Missed Classic: Stationfall - When Scientific Accuracy Attacks (Deleted Scenes & Alternate Puzzles)

Written by Joe Pranevich

Last time, I started into Lurking Horror (don’t forget to post your score guess!), but we still have a bit of business left with Stationfall. Infocom is a company that we have a unique level of understanding about thanks to leaked source code, developer interviews, and a veritable “cabinet” of internal documentation. I do not like digging in too deeply while playing so that I can approach the games as spoiler-free as possible, but now that we’re done I can dig in and report back what I find. Sometimes we find a lot (like all of the half-completed Moonmist variants!) and sometimes we find less. As Steven Meretzky is the source of much of the “leaked” documentation, we have a rare view at his development process all the way down to notes from individual testers. It is not a full view of the design of the game, but it provides a unique insight into how mature he had become in game design at this point and how little effort was wasted. 

Steve Meretzky was famous for keeping notebooks of game and design ideas that he would turn back to for inspiration. Our story starts out with a question: what game should he make next?

Decisions, decisions.

The Roads Not Taken

The road to Stationfall started in 1985, but the early history seems to have been lost. We know from later notes that a 1985 pitch for Stationfall existed and that Meretzky opted to work on Leather Goddesses instead. I am uncertain if that pitch was the same or different than the Stationfall summary he presented later, but there is some evidence (comparing the Zork Zero pitches) that these may have been largely unchanged between decision points.

Fast forward to August 1986 and Leather Goddesses of Phobos was wrapping up. Meretzky needed to decide what game to tackle next. He had a notebook full of ideas, but Infocom desperately needed another hit. At that moment, he did not know that Leather Goddesses would be their best selling game in years. But which idea should he tackle next? With a veritable menu of design thoughts to choose from, he assembled a committee of Infocom luminaries to provide him the crucial advice that he needed. These included Dave Lebling (Zork games, Suspect, and Starcross), Amy Briggs (prior to her work on Plundered Hearts), Jeff O’Neill (Ballyhoo), Jon Palace (editor/producer), Chris Reeve (director, contributor to Bureaucracy), Dave Anderson (Hollywood Hijinx), and Mike Dornbrook (head of marketing). This group was well-balanced with implementers, marketers, and testers, all of whom were passionate about the future success of Infocom. He presented ideas for ten different games.

For this evaluation, Meretzky provided each committee member with a description of each proposed game, as well as their genres, difficulty levels, and approximate labor. His Stationfall example looked like this:

The long-awaited sequel to Planetfall. Since your promotion to Lieutenant First Class in the Stellar Patrol of the Third Galactic Union following a commendable job on Resida five years ago, you have failed to advance your rank or career at all. Elevation to LFC has meant that your routine of constant mopping drudgery has been replaced by a routine of constant paperwork drudgery, and your harsh commanding officer Ensign Blather has been replaced by the bubbleheaded and bureaucratic Captain Measel. Your latest assignment: to shuttle over to Station Alpha Beta Epsilon Gamma Omega 78-C-985 Sector Delta Delta VH-98, in order to pick up a crate of new shuttle fuel requisition form requisition forms. You stop by the robot pool to pick up a robot aide for the trip … one familiar figure comes bounding towards you … “Hey, take Floyd! Floyd best robot aide in pool!” After a brief trip, you arrive at the station. It is deserted, your shuttle has stopped working, you can’t raise your ship, and Floyd has begun to act oddly…

He had a good idea of how Stationfall would start, even if he was not prepared with the mechanics yet. With nearly the same level of detail, he described nine other possible games:

