Trinity: design ruminations

This is not a detailed review of Infocom's Trinity, because Jimmy Maher has just finished that job. His sequence of posts (1, 2, 3, 4, 5, 6, 7, 8, 9) puts the game into its context in Infocom's history and, more broadly, in the history of the Atomic Age (remember that?) and the Cold War. Go read.

Inevitably Maher comes around to the question of the ending -- the "...what just happened?" denouement. (You can read just that one post if you're familiar with the game.) It's not the first time, of course. Maher links to a Usenet thread in which we went 'round this topic in 2001.

It's generally agreed that the plot logic of the ending doesn't really hold together. In fact, my teenage self was moved to write a letter of complaint to Infocom! I received a gracious response -- I think it was written by Moriarty himself -- which basically said "The game ends the way we felt it had to end." Which is unarguable. (This letter is in my father's basement somewhere, and one day I will dig it out and scan it with great glee.)

But today I am moved to be argumentative. If I were the author of Trinity, what would I have done?

(Oh, sure, I'm being presumptuous too. All due apologies to Moriarty. But we're both thirty years older; we're different people than the author and player circa 1986. It's worth a rethink.)

(I will assume that you've played the game and read Maher's post. If not, massive spoilers ahoy.)


As everybody has pointed out, Trinity is already constructed in the language of whimsy and metaphor; it starts out with a Lewis Carroll quote and builds from there. So expecting rigid logic is a fool's errand. Nonetheless, I do want a story to make sense when read at face value. (James Nicoll: "I don’t mind hidden depths but I insist that there be a surface.") Or, if the logic goes all Looking-Glass, it should do so in a thematic way.

Trinity offers the notion that the first atomic bomb would have "blown New Mexico right off the map" if we hadn't sabotaged it. Atomic bombs are vastly more powerful than we think. The little 20-kiloton blast that 1945 witnessed was "quantum steam", a side effect of changing history from a catastrophic New Mexino disaster to the timeline we know.

Maher discusses this in terms of eternal tragedy. Fine, I'd buy that -- except that it matters that atomic bombs don't work. Or work differently. Oppenheimer and Teller were wrong! All the physicists since then have been wrong. You can't just drop that into the story and not care what it means. Politics: all the mad calculations of MAD were orders of magnitude off-true. Science: the notion of fusion power, whatever that's worth, is built on quicksand. That's not a theme of "history is inevitable, we have come full circle" -- it can't be, because our history isn't what we think!

Or else the game isn't even about us, but about some other universe full of people. Sucks to be them.

But how else could the story have been cast?


Trinity could have followed through on its implied promise: you will prevent Trinity. Thus you prevent Hiroshima and Nagasaki; and the nuclear detente of the Cold War; and the envisioned nuclear conflict which ends civilization. A game which goes down this road is clearly pablum. It trivializes every triumph and disaster of our postwar history with a jovial "Well, don't do that then!"

Alternate history is tricky at best. We've all seen "if this goes on" think-pieces, which project some pet peeve into (inevitably) some variety of jackbooted dystopia, all in three smug pages and a glowering byline. They're laughable. You can just about build a respectable novel this way, if you spend the pages to develop an actual world and characters; if you have the human insight of an Orwell or a Walton. Infocom's shot at this was of course AMFV, and we generally agree that it didn't work. The world they packed into 256k of Z-code was just too sketchy.

For Trinity, whose body was a solitary metaphorical puzzle-quest, to develop a vision of a nuclear-free utopia in the last scene -- it would be a joke. We'd have no reason to care, and no reason to believe it beyond the author's "I said so." Scratch that plan.


Trinity could have ended by snatching the candy out of your hand. You begin in our history, foreseeing a nuclear war. You try to sabotage Trinity to prevent it. But you cannot: the Laws of Time (or whatever) are immutable. Thus, all comes to pass. We got the Bomb, they got the Bomb, we are rushing towards the end.

This would be bleak. (Bleak is already on the table, of course.) It would fit Maher's discussion of the moment of the abyss, the Great Change in the midst of inevitable tragedy.

