The Stack

I seem to have hit the point where this game suddenly becomes hard. I’m off in the depths of the forest trying to rescue Little John from basically an entire army, because that’s what it took to subdue him. For most of the map, my usual MO stands up: Render enemies unconscious, ideally by having Robin sneak up on them one by one, but with a Strong Merry as a back-up in case I wind up pulling aggro. Tie them up before they come to. When possible, help this process along with trickery like luring enemies into an ambush. (Honestly, the efficacy of ambushes is one of the things that really makes this game feel Robin-Hood-like.)

The big problem, then, is that this mission is the first to feature a mounted knight, and I don’t know what to do about it. He’s unprecedentedly fast-moving and hard-hitting, and I haven’t found a way to take him down permanently — using the Strong Merry’s biggest attacks, I’ve managed to stun him to the point where he’s no longer considered a valid target for attacks, but he recovers alarmingly quickly, and apparently can’t be tied up while he’s on that horse. (Would I have to tie up the horse too?) Up to this point, to minimize deaths, I haven’t been shooting people with arrows. For this guy, I finally break out the bow. He shrugs it off.

Clearly, then, the best way to deal with him is to avoid him altogether. The problem is that he’s positioned to notice when you go after Little John. Even if you avoid his line of sight, and do stealth take-downs of all the other guards in the vicinity, the act of freeing John from his bonds ineluctibly causes another soldier to appear and raise the alarm. Maybe the key is to just book it at that point, but even that seems tricky. I can’t outrun that horse.

One thing I’ve been shamefully contemplating is making a sacrifice. Get the Strong Merry into a fight with the knight elsewhere to distract him. But is that something Robin Hood would do?

Maid Marian has joined my party, at least for one mission. Her special unique ability is Spy, which clears the fog of war in a large radius around her. It’s not the best of powers, because fog of war really isn’t that big a deal in this game; people you haven’t seen are displayed, just rendered in a sort of mottled foggy grey devoid of details so you can’t tell if they’re friendly or not. (And you can probably take a pretty good guess about that anyway from their behavior; anyone flanking a doorway or marching in formation is probably a guard.)

At any rate, the choice of power is clearly meant to reflect her role in the story: she’s Robin’s “man on the inside”, maintaining a presence at court to supply the outlaws with information. (Which means it’ll be less narratively justified if she takes up residence in Sherwood, but these gameplay things are only loosely connected to their narrative justifications anyway.) All of the named characters seem to have such an ability like that, one loosely tied to their character. Robin’s unique ability is throwing coin purses, which isn’t quite giving to the poor (especially if he takes them back once you’re unconscious), but is at least adjacent to it. Will Scarlet, the youngest of the band, has a slingshot that can stun foes briefly. The manual tells me that Friar Tuck has a “put down beer” command, possibly to distract foes or make them less observant. Little John has the ability to give other characters a boost up walls — due to the reuse of maps in missions, I’ve seen walls with the cursor rollover for this action even though I still can’t perform it.

It seems strange that Robin’s special unique ability isn’t related to his famous skill with a bow. There’s a biggish problem with the source material in this game: it encourages you to avoid lethal damage, but that’s the only sort of damage a bow can do, so I’m avoiding doing one of the things that the title character is most famous for. It tries to make up for it a little in the non-plot-related ambush missions, where you attack various carriages and caravans carrying gold through the forest. These missions give you traps and snares triggered by shooting at targets, another fairly thin gesture towards the established character.

Two more critical path missions down, plus a couple of optional caravan ambushes to get extra cash. I don’t really have a use for all this cash yet, but the story has got to get around to asking me to ransom King Richard at some point. There’s nothing really unexpected about the story. It’s taking care to give all the Robin Hood fans out there what they want, introducing the old familiar characters one by one: Will Scarlet, Maid Marian, Friar Tuck. No Little John yet, although I’m sure he’s coming. Will Scarlet is the only one who’s actually joined my outlaw band back at the base in Sherwood Forest, though. The rest of my recruits are generic Merry Men.