• Zork Zero - A prequel to Zork and effectively the game that he would come back to a year later. 
• Minute Mysteries - A spiritual sequel to Moonmist where you solve mysteries in the same location but in different time periods. This would allow the scenery to change but the overall layouts to remain the same. 
• Attack of the Freons - A sci-fi/time travel game that would at first appear to be three disconnected games, but would eventually tie together.
• Blazing Parsers - A Western-themed comedy that makes fun of the mechanics of text adventures. 
• Interactive Bible - A continuation of Meretzky’s attempts to piss off the “Moral Majority” by tongue-in-cheek retelling bible stories through a comedic game lens.
• Titanic - A “docu-drama” game where you have to transport an object from London to New York aboard the Titanic, but leading to and describing the infamous sinking. 
• The Best of Stevo - A collection of 1-2 hour gaming vignettes. 
• Initiation Day - A fraternity-themed adventure that is “essentially an interactive fiction version of Animal House”. This would have been a collaboration with Dave Anderson.
• And finally an untitled “Viable Idea” that would bring several Infocom designers together to riff on a single idea and then try to merge it all into a game, sort of like a prototyping exercise gone wrong. Exactly how this would have been “viable” is unclear to me. 

There are so many “roads not taken” here, but I have to trust that Meretzky knew what he was doing, even if some of these feel half-baked. It may be notable that three of these games involved episodic content that could be solved with less time and effort than a full game. This may have been a concern around Infocom at the time, and may have contributed to the design of Nord and Bert.

A hand-drawn spreadsheet demonstrating their inability to select a consistent rating scheme.

The results were mixed and led to no consensus. Dave Lebling wanted Initiation Day, Amy Briggs voted for Zork Zero or Titanic, Jeff O’Neill also wanted Zork Zero, Jon Palace requested either Interactive Bible or the “Viable Idea”, Chris Reeve wanted Interactive Bible, and both Mike Dornbrook and Dave Anderson hoped for Best of Stevo. All of the team members used different ranking schemes, but none of them put Stationfall as their first choice. 

As an exercise, I worked out a scale (with some reasonable choices) to put all of those rankings on the same axis. Doing that math, it appears that The Best of Stevo was the most widely-liked idea across all of the decision-makers. Clearly, the group felt that giving Meretzky the reins to do whatever he wanted would be bankable. Zork Zero and Stationfall were tied for the next choice, but with Brian Moriarty starting work on Beyond Zork, it did not make sense for both of them to work on a Zork game. Exactly how the final decision came down to Stationfall isn’t clear from these documents, but it is obvious that Meretzky put sincere effort into deciding what he should work on next. Given the state of Infocom at the time, it must have been apparent to him that he wouldn’t get the chance to work on many of these ideas.

This pitch for Stationfall is essentially the final game. Barring some adjustments to how we discover Floyd and how our shuttle gets trapped (affected by the virus rather than run out of fuel), this accurately describes the final product. Jon Palace may have summed this up best when he wrote in the margin, “Another boring hit– which may not be such a bad thing.” Palace actually proposed that Meretzky go crazy by having our adventure with Floyd take us to other times, other Infocom titles, or even just down into a cave to merge Zork and Planetfall into one design. None of those suggestions were acted upon. (A time travel concept called “Futurefall” would eventually be worked on as the sequel to the Stationfall novelization, but that is years in the future and is likely unrelated to this note.)

Let's have a game with animals!

A Space Menagerie

At some point at the start of the design process, Meretzky expanded the one-paragraph summary into a three-page plot synopsis. This became the working document for the game’s play and is shockingly accurate in nearly every way, describing how we find Floyd in the robot pool, our arrival on the station, and all the way down to our final confrontation with our friend at the bottom of the station. The ending scene is taken nearly verbatim from this summary; even Floyd’s last words and Oliver’s arrival is written out exactly as in the final product. If it was created (as it appears to have been) early in the process, then Merezky had an uncanny ability to stick to a plan. 

While the document accurately describes the plot beats, it doesn’t quite land on the puzzles. This version describes fixing machinery as being important to our success. The document also implies that animals would be the key to many of the game’s solutions:

Since the station’s machinery is all malfunctioning, with increasing frequency and dangerousness, most of the puzzles involve repairing broken machinery (which would not yet have turned bad, because it hasn’t been on) or in figuring out ways to use living things or other non-machines to perform tasks normally done by machines. At the same time, you must stay alive despite the efforts of such machines as marauding hull-welders.