But, on the other hand, you'd have to make it work as a game too. It's hard to make failure work as a satisfying ending of a puzzle-quest. Possible, of course! But Infocom had already done Infidel (with mixed success, although teenage-me was satisfied). Repeating that ending would make it seem even more of a gimmick.

You'd have to rearrange the ending, anyhow. Infidel works because the final puzzle has powerful narrative momentum (Indy always finds the secret treasure!) and a direct link to the tragic ending (the tomb has One Last Trap). If Trinity's final puzzle is sabotaging the bomb, you're going to sabotage that bomb. Any other outcome would feel like a failure to solve the puzzle. If the puzzle were to reach the bomb -- and reaching it truly felt like a climactic moment -- then the player might accept some other denouement. But, more likely, it would feel like a cheat.

A variation would be for the protagonist to refuse to complete the sabotage. It's hard to imagine the player buying into this, though. You'd have to spend the whole game arguing for the preservation of history. Sure, erasing it all is empty polemic, but -- faced with the awful alternative -- the player engaged with the story has every motivation to try it.

So scratch that too.


We might leave the final choice unresolved; leave the future in the protagonist's hand, and thus in the player's imagination. This avoids both the unsatisfying failure and the just-so story of success. If done barefaced, though, it would be just as unsatisfying as an unsolved puzzle.

One can imagine ways to make it work. Perhaps build the entire story around choice, with visible glimpses of alternate outcomes for each scene. More ambitiously: have every major puzzle embody a choice, so that multiple solutions serve as multiple paths-not-taken for the story. These wouldn't have to form an exponentially-branching tree; a collection of independent (but irrevocable, in the story-world) choices would make the point. Have the paths-not-taken hover and haunt the player. Now the player, facing an unknown and unresolved ending, will do the work of imagining the alternatives for us. Or so we hope.


Being me, I have to suggest the indirect, metaphorical ending. You leave the conclusion open to interpretation: what was dream, what was metaphor, what was the hallucination of a brain being incinerated in nuclear fire?

Infocom went some distance in this direction, or we wouldn't have long blog posts about the ending to begin with. But I'd say they provided a single clear narrative for the ending -- terse, but clear. Other aspects of the story (such as the time-loop nature of the Wabewalker and their corpse) are left more open; to me, more satisfyingly open.

This stuff can be made to work, if you spend the game building up plausible hypotheses. And the author has to have a logical framework, even though it's not explained to the player. I'll admit up front that I have such a framework for Hadean Lands, and no, I won't talk about it... But I'll go through the process of imagining what might underlie this alternate Zarfian Trinity.

The hallucination-while-dying gag is even more of a gimmick than the Infidel ending. Go ahead, accuse me of using it anyway. Well, if Terry Gilliam can pull it off after Ambrose Bierce closed the book on it... But we won't try to repeat it here.

Nearly as common is the you-are-not-who-you-think-you-are gag. This, at least, can be varied to suit the storyline. We might decide that the protagonist is a guardian of history, a peer of the giggling narrator. Or that the protagonist is the giggling narrator, talking to themself across the timeline. Or maybe the protagonist is Oppenheimer?

Not Oppenheimer, let's say, but all of the innocent (or guilty) bystanders in each of the history scenes. You are not the London vacationer; you take their viewpoint temporarily, up to the point where they enter the explosion. Then you take the viewpoint of a Russian technician, and so forth. The realization that you are in a different body in every scene would arrive gradually. This would require a different approach to some scenes, of course. (There is no NPC viewpoint in space, and the Bikini test -- the dolphin perhaps?) Then, at the Trinity site, you are Donald Hornig, babysitting the equipment until -- contra real history -- you/he find yourself at risk. There's your crucial, personal choice.

I rather like this plan; it gives us a chance to read the story from a real person's perspective, rather than the Infocom-style everynerd. (Of course, at the time Trinity was being written, Hornig was teaching down the street at Harvard! There's a real-people-fiction discussion to be had there, but I won't get into it.)

