Super Punch-Out!! Blindfolded Race at GDQx 2019

Watching a lot of speedruns, as I’ve said before, can give you a distorted view of what video game difficulty is actually like. Speedrunning has been a going hobby for well over a decade now. New strategies are worked out and evolve. If they’re good ones, they become a part of everyone’s runs and are further refined. If they’re not good ones they’re discarded. What I’m saying is, the state-of-the-art advances. It doesn’t recede. People keep getting better. TASes are even more optimized than that, and are at this point really a completely different process, more scripting and exploring program function than playing a game in the traditional manner.

There may come a time, eventually, where, confident that runs have been perfected, speedrun becomes less focused-upon. Then after a period, people may come back and try to match the records of old. Or, maybe people will just stop speedrunning games, at least from the NES and SNES era. Many of these games are deep, but they aren’t an inexhaustible resource.

When you watch a speedrun, even one that’s three or four years old, you aren’t watching the effort of one person, but of a chain of people stretching back. Runners watch each other’s attempts and try to improve upon them. There aren’t many secrets.

Watching speedrunners who have played these games hundreds of times may cause you to think that the games are somehow easy. One way some challenge can still be preserved is in attempting challenge runs, like completing a game blindfolded. Like Punch-Out. Finishing Punch-Out blindfolded. That’s something that people do, but it’s still pretty challenging.

And it’s generally considered that Super Punch-Out!! for the SNES is the hardest Punch-Out game to do blindfolded. I’d think that that would be NES Punch-Out, since Tyson at the end is very random and can knock Little Mac down in a single punch throughout the first half of the first round, but the commentators on this video say it’s SNES Punch-Out, and I believe them. In this race, both players take a defeat at one point! That’s not something you often see at GDQ.

If you know what you’re doing Super Punch-Out is a fairly short game. This whole run (a race between two people) takes about 22 minutes from start to end. One nice thing about this race is that it doesn’t become a case, common in speedrun races, where one player jumps into the lead and stays there the entire rest of the race. The lead changes a couple of times, and is up in the air until the last fight.

If you’ve never encountered Super Punch-Out!! before, you might be surprised by how much it differs from the much better-known NES game. NES is very much a game of pattern recognition and exploitation. The SNES version brings back the two arcade games’ power meter, adds a dizzy mechanic that can affect every opponent, and just has a lot more randomness. Not blindfolded it may be a little easier than the NES game, it doesn’t have any opponents like Mike Tyson. But it still has its challenges, as much personality as the 8-bit game, and further, doesn’t lean nearly as hard on ethnic stereotypes, and those are all good things!

Make Random Items Appear Where You Want In Animal Crossing New Horizons

It’s three years after the release of Animal Crossing: New Horizons, famously introduced to a human contact-starved world right when efforts to contain Pandemic 2020 were at their height, unlike now when the world has largely decided to let the immuno-compromised fend for themselves. This isn’t the place to say what I think about that, but it is the place to write something that, had it been known in 2020, might have helped people out a lot.

The following is paraphrased from my Mastodon thread on the matter.

Every day, the game hides up to 10 100-bell coins, 5 wasps nests, and 2 random furniture items in trees on your island.

If you care about finding any of these things, there is a way to make the game put them where you want them. Selling wasps and items made from nests can bring in about 10,000 bells a day. The furniture can be given to villagers to help increase friendship. The coins aren’t worth much, admittedly.

Doing this, you can easily get the items you want each day without searching among all your trees. I use it to get the two random furniture pieces each day.

To make this trick work, you must have _exactly 17 non-fruit trees on your island_, enough to generate all the randomly-placed tree items. They can be cedar or other, plain trees.

If you don’t discover one of these items on a day, it’ll be left there for following days. It only places new items if the old ones haven’t been discovered, up to the maximum of each type. The trick relies on this fact.

Decide which of the categories of items you want to lock down the location of. Starting from that location, shake each tree until you find one of the objects you care about. In the example images I use furniture (the leaf icons), since those are a type of item it’s useful to search for quickly. You’ll probably want to have a net on hand, and maybe some Medicine, in the likely event you find one or more wasps’ nests.

Once you found the kind of item you want, stop shaking trees for that day. On the next day, all of the items you discovered will be found among the trees you shook that day, just in different places. Now, shake only the tree you want the item to appear in. If it’s not the item, keep shaking the trees you had shaken before until you find it. With luck, you’ll find it before you shake them all. Now stop shaking trees again.

