Intellivision’s revival was announced a year ago, at the 2018 Portland Retro Gaming Expo, and they’ve been fairly quiet since then. But it looks as though, unlike Atari, they have been working steadily toward product viability. They are now a year away from their announced release date, and they actually have some games in the works.
A recently posted teaser video showed a number of games, purportedly being played on Amico hardware, that look like re-skinned versions of popular early-80s classic games such as Moon Patrol, Asteroids, and others, but with modern graphics.
I’m not sure what to make of these games yet, but if they’re merely re-skins of old games, that will not be enough to make the Amico successful.
To be compelling, Amico developers need to imagine (and then realize) an alternative history. In the same way that steampunk is a retro vision of what science fiction could have been from the perspective of the 19th century worldview, Amico games should be an expression of the vision of what videogames could have been from the perspective of the late 70’s/early 80’s worldview.
This is a really difficult thing to envision correctly. There really is no one right way to do it, but its a rich space within which different designers could allow their imaginations to run wild, and come up with better or worse visions of what this alternative history might be. Really, we would not have a single alternative timeline, but rather a seeming infinity of branching alternatives.
There are several interesting main branches for this alternative future history.
Platform Steady State
Imagine that the original Intellivision never stopped production, that a new generation of computing hardware never arrived, but that games for the 1979 INTV system continued to be developed, using the same hardware, and the same constraints.
What would the games being developed in 1990, 2000, 2010 or 2020 look like for such a system? To answer that, we need look no further than the homebrew community. Or more broadly, look at the homebrew community for that generation of systems: Atari 2600, Intellivision, Colecovision. Over time, developers can become very familiar with the hardware, discover programming tricks and techniques to get the absolute most out of the system, and refine the quality of the software to its ultimate state.
Now try to envision the present day homebrew community, with the approval of copyright and trademark holders, creating authorized works. This gives us sequels to our favorite games, not just original new games, or thinly veiled “homages” or “parodies” of our favorite games.
This is one potential alternative history, and since it is one that we already have, in large part, in the form of the homebrew gamedev scene, it’s certainly viable. But it’s probably not the most interesting branch. The homebrew scene is awesome and it would be great to see it promoted on a modern platform that has mass appeal. But my hunch is that the appeal of the homebrew scene is narrow and niche. That is to say, the homebrew audience today is limited to a relatively small number of enthusiasts who have an old system that still works, and this number is most likely to dwindle as the enthusiasts die off and as the hardware breaks.
Unlimited Budget, 1979 tech
Now let’s go a bit further, and envision an Intellivision-like system that has just a little bit more capability. Not enough to put the system into a “next generation”, but more like what Intellivision could have been with an unlimited R&D budget and no compromises, subject to the limits of 1979’s tech. Whatever that tech is, it still has to fit inside a console-sized box, but it can be a no-compromises design within that box.
To properly envision this, we’d need to know the internals of the INTV rather intimately, as well as what other components were around at the time that could have been chosen for the system instead, but weren’t due to, primarily price considerations.
I don’t really have such an understanding, but imagine that we had a computer and electronics catalog from 1979, and it had all the components that went into the production of the INTV. Most likely, these parts were not the most expensive available, but the cheapest that could still do the job. Maybe in some cases they had to drop features entirely, in order to get the design to fit within a price point.
But you have unlimited budget, so you can select the top end parts, and include all the features. So maybe your CPU is speedier, maybe you have a bit more RAM, and the RAM is faster. Maybe this gives you the capability to do slightly higher resolution graphics modes, or more colors on screen from a larger palette. Maybe the IntelliVoice expansion module is built-in as standard equipment, and maybe it’s capable of speech synthesis that’s a little better. Games can have more sprites on screen without flicker, larger sprites, more animation frames, bigger levels, more items, etc.
Essentially this describes something like an “Intellivision III Plus” — the INTV3 was a backward compatible system released in 1982, which had a faster CPU, better graphics chip with double resolution capability and more sprites and colors, and yes, integrated IntelliVoice.
So this branch of alternative history more or less actually existed, but its life was cut short when Mattel canceled the INTV3 in 1983 due to the Crash. As such, we didn’t really get to see the full potential of what the INTV3 could have delivered. I’m not familiar with the homebrew community centered around the INTV3 hardware — most homebrew development focuses on more popular systems, such as the Atari 2600 or the NES, but to the extent there is one, the games it’s producing are essentially what this branch of alternative history might have looked like.
Moore’s Law Goes Linear
Let’s envision a future world wherein Moore’s Law failed early, and computers grew slightly more powerful, rather than exponentially.
Imagine that Mattel released a new system in 1983 or 84, and again in 1988, and then in 1993… and so on, until today. But rather than each generation of hardware roughly doubling in speed, memory, and so forth, the advances were merely linear. So, imagine that 40 years of integrated circuit micronization R&D results in a CPU clock speeds increasing from 1MHz to 40MHz instead of 4000 MHz. RAM increases from 2 KB to… just a few megabytes.
