Sunk Costs Fallacy

[Contains spoilers for Anathem, if anyone cares.]

I’m going through two intensely frustrating things at present:

  1. The end of my first semester of sound engineering school, and
  2. Reading Neal Stephenson’s Anathem

School: it’s TAFE, which means no exams worth mentioning, and they really don’t want to fail anyone if they can help it. That means the assessment tasks aren’t too difficult, and in many cases we finish them long before the end of semester so that there’s time for marking, resubmission, late submission, or whatever. The side-effect of this is that the last weeks of semester seem to be spent mostly sitting around not doing much. Last week I had a couple of days where we essentially did nothing — or nothing that I either hadn’t done before, or which I couldn’t do myself via Google or Wikipedia in a fraction of the time — which as you might imagine I found rather irritating. Wait, that’s perhaps too much understatement. I was literally bored to tears, and yes, I do know the meaning of the word “literally” thank you very much.

By Wednesday afternoon I’d started to think the whole TAFE thing was a waste of time. Perhaps I could do better working (paid or unpaid, or most likely a mix of both) in the industry and learning on the job. I’d almost certainly find it more fulfilling than sitting in class, and over the same time period I’d probably learn more and certainly get more hands-on experience and industry contacts. When I approached one of our teaching staff about this, asking for his opinion, he said that I “might as well finish what I started”. In other words: I’ve done a semester of a course that takes two semesters to receive a piece of paper (and four semesters to receive a more advanced piece of paper, but two semesters is the first relevant exit point). Now that I’ve sunk the costs into the first half-year, I might as well go through to the end of it, even if what we’re doing in class is of only limited use to me, and not all that good for my mental health.

On another note, Anathem: a few weeks ago, probably because I was missing Wiscon, I found myself in an SF-reading mood. I wanted to catch up on a lot of the books my friends had been talking about over the last couple of years, while I’d been reading other things. I ordered an ebook reader, which would take a couple of weeks to arrive from the US, and in the meantime I hit the public library and borrowed a couple of books I’d been meaning to read or re-read, to tide me over. One of them was Anathem.

Before I start panning the book, I should mention that I’m actually a moderate Stephenson fan. This website is named after a term I found in one of his books, after all. I first encountered his work when found Snow Crash on the shelf of a general bookstore in Ballarat, sometime in the early 90s. I picked it up because the cover looked cool and bought it because the first paragraph grabbed me and wouldn’t let go. Cryptonomicon came out the same year the geeky consultancy company I was running was working on a gambling project on an offshore data haven; all our servers were named things like “kinakuta” and “raft”, and my laptop was “yt”. Hell, I even got hold of a copy of The Big U and read it. So, I’m not generally averse to the guy, and I have a fairly high tolerance for his diversions, random infodumps, and half-assed endings.

It was only when I got to the Baroque Cycle that I couldn’t handle it. At the time I was reading a lot of historical fiction and had pretty firm ideas on what constituted good writing in that genre. Quicksilver rubbed the wrong way against those genre conventions once too often, and since Stephenson was a relative newcomer to a period I already knew a bit about, his geeky fascination with things I considered commonplace (muskets and slow-match, for example) started to grate. Quicksilver was the first of his books that I didn’t finish, and I didn’t pick up another one until now.

Anathem is about a monk-like order who have survived thousands of years, who remain cloistered for up to a thousand years at a time, and who have a daily service of winding their giant clock, which has not just minute and hour hands, but year/decade/century/millenium hands too. It came out when I was working at Metaweb, on Freebase. The company had been named after Baroque-cycle-affiliated wiki of all knowledge, “The Metaweb” (now defunct, but you can see it on the Wayback Machine), and was founded by people closely associated with the Long Now Foundation, who are actually building a 10,000 year clock. Long Now talk was common at the office when I worked there, and there was lots of enthusiasm for Anathem when it came out — I remember there being an offer of tickets to a launch event or author talk or something for Metaweb staff — but I wasn’t in an SF-reading phase, so I skipped it. When Metaweb was acquired by Google, one of our founders gave a speech at our acquisition party talking about how Freebase was meant to be a repository of information that would last 10,000 years, and getting it into Google was the best possible way of furthering that goal. (True? Not sure.)

Enough background. A couple of weeks ago when I was standing in the Darebin Public Library’s Adult SFF section wondering what to read, I saw Anathem and grabbed it. I figured it would fill the time before my ebook reader arrived, I’d get to see what connections it had to Metaweb-the-company-where-I-worked, and it couldn’t hurt to have some of pop-cultural awareness of what it’s all about, the same as how I went to see Avengers, even though I don’t have much interest in the franchise, just so I’d know what people were talking about. All these were reasons to have a shot at it even though I knew there was a risk that I might find it as tedious and annoying as Quicksilver.

Surprise! It’s tedious and annoying! Stephenson finally found a way to add even more tangential infodumps into the story, by having almost the entire cast of characters be philosophers/theoretical scientists who spend most of their time lecturing or in Socratic-style dialogue about things like geometrical puzzles or the sensory perceptions of worms. Most of it ties in to the overall plot development, which at least is an improvement on some of his previous works.