Doing this day after day, you can get the item narrowed down until it appears where you want it to be generated. Once it appears there, stop shaking for that day, and then don’t shake it again on following days. Start over with another of the type of item you want to narrow down.

By working like this, probably within a couple of weeks you can get all the items you want generating where you want them. So long as you don’t shake any other trees, those will always produce the ones you want. If you shake other random trees, you’ll introduce uncertainty into what’s generated.

In this way, I have produced two trees that always produce furniture every day, generally without fail. This trick has been tested for months on my island.

The only drawback that I can find is, a couple of seasonal events (Christmas and Easter) are known to disrupt it, since they can repurpose some of your trees as non-random types for a little while. When the event ends, you’ll probably have to set it up again.

Magnavox Odyssey Ski

I turned fifty in February. No one believes me when I tell them my age, and for that reason I’m not too loud about it in person-one can very easily get tired of hearing “no way” in response to an admission like that. I don’t quite know what to make of my generally-youthful appearance. I know that things like that don’t last forever, so I’m trying to enjoy it without too much trepidation.

Anyway, fifty years in 2023 is a neat match for the history of commercial computer gaming. I was an early reader, so in the back of my brain I still remember hearing about the introduction of the Atari VCS, a.k.a. the 2600, in 1978, when I was but five. But there is a game system older than me: the Magnavox Odyssey.

The Odyssey was such a strange beast, in several ways. The first commercial TV-based home gaming computer, it didn’t have true interchangeable game cartridges; all of its games were hard-wired into the console. All of its cartridges came with the unit, and inserting a cart simply completed a circuit that told the internal electronics which game to run. The games had extremely simple graphics, we’re talking pre-Pong-level. Games cleverly used screen overlays, with translucent elements, to provide playfields and tracks. The computer didn’t even count score on-screen; it relied on players to keep track of that themselves.

The Magnavox was visible and remembered, moreso than really obscure machines like the Fairchild Channel-F and the Bally Astrocade, enough that it inspired a much more powerful (but still pretty weak) successor with true software called the Odyssey 2, or, as the machine’s trade dress stylized it, the Odyssey2. It’s funny: I have never once in my whole life ever heard anyone call it the Odyssey Squared.

Stories from the time tell us that the Magnavox Odyssey was stymied in the marketplace through an expectation, they say created by its advertising, that one could only use the Odyssey on a Magnavox TV. That wasn’t true, one didn’t need a Magnavox TV to use the Odyssey, but since a major component of each game had to be physically affixed to the screen, and its location and size had to match up with what the game’s design expected them to be, one did have to have the right size of TV to play properly.

The Odyssey was actually a little bit more of a success than commonly represented, surviving its first Christmas season without being discontinued, and even inspiring the production of some cheaper cut-down versions that only played some of the original’s games.

If you’re wondering how this worked out in practice, let’s take a look at Ski, one of the games that the Odyssey played. Here’s a page describing it, which links to its printed rules that came with the machine, and here’s video of two young people gamely trying it out:

I love that the TV in this video has TOSHIBA prominently displayed across the bottom of its screen.

Infamously, the patent that Magnavox owned on the Odyssey was used to terrorize the game industry for a while. According to an NPR obituary on Odyssey team leader Ralph Baer in 2014, Magnavox eventually garnered over $100 million on infringement lawsuits, far more than it ever earned in sales, up until the patent expired in the early 90s. Consider: a patent issued on a machine invented before I was born was used to attack game makers into the SNES era. And this wasn’t even a software patent, which I would hope everyone recognizes sucks by now: the Odyssey was a physical machine, and its patent was of the ordinary kind.

(Yeah, it’s 2023 and I’m still banging on about patents. I know there are legitimate uses for them. In some industries, unpatented inventions are ruthlessly copied by others. I’m not here to argue about them in general.)

Ski’s overlay

To play Ski, the player uses the weird two-dial controller that came with the Odyssey, that worked kind of like the dials on an Etch-a-Sketch, to move a square on the screen. The square is the only visual element of the game: the whole rest of the screen is a black field. Over this, the player puts the Ski overlay, which depicts a simple course as a dashed line, winding around drawings of trees and mountains. The overlay is opaque except for the dashed line, so as the player moves the square with the controller, it lights up the line. The idea is to get the square from the start location to the end by one of several routes. The player is “penalized” for hitting obstacles only in the sense that they are left to apply their own penalties: participants are expected to be honest in this, maybe with the aid of a referee to do the time/score recording.

