So, if you're done reading, you should do the next logical thing.



There are no comments for this topic.

To enter a comment, log in using the

This page is Copyright © 2016 Christopher Allen.