A definition of a generation needs not only a technical differentiation but also impact in the way of sales volume. This lost generation had the former but not the latter and in fact existed during the decline prior to the NES. I remember this vividly as I worked at my family's computer/game store selling Atari and ColecoVision.
I really do like having this all documented and brought into awareness to all the interesting consoles that get mentioned in passing but not collectively discussed. The gen 2.5 naming is fitting.
I was very much in the Atari camp--I had C64 friends too. I felt like Atari really dropped the ball--when rolling in cash and had all the tech to release something like the Atari XEGS[0] in 1979/80 but instead they were milking the VCS/2600 profits.
And this theme of memory basically holds true through the end of the "Fourth" Generation: games on console were assumed to be memory-starved, and advancements within those generations occurred through larger ROM enabling larger and more animated sprites, more graphically distinct "worlds," full soundtracks and scripted cutscenes. In 1988, the year of the great DRAM shortage, more of the games released used smaller ROM space, and it caused a notable regression back to technically simpler productions.
When you get to the Playstation/Saturn/N64 the approaches diverge quite a bit because the bottlenecks shift according to which platform you talk about. Memory is still a major consideration, but data streaming can be either relatively easy(N64) or a considerable engineering problem(everyone else), while the inverse is the case with data storage capacity. It's easier to discuss specific platforms as stable entities from that era onwards, since there's a cleaner separation of computing resources from game content.
The author says that the 3rd gen of consoles really begins with the NES (by which I guess he means 1985), and so these early 1980s consoles can't belong to the 3rd generation.
Ok, but the SG-1000 was only sold on the Japanese market and was released in 1983. Nintendo also released the Famicom (NES) in Japan that same year.
So I think it makes more sense to say that the 3rd generation begins in 1982/1983 with the Colecovision and Atari 5200. I don't think it's fair to exclude them from the 3rd generation just because they weren't as successful as the NES.
It's not like the NES/Famicom was radically more advanced than those consoles either. The 5200 and NES have very similar hardware (both are running MOS 6502 variants). The NES was a little more powerful but that's not why it was more successful. The NES beat the competition because it was a better designed product that was ruthlessly marketed and supported by Nintendo.
The NES had super durable, and (for the time) ergonomic controllers while the 5200 and Colecovision had shitty, stiff joysticks with 9-digit numpads attached to the body. Nintendo sought out top game design and engineering talent to put out compelling, boundary pushing original titles while Atari and Coleco were still stuck in the arcade port paradigm.
>The author says that the 3rd gen of consoles really begins with the NES, and so these early 1980s consoles can't belong to the 3rd generation.
Rather than the author, I would say that it's everybody else - a general consensus - making that claim.
>So I think it makes more sense to say that the 3rd generation begins in 1982/1983 with the Colecovision and Atari 5200.
I think is precisely the premise of this article - to show that those systems, had it not been for a crash of 1983, would have been categorized as early 3rd gens instead of late 2nd gens and to give a way to somewhat correct the mistake by calling them "generation 2.5"
Sure, I guess I'm just disappointed that the author didn't take his argument to the logical conclusion and say that actually the 3rd generation of consoles began earlier, but that it was distorted in the US due to the 1983 video game crash (which was really only a thing in the US).
I think it comes down to online discussions of video game history often being very US and Nintendo centric, and the concept of a "generation" being quite flawed (the author does mention that).
> Six years after the start of the Second Generation, a new generation of consoles appeared (...) For reasons that are not entirely clear, they are considered part of the Second Generation.
As well as in last paragraph:
> (...) ‘What to call this so-called “Lost Generation?’” Should it be renamed to the Third Generation, and the Third to the Fourth and so forth? That would be confusing. While I like the term “Lost Generation,” and it lends well to how these machines were (...) doomed by the console market crash, (...) I suggest calling it the 2.5 Generation; this avoids any renaming and places it in the correct chronological order.