Magnavox developed a new version of Ski, called Ski Festival, that was planned to be released for a successor to the Odyssey, not the Odyssey 2 but a different successor that was cancelled. Little is known about it, other than an image in a sales brochure. People have zoomed in on the image and attempted to recreate the game from it. Video of this valiant attempt is here.

Both of the videos in this post come from the Odyssey Now project (Youtube) from the Vibrant Media Lab at the University of Pittsburgh, which seeks to preserve and provide information on this important console.

Dall-E Sucks At Drawing Most Classic Video Game Characters

I had originally scheduled a post on this for a couple of weeks ago, but WordPress gained what I will euphemistically call a personality at that time, and the post developed a “critical error” whenever I tried to edit or view it. I kept pushing it back in the hopes of being able to figure out what was the trouble, but the trouble refused to be be figured out. So eventually I just remade the post.

Whether it’s intentional or not, if you ask Dall-E to depict a number of classic video game characters or elements, it’ll show itself to be surprisingly clueless. Here’s what I got from it:

Pac-Pac

From 2016 comes Pac-Pac, a Pac-Man style arcade game for an unusual platform: the Commodore Plus/4!

The Commodore 64 was famously intended to be a family computer that could also play games. The Plus/4 was intended as more of a business machine, without hardware sprites or the 64’s capable sound chip. It still had 64K of RAM though, and some productivity software included built-into the system in ROM. It could also output more colors than the C64, was clocked at a higher speed, and had a simpler design with fewer chips.

Still though, the lack of hardware sprites was a big limiter for games, which remained a driving factor for microcomputer adoption. Having no sprites, in Pac-Pac, the player’s surrogate character and the ghosts are both drawn on-screen in software, which consumes a lot of processor time. The game still runs at a decent rate though, and is fairly fun to play.

It’s best not to play Pac-Pac like Pac-Man. Despite a superficial resemblance it’s much the different game. The ghosts don’t have different personalities, and don’t coast confidently through the maze, but jitter about uncertainly, and randomly. This makes them generally easier to avoid, but it also means they’re prone to camping in the vicinity of uneaten dots. You’ll find you’ll have to lure them away from the last dots in the maze to get to them safely. You’re more likely to lose a Pac from daring their presence a little too closely.

Unlike Pac-Man there are no energizers, so there’s no way to attack the monsters yourself. On later boards the ghosts slowly get more aggressive, and they move faster. There’s also a timer to force you to go after dots. Eating randomly-appearing fruit replenishes the timer by a bit. There are also Question Mark items that appear in the maze, that can produce good or bad effects. They’re usually good though. The only ways to earn extra lives are by earning 5,000 hard-won points or, occasionally, from a Question Mark.

To play it you’ll probably need an emulator, such as the one from WinVICE. RetroArch can play it with its xplus4 core, which comes from the VICE project.

I am informed that the author of Pac-Pac, Skoro, passed away earlier this year. He made a plethora of work for the Plus/4, as shown by his page on Commodore Plus/4 World, from 2019 to all the way back in 1988. 31 years is a good long while, and I hope that the fruits of his labor will be enjoyed for decades to come.

Pac-Pac, from Commodore Plus/4 World

The Battle Vortex Audio Show

Ultima Online is a wonder. World of Warcraft debuted in 2004; Ultima Online started in 1997. And it’s still going!

When it was new podcasts were not yet a thing! Podcasts arose from the fusion of periodic MP3 audio content and RSS feeds, in October 2000. Yet when UO was new there was an audio show called Battle Vortex that covered it. So we can’t call it an Ultima Online podcast, because those didn’t exist then, but it was a whole lot like one.

Battle Vortex had been gone from the internet for awhile, but now the whole show, 156 episodes, has been uploaded to the Internet Archive! It is a priceless snapshot of the early days of MMORPGs, and it’s heartening to see it housed someplace that will preserve it.

Battle Vortex (Internet Archive)

Why Speedruns And Not Score Attacks?

This editorial doesn’t necessarily reflect the views of this blog. However, blogs don’t have views anyway, so what would that even mean?