All of these storytelling gimmicks, while certainly gimmicks, serve to refocus the player's attention on the story. That's why I keep coming back to them. Rethinking everything that's happened from a new perspective is, well, thinking about everything that's happened! And when your ending is difficult to accept, it always helps for the player to figure it out rather than being handed it on a plate. It gives 'em a sense of investment, right? That's the point of interactive narrative in the first place.


Finally (for this post) we have the ending in which you choose between our history and some more terrible one. This was Moriarty's option, and I think it's workable. My objection is to how Trinity framed that choice: as a forked history in which neither choice is really our world.

Can it be reframed? Not, I think, with "sabotage the bomb" as the final puzzle. If the winning outcome is our world, the bomb must go off as planned. Perhaps the player discovers some deeper threat -- aliens? time police? paradox itself? -- and must divert, at the last moment, from sabotaging Trinity to defeating this enemy.

"Paradox itself" is a tidy way to frame the threat: the bomb must go off, or history evaporates in a puff of logic! Except that this really falls back under the "immutable Laws of Time" scenario we covered earlier. It comes off as a cheat.

No, we need an enemy that the player will feel good about defeating. Aliens are too out-of-the-blue. Nazis are too Godwin (even in a WW2 game scene). Time travellers could work; a faction from the collapsing Soviet Union, perhaps. (Science fictional in 1986!) Say they pose an extreme threat -- say, a plan to change the outcome of the war, followed by a joint Nazi-Soviet hegemony of the world?

This would have to be developed at some length, and again, it's unclear whether Infocom had the resources to pull off a solid alternate history. But it's the option I'd try. If, you know, I knew anything about history.

This entry was posted in Zarf on Games  and tagged  , , , , . Bookmark the permalink.

16 Responses to Trinity: design ruminations

  1. Duncan Stevens says:

    You are not the London vacationer; you take their viewpoint temporarily, up to the point where they enter the explosion. Then you take the viewpoint of a Russian technician, and so forth. The realization that you are in a different body in every scene would arrive gradually. This would require a different approach to some scenes, of course. (There is no NPC viewpoint in space, and the Bikini test -- the dolphin perhaps?) Then, at the Trinity site, you are Donald Hornig, babysitting the equipment until -- contra real history -- you/he find yourself at risk.

    Interesting! But--not to nitpick--how do these vignettes fit together into a single narrative? Trinity made you travel to each scene and do something/retrieve something necessary for a puzzle of some sort (artificial, but it kept things moving forward), but if you're a different person each time, it's hard to see how that could work. The dolphin retrieving the coconut doesn't get the coconut milk into a recipe in the out-of-time world. And if you're just there as a witness with nothing to do, it's not as engaging.

    Maybe one way is that you're a single intelligence all along that gets transported into the mind of various different people, and you acquire *information* that you need, rather than actual objects. The Nevada technician/dolphin/Russian scientist/Nagasaki plane crew member each hear some vital tidbit of information about the alien invasion or whatever the bigger plot turns out to be.

  2. Jimmy Maher says:

    Options 3, 4, and 5 are all very interesting, but would make of Trinity to a greater or lesser degree a radically different game from the historical tragedy that I think Moriarty really wanted to write. If you're giving the player a choice at all, as in option 5, you're directly undercutting the theme of "the weight of all that history, crushing you." Options 3, 4, and 5 could also all to a greater or lesser degree be very problematic to pull off within a 256 K Z-Machine, although obviously if you're thinking in terms of "how I'd make Trinity today" this doesn't apply.

    One thing that came up in the email discussion Duncan and I had was just when the ending was actually added to Trinity: i.e., was it planned all along or something of a kludge added at the last minute to make the theme of historical tragedy work while still giving the player some sort of satisfaction at "winning." There is a beta on the IF Archive, but it's a very late beta, probably the last before final release, and so doesn't tell us much. The Infocom Fact Sheet lists a number of earlier alphas and betas I believe, but they don't seem to publicly available anywhere.