The generics all have individual names, and can be trained up in combat individually, and the game tracks their individual health and inventory from one mission to the next. In terms of abilities, though, they come in just three types, which the manual calls Strong Merry, Aggressive Merry, and Mustachioed Merry. I’ve been finding the Strong Merry to be by far the most useful in combat: under player control, he can do big sweeping moves that knock out multiple people at a time (including any friendlies standing too close). But you can’t neglect the skills that the other Merry types bring to missions, like healing and lockpicking. Robin Hood stories and Robin Hood games sometimes make the mistake of making Robin simply the best at everything anyone does, making you wonder why he’s keeping everyone else around at all, but I think the RTS influence in this game’s ludic makeup helps it to avoid that, making the focus on the coordinated actions of a team.

What do all the extra duplicate Merries do while you’re on a mission? Whatever jobs you’ve assigned them to back at Sherwood HQ. They can produce supplies to take on missions, such as arrows and healing herbs and throwable coin purses (with extra-weak seams to make them burst on impact and scatter their contents over a wide area), or they can train in hand-to-hand combat or archery, or they can just rest and heal. This is all done with the same UI as missions: home base is a kind of Sherwood Treehouse Playset that characters walk around in, with stations you can leave them at to tell them what to do. The annoying thing about this is that with dozens of controllable characters, you can’t select who to control in the more convenient ways, like the 1-5 keys. To select someone, you have to click on them, which can be difficult if they’re walking around. If you try to click on someone and miss, the game interprets it as telling the currently-selected character to walk to the spot you clicked on. I really wish sometimes that I could pause the game and still give orders, Baldur’s-Gate-style — not just here, but in combat too. But I suppose that’s the downside of the RTS influence.

I haven’t quite gotten through the second mission yet. This is a save-a-prisoner-from-the-gallows mission, one of the classical Robin Hood scenarios, and it’s quite daunting from the start, especially if you’re trying to cooperate with the game’s discouragement of killing. The place is crawling with soldiers, and worse, they’re clustered in small groups and watching each other’s backs. With only one playable character, you only have two nonlethal ways to dispose of enemies. First, you can sucker-punch them before they draw their weapons, which only really works on isolated individuals, and only on the weaker sorts at that. Secondly, you can throw a purse of gold into their midst, causing them to fight each other over it. Ideally, the brawl leaves only one guard standing, thus turning them into an isolated individual who you can sucker-punch. But again, the tougher guards are immune to this trick, and you can only do it so many times — you can retrieve the gold from the unconscious guards, but oddly enough, throwable purses are a limited resource.

The whole deal, then, is that you can’t do much of anything on this map with just one playable character — because this is the level that teaches you how to use multiple characters. With the man you’re rescuing, and three other condemned prisoners who just happen to be there too, you have a party of five, which seems to be the maximum the game accommodates, judging by the UI. Different characters have different abilities: this guy can pick locks, that one knows how to use healing herbs, and so forth. The two most relevant skills for the above discussion are the ability to tie up unconscious foes so they don’t pose a threat when they wake up, and for big strong guys, the ability to pick up the unconscious, dead, or bound so you can hide them where they’re less conspicuous. In short, your standard stealth-game stuff, but it takes multiple people to do. Most of the time, I’m using just those three characters: the big guy, the bondage guy, and Robin, who knows how to sucker-punch.

Not that sucker-punching is always necessary! The big guy carries a bludgeon that lets him deal nonlethal damage in combat mode. Still, the most effective way I’ve found to conduct combat is: One of my merry men engages the enemy in combat, and while he’s thus distracted, Robin comes in from the side and sucker-punches him. I’ve basically got it all down to a science now, and have been indulging in the same sort of foolishness as I did in Deus Ex: maximizing my freedom by clearing the map of all threats and stuffing them into the same few picturesque half-timbered buildings, where they are no doubt stacked like logs. This is why it’s taking me so long to finish the level.