Here is a question you might not of thought of. Speedruns are, after some years, still very popular, streamers still chase records, sometimes a matter of shaving off tenths of a second off of the previous time, and AGDQ and SGDQ continue to bring in millions for charities.

But, why? Why is it speedruns that have gained the interest of so many runners and spectators? Many games have their own method of measuring player skill: points! The score record chase is even much older than speedrunning, dating back to the heyday of Twin Galaxies. It’s even encouraged on the attract-mode vanity boards of countless arcade machines. So why is it that quick-playing has attracted so much attention, and not high-scoring?

The first thing, of course, is that time attacks (playing to finish quickly) is universally applicable, while not all games track score. Score keeping has become a lot less common in recent decades. What does it even mean to score attack Metroid Prime or Resident Evil IV? And often a quick-playing game is a lot more exciting to watch than one where the player just seeks to increase some abstract value. Fast play is easy to understand, but high point awards are often not immediately accessible to a viewer who isn’t already familiar with the game’s systems.

But more than that, many games have very sloppy point discipline. If a game doesn’t have a timer, but does have a score and respawning enemies, then there is no reason, from a point maximizing standpoint, to not just regenerate the same enemy over and over, a boring way to play but still, by the strict rules of the game, valid. To some extent this can be accounted for through out-of-game rules, like how Twin Galaxies will disallow certain types of play that just seek to increase points in an empty manner.

It’s not always easy to decide what counts as actually playing the game and what’s meaningless farming, which makes the allowance of some types of play a judgement call, and any time an official’s subjective opinion becomes an important part of the legality of some behavior, you’re going to end up with people trying to push the boundary of what is allowable, and as we see from professional sports, that means no end of arguing about whether a referee or umpire’s call is valid or not.

This doesn’t even cover scoring randomness. Ms. Pac-Man is a great game in many ways, but one aspect of it that makes it less suitable for score attacks is that, in long games, the point values of fruit becomes such a huge part of the score. After the seventh board, the fruit generated in Ms. Pac-Man are random, and can be worth anywhere from 100 points for Cherries, to 5,000 points for a Banana. Up to two of these can appear on each board, and once the game progresses past the point where ghosts can be eaten for points, the value for the higher-valued fruit easily overwhelms all the other scoring in the game, up to the kill screen at around board 144. (Ms. Pac-Man doesn’t have a definite kill screen like Pac-Man does, but a variety of possible screens.)

When point awards are random like this, getting a score record in an individual game becomes a matter of luck. What that practically means is, players who attempt more runs are more likely to get a lucky game that gets a record. Essentially, record chasers must utilize the law of averages: a person who plays 100 games is much more likely to get a lot of Bananas in a single game among all of them than a person who has only played ten.

But even so? Lots of games were made explicitly with scores in mind. After decades where it was a common, sometimes even primary mode of play, I feel like playing for points is fairly neglected now. I mean, I’m not going to go on a rant about young folks trying to get their games over with without stopping to savor them. Just, you know, it’s not bad to play for points.

I fully realize that this is a topic that no one cares about, and even I am not really that concerned with it. It’s just an arbitrary value to maximize. “Yay, I’m X good! I’m Y better than I was before!” Yeah not really terribly important.

Romhack Thursday: Mario 64 Character Swaps

On Romhack Thursdays, we bring you interesting finds from the world of game modifications.

The world of romhacks ranges far and wide, from dumb graphics hacks that put Wilford Brimley in place of Mario to full games that are unrecognizable from the software they were made from. We usually try to focus on more substantive fare, but today we present three hacks that mostly leave Mario 64 unchanged, except for giving the overall-wearing movie star a rest in favor of one of three understudies.

Super Cream 64

In the case of Super Cream 64, it feels like there’s enough to go by despite the core game, in most senses, being the same as Super Mario 64.

Saying that it’s a simple character swap is both dead accurate and wildly understating the effort that went into this. Nearly all the characters have new models, and there’s a few more in there as well. Mario has been replaced with Cream the Rabbit from (a couple of) the Sonic the Hedgehog games. Cream’s one of those characters that barely got any main game appearances before being relegated to the likes of guest roles in kart racers, so unless you’re as soaked in the deep Sonic lore, as I appear to be, you might never have heard of her, or Cheese, a Chao that follows her around (don’t ask me what a Chao is, I could tell you but the answer would probably not be useful to you) her mother Vanilla, or her friend Blaze the Cat, who are also in this hack. You may know of Amy Rose, who’s also here.