The closest we came to fixing anything in the final game was swapping out some diodes. As for the animals, we are left with only two of those puzzles and those play (to my mind) as some of the most nonsensical in the game. How much sense did it make to ride a balloon creature? Or retrieve something from a vending machine using an ostrich? Not much. Perhaps, as Meretzky pulled this all together, he realized that a sci-fi game based on animal puzzles might not have worked out. 

One final difference in the design document was related to the game’s last puzzle. In the design, he describes simply shooting the pyramid after blasting Floyd. The idea that platinum could be used to defeat the pyramid does not seem to have been considered yet.

The original text of the alien’s message.

No Platinum, No Problem

Many of the day to day decisions as the game was being created have been lost now, but we still have a few curious nuggets of information. By the end of 1986, the final station map seems to have been in place (or nearly so). There are references in the files to questions about the placement and number of bathrooms and whether or not they would be gendered. (Notably, the devs went with non-gendered dorms but kept gendered bathrooms.) 

As previously mentioned, the use of platinum in the final puzzle solution was not present until after January 1987. The alien message that would, in the final game, clue us in to the metal’s use was less helpful:

Our ancient enemy will soon taste or vengeance, and the souls of our countless ancestors will at last rest in peace.

It’s obvious in retrospect that this message is missing something. Having gone through the trouble to decode the alien “writing”, not having it be an important clue would have been a mistake. This also changes how we feel about the Hunji body that we discover in the ship: there is no trace of his heroism and self-sacrifice here, only a statement of war.

Once platinum was added to the game, there were multiple iterations before that puzzle reached its final state. Initially, the foil would be found in the rec. center, presumably in the holo booth that we explored but had nothing of value. In February of 1987, tester notes described the platinum puzzle as being too difficult. Perhaps in response, the foil was moved to the barber shop mirror and a platinum detector was placed in the bank. A different tester complained later that platinum would make a terrible mirror, but Merezky appears to have been satisfied with the level of difficulty.

A visual aid to teach a game designer about chemistry.

A Refrigerated Bomb

As you will recall, Stationfall had a multi-part puzzle where we needed to find a quickly-melting explosive in the “cold storage” of outer space, bring it inside while keeping it cool, and then use it to complete a bomb in the commander’s quarters. This puzzle went through multiple iterations as Meretzky seems to have been fixated on the need to keep the explosive cold, but not sure how to do it. 

The first of these iterations involves a “bottle of CO2” that we would discover in the trading post. Investigation of the bottle would reveal that it does not contain regular carbon dioxide but rather dry ice. They are chemically the same, but obviously one is a lot colder than the other. This ice could then be used in some way to keep the explosives cold during transport, most likely by just storing the explosives in the chilled bottle. “Gary” (possibly Gary Brennan, a member of Infocom’s test team) argued at length against the scientific plausibility of this. He even provided diagrams (reproduced above!) He argued about the pressure required to keep the CO₂ solid, as well as the frostbite that we would receive if we carried the bottle without gloves. Whoever Gary was, he was a hero to detail-oriented nerds everywhere. Meretzky relented and that version of the puzzle was dropped, although it may have been adapted for Lurking Horror instead, substituting liquid nitrogen for dry ice.

A second version of this puzzle involved finding (or creating) a block of water ice outside in space, right next to the explosives. The player would need to drill an appropriately-sized hole in the ice, put the explosives in the hole, and then carry the block to the office. Code for this solution partly survives in the final version. Testing notes suggest that this path would lead to parser ambiguity as you would have a “hole” in the ice and a “hole” in the safe in the same location. Exactly what sank this idea is not documented, but it may simply have been too complicated.

The final game used a third version of this puzzle that is effectively the same as the first: eliminate the “bottle of CO2” and replace it with a Thermos. We’d get that Thermos in our survival kit, but it otherwise would have worked the same without all of that hokey “science” getting on the way. A subsequent tester had a problem with frozen explosives being kept cold in a container designed for soup and suggested (presumably with no trace of irony) that perhaps dry ice could be found somewhere in the station and added to the Thermos first. I can still hear the howls of rage that must have come from Meretzky’s office the morning that memo came through…

Map of a nearly final game. Notice the CO2 in the trading post and foil in the Rec. Room.

