In console videogame development, it’s said that games created at the end of a console’s life tend to be significantly more technologically advanced than games created towards the beginning, because development companies will have learned new tricks and ways to fully exploit the power of the hardware they are working with.

This is “common wisdom”, but I’ve never really noticed it in any consoles I’ve played with. Admittedly, Super Mario Bros. 3 seems a lot more advanced than Super Mario Bros. — the former had many more states for the hero to be in (tanooki suit, raccoon tail, giant boot, etc.), slanted terrain to slide against, different terrain types (such as ice, quicksand, etc.) which had different properties. — but none of these features seemed like technological marvels beyond of the capacity of the NES.

If anything, they seemed more the limitations of the imagination of the game designers, or an inherited relic from when all Nintendo games were arcades games (see Donkey Kong), intended only to entertain for a couple of minutes, not hours. Mario 3’s innovation seemed to only require more storage space (easily added on the cartridge), rather than a faster CPU, nor more memory.

On the Sega Master System, one of the first games was Alex Kidd in Miracle World, and one of the last was Wonder Boy in Monsterland 3. Again, I did not notice any significant technical innovation. From the screenshot, it looks like Wonder Boy 3 may have had a parallax scrolling background whereas Alex Kidd did not (but it’s been too long since I’ve last played that game, so I can’t recall if the background was parallax or fixed).

Genesis has Altered Beast from 1988 versus Bugs Bunny in Double Trouble from 1996. Super Nintendo has Actraiser from 1990 versus Harvest Moon from 1997. In both cases, there did not seem to be any significant technological improvements.

Since this idea — that games at the end of a console life are technologically superior — is accepted as common wisdom, I must assume that it is true, but that the technological advances are so subtle as not to be perceptible to even relatively technologically savvy people such as myself.

This is why I appreciate this Making Of Shadow of the Colossus, in which the developers highlight some of the technological innovations they are most proud of. First, let’s compare this late Playstation 2 game with an earlier PS2 game. Armored Core 2 from 2000 versus Shadow of the Colossus from 2005.

I won’t do a specific comparison between SotC and ArmoredCore 2, but what I will highlight are the features which many next gen (i.e. Xbox360 and PS3) games are bragging about today (2009), which SotC had already implemented back in 2005, on inferior hardware.

• Metal Gear Solid 4 brags about Self-shadowing on the PS3, but SotC did it on a PS2.
• Ghost Recon Advanced Warfighter brags about HDR and Blooming on the Xbox360, but SotC did it on a PS2.
• Flower brags about fur on the PS3 (grass bending in the win is basically implemented as fur), but SotC did it on a PS2.
• brags about motion blur on the PS3, but SotC did it on a PS2.
• Oblivion brags about LOD for landscape rendering.
• Assassins’s Creed brags about Inverse Kinematics and ragdoll physics.
• Killzone 2 brags about particle effects to simulate smoke.

I never really noticed any of these features when playing through SotC, but now that they’ve been pointed out to me, I must admit I am very impressed with what the developers, Team Ico, were able to pull off with such limited resources. I would love to see more coverage like this for the other systems, so that I can more fully appreciate how close to the edge the Nintendo team pushed the NES when they made Mario 3, for example.