That’s Amy, right there.

There are some play differences. Somehow Cream can fly, kind of, in a gliding sort of way, which makes the game a little bit easier. A small number of areas have been changed. But mostly this is a game for people who haven’t gotten their fill of Super Mario 64 already, and who want to play it as a different character. It’s almost as light and fluffy as its protagonist, but it’s evident that a lot of care has gone into it.

Here’s a bit of gameplay to show you what it’s about:

Super Cream 64 came to my attention when a friend was putting together a console-playable copy of it for her Sonic-obsessed kid. It’s amiable and mostly harmless. I can’t say it’s my usual kind of thing (the game itself really is mostly Mario 64, and Cream is a little too cutesy for me), but maybe it’s more to the taste of some of you out there?

This post started out being just about SC64, but here’s a couple of other character replacements that may be amusing, for a few minutes at least.

Captain Falcon 64

Excepting his appearances in Super Smash Bros. games, Captain Falcon is rarely playable on foot. Captain Falcon 64 suggests a run style and moveset that’s pretty much how one would imagine he’d play in an officially-made 3D platformer.

Super Mario 64: Sonic Edition

The logical intersection between Cream 64 and Falcon 64 would, of course, be Sonic 64. He’s even faster, and harder to control, than Captain Falcon. There are some gameplay changes here, including the ability to become Super Sonic, but I couldn’t tell you how. Maybe explore it yourself and see if you can figure it out.

Super Cream 64, by Gamebun (Sonic Fan Game HQ)

Captain Falcon 64, by PastaPower and Blakeoramo (SMW Central)

Super Mario 64: Sonic Edition, by Thodds (SMW Central)

St1ka Points Out Neglected 8-Bit Games

St1ka is an interesting Youtuber from Portugal who often covers the Brazilian gaming beats. I should link to him more often when he does videos on his standard beat, as the gaming scenes outside of the US, Europe and Japan don’t get nearly as much exposure as they should.

This time though he is looking at retro gaming with a more general focus, pointing out interesting titles from decades past. They’re not making nearly as many 8-bit titles as they used to, so finding a cool old game you’ve never tried is almost as good as if they were making them now!

The games covered are:

  • Konami’s Getsu Fuuma Den, Japanese horror for the Famicom,
  • Rod Land, a single-screen arcade platformer with a good Famicom port,
  • Kickmaster, a deep combat action game for the NES,
  • Lunar: Walking School for Game Gear, a Game Arts-produced spinoff JRPG of the main Lunar series that doesn’t have much to do with its origin, and has a happy slice-of-life anime feel,
  • Psychic World for the Sega Master System, a neglected platformer where you pick up magic-like powers as you progress, which got a Game Gear port with unexpected differences, and both deriving from a Japan-only MSX version with its own differences,
  • Power Blade and Power Blade 2 for the NES, and the Japanese version of the first one, Power Blazer, stylish action platformers akin to Mega Man, but staring an Arnold Schwarzenegger clone who wields boomerangs, except in Japan where you play as a serious-looking little kid,
  • Cave Noire, a roguelite game for Game Boy that I’ve written about in the past, and personally vouch for,
  • Daikatana for Game Boy Color, a 2D and rather improved version of the infamous 3D PC game,
  • Power Strike II for the SMS, or its Game Gear remake that’s completely different, but they’re Compile shooters so you know they’re going to be awesome, and really you can’t go wrong with any Power Strike/Aleste game, or ZANAC or The Guardian Legend on NES come to mention it,
  • Queen Fighters 2000, a bootleg game for Gameboy Color that outright cribs the style from Gals Fighter for NeoGeo Pocket Color, but also includes a bunch of random characters from other properties for the hell of it, since being a bootleg game anyway why not,
  • Aliens Neoplasma is a 2019 release for the Spectrum that makes excellent use of that system’s graphical quirks to increase the game’s atmosphere,
  • the action RPG Dark Arms for the NeoGeo Pocket Color,
  • early survival RPG Survival Kids for the Game Boy Color,
  • Phantis, a cross-genre sci-fi game for the Spectrum,
  • Shatterhand for the NES, which feels inspired by Batman on the same system, and
  • Kabuki Quantum Fighter, also for the NES, which also seems inspired by Batman in its play style, and was also developed by Hal Laboratory.