Going Ballooning

If you thought science nerd “Gary” would only have a problem with the dry ice puzzle, guess again! In fact, the balloon puzzle was rejected by him and several other testers due to its implausibility. In the final game, we led a living balloon creature to the station’s chapel. By scaring it just a bit while holding onto its leash, it would propel us up to the ceiling so that we could collect a star-shaped light. I didn’t like that puzzle much and it seems that I was not alone.

To make a long story short, much attention was given to the relative mass and density of the balloon creature and what would be needed to propel an adult to the ceiling. Calculations were made and it was determined that the balloon would have to be the size of a truck to have enough lift for the puzzle as designed. Options were discussed and at one point it appears that an anti-grav belt would be added to the game. This belt would reduce our mass just enough that the balloon could lift us. Perhaps this made the puzzle too difficult or perhaps introducing such an object would require adjustments to other puzzles or text as well and it was wisely left out.

Meretzky still had a problem, but he solved it in the simplest possible way. He added a footnote and moved on. This is the footnote we receive while solving the puzzle in the final game:

This is, of course, impossible. To do this, the creature would have to be an order of magnitude larger. There are two possible explanations: 1. The creature actually extends part of its volume into a parallel dimension. 2. A callous disregard for scientific accuracy on the part of the author.

Clearly, there was a limit to how much patience Mr. Meretzky had for scientific accuracy getting in the way of (in his mind) good puzzle design.

A play-tester's note.

Searching the Code

I hope you enjoyed that look through the design and testing notes! They represent an amazing window into the development of this game and the relative stability of the plot and puzzles. For each puzzle that was iterated to completion, there were many that went unremarked. Meretzky maintained a surprisingly consistent vision of this game from beginning to end.

Before we close out, I also took a dive through the source code to find aspects that were started and then commented out or were present in the game but not usable. My assumption is that these would represent ideas that had made it far in the process, but it’s impossible to know for sure. With the relative expense of storage back in the 1980s, I imagine that they would have avoided keeping too much dead code around and Infocom often used every last kilobyte of that era’s paltry memory sizes. 

In no particular order, here are a few things I found while digging through the code:

• The game contains the verbatim ending to Planetfall, commented out. It is unclear whether this was kept in as “inspiration” or if the game was intended to open with the final scene of the previous game. 
• While sleeping, our character can dream. Most of these dreams are jokes (such as a scene where we imagine Floyd like Hal from 2001), but two were commented out. These two describe our youth on the planet Gallium, our cat Swanzo, and us being plagued by a schoolyard bully. I don’t remember dreams in the game at all (but may have skimmed over them), but I assume those two dreams were removed to keep our character as “everyman” as possible. 
• There is code for Floyd being deactivated, for discovering Plato while Floyd is not present, and for Plato turning Floyd back on. I never tried to turn off Floyd in the final game so have no idea if there is a way to do any of that. (Some of the testers talk about Floyd appearing when they thought he was off so it must have been used in the versions they checked.) 
• The space village previously had a church, but it was removed. Presumably it would have been too confusing or similar to the chapel in the station.
• There is code present for driving the forklift. The forklift was hinted at in the game as being needed to transport a pallet of forms, but we only found it in the endgame when it tried to kill us. This code may have been copied from Lurking Horror or perhaps the original idea for the Stationfall puzzle was shifted to that game. 

This ends our coverage of Stationfall! Another Infocom game down and only a few more to go. We will be returning to this sub-series later however as I will be looking at the novelizations by Arthur Byron Cover when we get to 1988 and 1989. After that, we’ll have a third (non-published) novel as well as two attempts to revive the series to research. I’ll look forward to discussing that when we get there. 

In the meantime, I’m back to writing about Lurking Horror. As may be apparent from some reference above, I have been diligently playing and recording my journey. Barring any surprises in my workload, I hope to have the next post out in a week.