This is basically what personal computers were in the early-mid 1990s, only it happens 30 years later in the alternate history than it did in the real world. I had a Macintosh Centris 610 purchased in 1993, which had a 20MHz Motorola 68000 series CPU, 4MB of RAM (upgradeable to 16MB or so), and an 80 MB hard drive. A game console in the mold of the Intellivision likely wouldn’t have had a hard drive, so swap it for a cartridge slot, but let each cartridge hold ROM images of around 1-4MB on average, with maybe a spectacular, premium game topping out around 12-16MB. Graphics on my system were limited to 8-bit or 16-bit color depth, depending on resolution mode, and I think my 14″ monitor did 800×600, but maybe it was only 640×320.
If you programmed a system with specs like that at the bare metal, in assembly, dedicating the hardware fully to running the game, and not an operating system with layers of services for applications, that’s actually quite a powerful system, which would be very capable, even without hardware accelerated 3D graphics.
Keep in mind that software developers would have a very strong incentive to push hardware to its limit, and much more time to perfect performance optimization techniques for a stable hardware platform that doesn’t fall into obsolescence after 18 months. There wouldn’t be the mentality of just making software that’s good enough and shoving it out the door, and trust that next year’s computers would be powerful enough to run it acceptably. In such an alternate history, software craftsmanship would be highly advanced compared to in our world, and developers would work in low-level languages.
The thing is, I feel that a few generations advanced version of what Mattel could have evolved the Intellivision into doesn’t really feel like an Intellivision. It feels like… well, like an early 1990’s pre-PowerMac Macintosh. But it’s not really what Intellivision would need to be going with the Amico for if they wanted to tap into nostalgia for what Intellivision was in 1979-1982.
Something else then?
To do this, we need to understand the resource constraints of the Intellivision in a bit more detail, and then come up with some way to relax only one or two of them, in just such a way that it enables some new possibilities for games that are similar to the games that typified the Intellivision era, yet were just out of reach of what the console was actually capable of.
As an example, we might look at what sort of games we could have seen with twice the resolution, or an extra color per sprite, or removing the limitation on the number of sprites per horizontal scan line. Or ROMs that could be an extra 4KB in size.
Another way to think about it would be to think about ways we could augment the Intellivision by granting one capability from present day systems. For example, what sort of games could an Intellivision have played if it had always-on, low-latency network connectivity, that enabled it to connect to the internet and communicate high scores or enable online multiplayer? But (since the games are still based on EEPROM technology that is not re-writeablee) cannot apply updates or patches (other than new content). Or what would an Intellivision game look like with WSVGA resolution and thousands of sprites instead of NTSC resolution, but the same color palette?
We’d want to remember that these are not actual limitations we’re designing into the actual hardware, but rather artificial limitations we pretend to exist, accepting them as design constraints, in order to force us to come up with creative solutions for the problems that would arise in game development as a result of them, so that we could end up with games that have a distinctive flavor to their style that evokes an Intellivision aesthetic.
So summing up:
Do I want a game that plays exactly like Astrosmash, but has 4K resolution graphics, with 32-bit color depth, and the sprites look photorealistic, but it’s just a re-skin of the original Astrosmash? Nah, I’d probably rate that about a 4/10 (and the original Astrosmash is easily an 8/10). You don’t get a better game just by multiplying the graphical output.
What I would be excited about is a spiritual successor to Astrosmash, with extended play mechanics, and blocky, low-res graphics that evoke the original, but have enhanced effects, maybe some trails or motion blur, some neon glow, something like that. Something like what Tempest 2K was to Tempest.
A retro-inspired game like Geometry Wars would be a good fit for the console.
I’d also love to see a game that looks like it could have been drawn by an INTV or Atari 2600, but which is much larger in scope than what those games were capable of. Think Pitfall 3, and it looks like Pitfall I and II, but it has a larger world, more variety of obstacles and creatures, and additional play mechanics that weren’t possible on the original hardware.
Intellivision’s FAQ on Games suggests that this sort of thing is pretty well what they have in mind for the console, although what the games actually end up like obviously remains to be seen. They are looking to launch with a catalog of about 40 games, which is a respectable number of exclusive titles to start out with.
It is still a very difficult market to establish a new console in, though. Intellivision’s approach is to target a different market segment, rather than try to compete directly with Microsoft, Sony, Nintendo, and Google, and this seems smart, but whether it will succeed is up to Intellivision’s execution, and the market to decide. The casual/family gaming market is already pretty well served by mobile devices, and it’s difficult to say whether the Amico will appeal enough to consumers to get them to buy enough to make it a commercial success.