INCREDIBLE 8-BIT Hidden Gems You Never Played (Youtube, 41 minutes)

Fumble Dimension: Breaking 1,000 Points in Madden 22

It’s been about five months since we looked back at Jon Bois’ Breaking Madden’s Super Bowl game where they set one team up to maximum stats, the other team at minimum stats, and just ran up the score. They found out that Madden 2013 stopped counting score at 255 points. Counting manually, at around 1,500 the game called a penalty on a play that wasn’t run, and the instant replay presented the most frightening image a sports video game has ever generated, the Football Fetus:

Soak that one up for your nightmares.

They stored a team’s score in a single byte in 2013! It was a striking example of how EA Sports, without competition for decades, basically views Madden as a no-effort money printer. Licensed NFL football is either Madden or nothing. What are you going to do, play with no-namers? Pshaw!

Breaking Madden is like ten years old now, and the rights situation hasn’t changed. EA Sports continues to squat possessively on its golden football egg, with no end in sight. But Breaking Madden made some internet waves in the time since. Maybe they’ve gotten their development act together? Maybe?

In the time since then SBNation has switched these kinds of things that they do to Youtube videos, and rebranded them as Secret Base. Jon Bois has become quite the Youtube sensation there in the time since, making a lot of very well-regarded internet documentaries.

One of the Secret Base subseries is called the Fumble Dimension, which is a similar kind of attempt to break sports video games, just in video. About 11 months ago they (mostly Jon’s associate Kofie) again took a hammer to the most recent Madden, 22 at that point, and tried to run up the scoreboard. This time though they did themselves. No, they didn’t try to win against an inferior team. They played a team to lose, against a team set at maximum AI, and tried to let the computer score as much as possible. Here it is (23 minutes).

The good news is, the score no longer ceases to count after 255 points. The score is free to rise up over 1,000. The game’s final score broke 1,700, and didn’t break 2,000 only because Kofie was feeling hugely bored playing terribly on purpose for play after excruciating play. They theorize that 7,000 may well be popular with optimal sub-optimal performance, but they leave that demazing feat to some other intrepid failure.

But while they have wisely decided to store player scores in more than one byte now, there were several other hints that Madden 22 is just as haphazardly constructed as Madden 13 was. Players would try to run off the field, held back only by the walls of the stadium. The announcers announced a lengthening series of safeties each as the “second safety of the game.” And while the score counted correctly, an assortment of player stats were scored increasingly inaccurately as the self-induced drubbing continued, some dipping into the negative as they grew, a sure sign of uncapped signed values.

When I posted on Breaking Madden, I took the opportunity to diatribize about the decay of the Madden games, and how the series should either be given renewed resources or the license just be allowed to pass to other hands. I won’t bore you with yet more harping on the point now, I’ll just say, please NFL, hand your license over to someone who actually seems to care.

Fumble Dimension: We tried to break the Madden scoreboard (Youtube, 23 minutes)

Another Retro Blog: Retro365

If blogging is ever going to come back from its loss to social media, it’s going to have to be from going more social itself. By that I mean links between blogs, making it easier to surface sites to others. Not only directly, but by helping to raise each other’s Google rank, although I think time has shown that Google is a fickle friend to people producing material for the Web, any site prominence you gain can easily be wiped away the next time they change their algorithm. Bigsites naturally get traffic just from being established, and other sites try to become big by gaming their placement with hyper artificial SEO techniques. Meanwhile us littlesites have to succeed largely by being interesting and direct views, as well as what traffic we can gather through followers through RSS, social media, Patreon and other sources. And there’s no reason not to help each other out. We’re not in competition between us. Any cross link, wherever, strengthens us all.

Here’s one from me. Retro365 has a vast collection of gaming media from the earlier days of home computing, and has been going for about three years now. They’ve got lots of demonstrated software on their Youtube channel. If you have an interest in learning about, or just seeing this stuff, they’ve got plenty for you.

Here’s a few choice items from their channel. There’s the classic CGA DOS game Paratrooper (the player doesn’t last long, only a minute):

Dungeon! for Apple II, published by TSR themselves in 1982 (32 minutes):

Oil’s Well for the Atari 400 and 800, a variant of the arcade game Anteater (8 minutes):

And a complete playthrough of comedy adventure game classic Sam & Max Hit the Road for PC (an hour and 47 minutes):

Retro365 blog, and on Youtube.