A random conversation got me thinking about this game, so I pulled it out and played it a bit. When I first tried it in the early 2000s, it struck me as fairly original in concept: it looked and controlled like a RTS of the time, but it had stealth mechanics and puzzles. I’ve since learned that the Commandos series did it first, but I haven’t played those.

At any rate, this is a game that supports both stealth and combat approaches, and allegedly rewards choosing stealth: Robin’s ability to function as a folk hero and attract followers depends on people’s opinion of him, which is affected by how many murders he’s responsible for. I haven’t gotten far enough into the game for that to be a factor, though. All I’ve gotten through so far is the first mission, in which Robin Hood isn’t even really Robin Hood yet. Back when the game was newer, I didn’t even get that far; I got stuck in perfection paralysis, repeatedly realizing that there were better ways to do stuff as I learned the game’s vocabulary and restarting the entire level.

When run as-is under Windows 10, the game is unplayably sluggish. To get around this, I installed a patch that turned out to just be a little wrapper for running it under DxWnd. I’ll have to remember to try DxWnd for other games that display similar symptoms. It does create two new problems, though. First of all, the cursor leaves trails behind it in the game’s menus, including in-game message boxes. Secondly, it breaks the game at higher resolutions. By default, the game runs at 640×480, but the options menu lets you dial that up to 800×600 or 1024×768, and it’s a shame that I can’t take advantage of that. I’m basically stuck with 640×480, which looks brutally coarse to me, although I can already feel myself adapting to it.

[Addendum 22 Sep] It seems like the failure of the other graphics modes must be linked to the introductory FMV cutscene. This plays automatically when you start the game, and it always plays in 640×480. So if you have the game set to play at any other resolution, it has to change graphics modes on the fly. DxWnd doesn’t appear to handle that well. A lot of older games for Windows do this sort of thing, switching resolutions for FMV, and it’s never really worked very well. On every single PC I’ve ever owned, switching graphics modes takes a few seconds, with the typical result that you wind up missing the start of the video.

Admittedly, ParserComp 2023 is old news by now, but I’m not quite done with it. I declared my intention to play all sixteen of its games, and I got through only seven before the judging period’s deadline. My mistake: I got a very late start, but thought I could still get through them all, because I was really thinking in terms of IFComp, which has the famous two-hour rule to encourage smaller games. ParserComp has no such rule, and this year’s entries included two more games by the author of Finn’s Big Adventure. I did consider continuing to play and write up the entries here after the deadline passed, as blogs have no deadline, but I didn’t do that.

Still, there’s one entry that I really want to comment on here, because it’s so of-the-moment. The Fortuna, a thriller set on a cruise ship, uses an LLM to generate text. Now, from where I sit, the judgment I’ve seen passed on LLMs and generative AI in general has mostly been negative: that their lack of any mooring in reality makes them unsuitable for dealing with facts, while applying them to fiction and creative works is a threatens the livelihoods of human artists (whose livelihoods are precarious enough already) while ultimately resting on their labor without fair compensation. But the author of this piece is a true believer in the liberatory potential of the AI revolution. (I thought for a while that they might be pretending, mainly because of the illustrations: these too are AI-generated, and, while the first few shown are reasonable enough that I mistook them for stock photographs, they start turning toward subtle monstrosity before long. I’ve now played enough of the game to be reassured of its sincerity, but it’s conceivable that there’s a twist that I simply haven’t had the patience to reach.)

Now, you may be thinking “LLM-generated IF? Like AI Dungeon, then?” Not quite: the only generated text in the game is NPC dialogue provoked by freeform input. It’s in some ways a natural choice, supplementing one of IF’s traditional weak spots while limiting the damage that the LLM can do to the narrative, but with the current state of the tech, the effect is deeply peculiar. It makes the dialogue unnatural in a characteristically ChatGPT-ish way. Characters don’t really converse so much as use your words as prompts for paragraph-long essays with lots of “I believe” statements:

Journalism is an incredibly important field that allows us to gather, analyze, and share information with the world. It is the perfect blend of storytelling and investigation, allowing journalists like myself to shine a light on important issues and hold those in power accountable. Journalism has the power to shape public opinion, influence policy, and promote social change. As a woman in journalism, I am particularly passionate about using my platform to advocate for women’s rights and to amplify the voices of marginalized communities. It is crucial that we have diverse perspectives and stories represented in the media, and I am determined to be a part of that movement.