    That said, what feels to me like the perfect solution just hit me:

    Have the game play exactly as it does, let the player sabotage the bomb... but *it doesn't make any difference*. The Kensington Garden attack still happens. When the scientists went to look at the bomb, they found the cut wire, beefed up security, and did the shot successfully a week or two later. According to real-world (as opposed to adventure-game) logic, this is after all exactly what *would* happen. It's not as if they're going to scrub the entire Manhattan Project because of one failed test, especially if that test was quite obviously scuppered by sabotage. This might bring some slight variations on when and how World War II ended, etc., but the arc of history would remain intact. Such an ending would perfectly support all of the game's themes: the weight and inevitably of history; my hobby horse of Trinity as tragedy; etc. And it's also a neat and much more logically consistent time-travel puzzle -- the timeline you inhabit has *always* been the one where the first Trinity test was sabotaged. The trickiest part would be figuring out how to convey this with sufficient grace and subtly to the player. Left unexplained, the ending would come off as a complete non sequiter. But if anyone at Infocom had the writing chops to pull it off, it was Moriarty.

    • Duncan Stevens says:

      Also interesting. The problem, to my mind, is that there's no motivation for the protagonist to go through the loop more than once; if going back to 1945 is only going to change the date of the test, why bother? There's certainly motivation to get out of London, but at that point I would expect the protagonist to live out his/her days in the out-of-time world. But if there's no motivation to keep going, then why bring the umbrella to Nagasaki? Introducing the avert-a-larger-evil goal at least establishes why the loop has to keep going.

      • Jimmy Maher says:

        But doesn't the current version to some extent fall victim to the same problem? Wouldn't the Wabewalker at some point just decide to let the test happen on the theory that it's better to blow up New Mexico in 1945 than to blow up the world fifty years later?

        One thing that's not entirely clear to me: does the Wabewalker actually *remember* after each trip through the loop? If so, the game we play is his first time through. If not, it could be his millionth or billionth -- which is somehow creepier.

        • Duncan Stevens says:

          I guess the problem with "New Mexico died to avert further nuclear war" is: who's to say an expanded 1945 explosion would have been the last? Even minus the inhabitants of New Mexico, doesn't the U.S. want something to hold over Stalin, and doesn't Stalin have a motive to get one of these atmosphere-igniting bombs to threaten the U.S. with? To be sure, the arms race would have looked different--one or two of those things would be plenty for MAD purposes--but it's not clear to me that it wouldn't have happened.

          The way the epilogue is written suggests that Wabewalker does remember--indeed, it suggests that the memory comes back when Wabewalker sees the old woman. I don't have the text on hand, but you stand there staring, even after the air-raid sirens go off, and eventually the roadrunner nudges you, and the Voice in your Head says "It's time." As you don't act the way you did the first time through, the suggestion is that now you have new knowledge/memories.

    • Andrew Plotkin says:

      Duncan Stevens: "...how do these vignettes fit together into a single narrative?"

      Your idea of information puzzles is a good answer, yes. It's also not inconsistent to allow each character to retrieve some physical item from the world. Part of the game arc would be figuring out who *you* (the Wabewalker) are; what rules apply, what your goal is. You're body-jumping for a reason, so fetching items can be part of that.

      Jimmy Maher: "...but would make of Trinity to a greater or lesser degree a radically different game from the historical tragedy that I think Moriarty really wanted to write."

      Sure, I said this was presumptuous.

      And who knows how much what Moriarty "really wanted" was a function of what ideas he had(*). My story ideas and puzzle ideas always influence each other, so if I'm redesigning a game thirty years on -- and as a different person -- I'll allow the underlying story to change as well.

      (* Brian, you don't have to jump in here, but feel free...)

      "...was it planned all along or something of a kludge added at the last minute"

      Speaking as an author, it seems unlikely that you'd set out writing this game without knowing what the ending was going to be. Yes, there's always the possibility that your idea falls flat in early playtesting and you decide to rethink it.

      I like your week-later ending, but it does sort of fall into the sector I labelled "snatching the candy away". That is, the player's natural response is "Oh, I almost solved this puzzle but I missed a step. Restore and try again." You'd have to convey *both* "you've really and truly solved the game" and "your actions were futile." As you say, it would require chops.