The game tells you to befriend these characters to advance the plot — I could comment on how pathologically utilitarian this view of “friendship” is, but that’s really a problem endemic to IF as a whole, not this game in particular. Befriending a character basically means asking them about their pet topics repeatedly. Which they tend to circle back to regardless, often with suspiciously similar verbiage:

> ask nancy “What are you doing here?”

I’m here as a reporter for the New York Times, covering the recent women’s rights rally in the city. It’s important to give a voice to the issues that affect women and make sure their stories are heard. This rally is a powerful platform for women to come together, raise awareness, and fight for their rights. As a journalist, it’s my responsibility to report on these events and shed light on the challenges women face in our society.

> ask nancy “I mean, what are you doing on this deck in particular?”

I’m conducting research for a new article I’m working on. I’m investigating the gender wage gap in the corporate world and the impact it has on women’s rights. I believe it’s important to shed light on these issues and advocate for equality and fair treatment for women in the workplace.

> ask nancy “Why are you on the pool deck?”

I’m on the pool deck because I’m investigating the recent increase in drownings in the area. As a reporter, it’s my job to shed light on important issues and bring attention to them. It’s not enough for me to just report the facts; I want to use my platform to make a difference. I believe that by raising awareness about the importance of water safety and advocating for women’s rights, we can start to address this problem and prevent future tragedies.

For this player, at least, the end result is that I gave up on trying to treat this as anything other than Infocom-style “ASK/TELL” dialogue with unusually elaborate “I don’t understand” messages. The LLM simply doesn’t have enough knowledge about its context within the game to be able to converse about that context meaningfully. It’s been said that the ability to figure out what you do and don’t need to pay attention to, and to filter out the unimportant, is the quintessential skill in all videogames, and it’s possibly even more paramount when the text is being produced in a void of authorial intent, where failing to realize that the nonsense is nonsense has the potential to send you down infinitely-deep but ultimately empty rabbit-holes. So there’s the paradox of this game: the whole thing is a showcase for a feature that it structurally encourages you to ignore! (But I suppose the same could be said of a lot of graphical games…)

It’s got me wondering about other possibilities, though. The dialogue feels meaningless because it’s disconnected from the rest of the game — even the “befriending” mechanic seems like it’s just scanning your input for keywords and ignoring the response:

> tell david “Physics sucks and physicists are bad people. People like you are responsible for Hiroshima. I hate you. I hate you. I hate you.”

I’m sorry if you feel that way, but as a research scientist in the field of physics, my goal is to make a significant contribution to my field and to help others through my work. I understand that there have been events in history, like Hiroshima, that have used scientific discoveries in destructive ways. However, it is important to remember that science itself is not inherently good or bad – it is how we choose to use and apply it that determines the outcome. I am driven by a desire to contribute positively to the world and leave a lasting legacy, and I strive to do so with compassion and responsibility.

David Millman is now your friend!

But what if it didn’t ignore the output? What if it tried to determine whether you had befriended a character on the basis of what they said?

Well, for one thing, you’d have to retune the output considerably to make that work. In this game, I don’t think I’ve seen any character actually express camaraderie or affection (or indeed any emotion at all). But assume that’s a solved problem. Would it make for a better game? I don’t know that it would! It might be more responsive, but it would be less intentional, less about the story the author wants to tell and more about figuring out how to manipulate the LLM.

But on the other hand, game devs routinely let general-purpose libraries handle things like physics and 3D rendering, sometimes with unanticipated effects. Will it someday be possible for developers to simply give an AI actor parameters and direction and motivation, and trust it to ad-lib its lines? I suppose that’s the dream this game represents. And I’m willing to keep an open mind about it. But we’re not there yet, and it’s not at all clear that LLM are anything more than a dead end in its pursuit.