      • Jimmy Maher says:

        I think you could pretty clearly signal that the game was, in fact, "won" without too much difficulty. I read a lot of vintage reactions to Infidel in preparation for that article, and I don't recall anyone actually being confused by it to the point of trying to replay to change it. One dead giveaway is of course a maxed-out score. No, the difficulty would come in trying to convey what really happened quickly and gracefully, without subjecting the player to a huge wall-of-text infodump for a final grace note.

      • Duncan Stevens says:

        The Infocom tester who posted on rgif back in 2001 said that a lot of the key text came along late in the process:

        https://groups.google.com/forum/#!searchin/rec.games.int-fiction/trinity$20%22chip$20hayes%22/rec.games.int-fiction/cWLE0_KC64c/CuG5cCj2Jw8J

        ...which doesn't mean it wasn't in Moriarty's head in some form, of course.

        • Jimmy Maher says:

          In his Get Lamp interview, Jon Palace said that Moriarty's working methods were quite unique amongst the Imps. He liked to first test just the skeleton of his games, the mechanics of the puzzles and the storyworld, with barely a Scott Adams level of text. Then one day you he would send out the latest version for testing and it would suddenly be filled with rich descriptions. Just FWIW...

          • Andrew McCarthy says:

            Sounds very much like the sort of thing done by the designers of classic graphic adventures. I've heard that Ron Gilbert had essentially the same working method on The Secret of Monkey Island, for instance.

  3. Peter Pears says:

    Please include the information that this blog post contains spoilers. I've seen in Planet IF; I haven't played Trinity yet and completely skipped the Maher post that, he warns, has massive spoilers. I was able to recognise early on that you speak quite freely about the ending of Trinity in your post; please add the necessary "Here there be spoilers" for everyone who follows PlanetIF and has yet to play this masterpiece. Thank you.

    • Andrew Plotkin says:

      My time-travel powers have allowed me to add the line "massive spoilers ahoy" to the original post. *Everybody but you remembers it as always having been there.*

  4. Andrew McCarthy says:

    That's really interesting that you point that out--nearly the same quote occurs in the ending of Moriarty's LOOM.

    "Bobbin. It's time!"

    Of course, the line is different in the talkie CD version, which rewrote almost all the dialogue to accommodate early voice technology.

    • Andrew McCarthy says:

      Oops--I'm afraid I was trying to reply to Duncan Stevens' earlier comment about the epilogue of Trinity.

  5. Duncan Stevens says:

    Another possible ending:

    You sabotage the test, successfully enough that Truman pulls the plug. (This might involve more than cutting a wire.) And, through a heretofore-unnoticed exemption (Section 802(c)(6)) in the Time Travel Paradox Prevention Act, you find yourself back in London, in a different future.

    But this future isn't all strawberries and cream. It seems that defusing the nuclear age turned the Cold War into a hot war starting in 1949. World War III, unlike the first two, lacked a conclusive ending; tens of millions died before the war finally petered out in 1958, and by that time Europe was in a pretty grim state, to say nothing of Japan and the Middle East. Paris, Istanbul, and Tokyo are mostly memories at this point; London didn't fare too well either. Furthermore, the postwar world economy has been largely stagnant; millions remain out of work, with no prospects of finding any. The underlying message: the will to power, and the willingness to inflict suffering on others in the name of some abstract principle, didn't disappear when the test was sabotaged. Atomic weaponry were simply a more efficient means of inflicting that suffering. The game could leave it ambiguous whether the player goes back through the loop--or which future he/she chooses. (Obviously, something other than a nuclear attack would have to give the player the escape route, but there could still be an intriguing white door somewhere.)

    Obviously, this would be a tricky line to walk. Too gloomy a forecast, and it's just another failure message--and the game comes off as pro-MAD. Too sunny, and the player doesn't feel the "weight of history" or view the meddle-with-time project as a mixed bag at all. But, as you've said, Moriarty had the chops to pull it off.

Leave a Reply

Your email address will not be published. Required fields are marked *

*

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>