The other thing that annoyed me was his worldbuilding: it’s set in another world where the people in it have “jeejahs” that are almost identical to our mobile phones and tablet devices; where the plebs wear baggy pants and sports jerseys with numbers on the back; where the dominant religion has a schism directly equivalent to the Reformation; and where details ranging from canvas-covered military transport vehicles to bucket-sized “sugar-water” drinks are all surprisingly familiar. The overall effect was of the kind of lazy worldbuilding where everything gets an “alien” name full of Zs and Qs and apostrophes, but is otherwise exactly the same as our world.

And then, over the course of hundreds and hundreds of pages, you eventually realise — SPOILER — that it’s all because parallel universes blah blah. Wait, you’ve been irritating me with your sloppy worldbuilding and “jeejahs” for all this time just so you could go SURPRISE! ALTERNATE EARTH!? And I’m meant to go “oh, wow, you’re not sloppy, you’re actually BRILLIANT!” Sorry, not feeling it.

So, I’m seven-hundred-something pages in to the book, and about ready to throw it across the room. And yet I find myself thinking, “Well, I’ve come this far, I may as well finish. Maybe it’ll get better.” At the same time, I have books ready and loaded on my ebook reader that I could be reading now, and probably enjoying more.

So the questions I’ve been asking myself, and which I ask you, if you care to take a shot at them: Firstly, with about two hundred pages to go, should I finish Anathem? Secondly, should I stay in school? If your answers differ, then why?

12 thoughts on “Sunk Costs Fallacy

  • Mary

    Not that this is directly helpful, but I invoked the sunk cost fallacy 5 days before I submitted my PhD thesis, and then reasoned that if I’d been teleported into the world with 5 days left to go on a PhD thesis and the right knowledge transferred into my brain, I probably would finish it because 5 days is cheap for a (hopefully) PhD.

    I guess that’s the flip side of the sunk cost fallacy: if you ignore investment up until the point at which you are making the decision, you sometimes get a rather cheap payoff. You (you-now) get to have finished Anathem for the cost of reading 200 pages, or a two semester certificate for the cost of one semester. The sunk cost fallacy really bites when you aren’t quite sure how much effort there is left (eg, in the “middle” of a PhD — or really most large projects — when people often have 1–2 years longer to go than they believe they do). That’s not the case here.

    I guess my answers are: give up on Anathem. I haven’t read it, but the last 200 pages of Stephenson novels are rarely my favourite parts anyway. (Which reminds me, I need to finish my writeup of Reamde sometime soon.) The potential payoff sounds rather low (based on this entry), only the potential to retrospectively understand a couple more inside jokes at Metaweb.

    The school thing is more complicated because I don’t know the potential payoff: you’re still by the sound of it planning to work in the industry, so does the industry care about this qualification? To compare, if you were a semester off a computer science Bachelors and intending to go into programming for the first time I’d say finish it if at all possible (even by, eg, two or three part-time semesters), because enough employers want to see it that it’s worth it in the general case. If it has no impact on your job prospects as it sounds like you suspect, I guess I’d also lean to dumping it.

  • Mark Smith

    Book — don’t really care. I finished it but wasn’t overwhelmed, and you won’t really gain anything (much of anything anyway) by having finished it.

    School — the sunk cost fallacy is true, but I’m not sure it applies in this case. It is only a sunk cost if the ultimate value is nothing (for you). If you get this piece of paper, then will that be worth something? A job, satisfaction of completion, better pay, something?

    If no — then sunk cost applies. Continuing to do something that isn’t making you happy is not a good call. But if yes, if that paper is worth something, then the time you put into it is inherently worth something. (But it may not be a lot.)

  • Vera

    Book: turf that sucka.
    Course: finish. Paper has some value to show your base level of knowledge to prospective gigs. Also might effect award rates?

  • Skud Post author

    Mary and Mark: it’s probably worth knowing that the qualification has very little value in the industry — about as much value as, say, a sub-university-level certificate in IT from a vocational school does. Which is to say, all it certifies is that you’ve been exposed to some stuff, but employers will actually want to know about your experience and clue rather than just see the certificate. It is *not* comparable to a Bachelor’s degree in terms of opening doors. (And note that I don’t have a Bachelor’s degree — actually dropped out in my final year — and think I’ve done okay so far.)

    The reason I’m doing this is not for the qualification, but for a way to spend a couple of years learning stuff in a somewhat structured way. What I lose by not doing it is a) the structure, and b) the government subsidy that I get as a student, which runs to a few hundred bucks a month. I can get a similar subsidy if un- or under-employed for the cost more paperwork and bullshit. Or of course I can find actual paying work (and it’s worth noting that I already have at least one offer of paid sound-industry work, but it would be hard to fit around school.)

  • GC

    Quit both now. There are not enough minutes in your life to read all the books worth reading, which means you don’t have time for the shit ones.

    Your course means virtually nothing in the field you’re going in. Hands on experience and knowing the right people is what you need now. Unless you’re going to work for a large corporate events management company or something.

    Sorry I am typing from my iPhone so this message probably isn’t delivered with as much sugar as I usually prefer.

  • Nathan Myers

    I suggest finishing both, but for different reasons.

    First, Anathem actually only gets really interesting in the last couple hundred pages. What came before I found very irritating — and I remember remarking frequently that he was really drawing down his trust account — but it turned out to seem necessary. The real point of the book isn’t the world-building or didacticism, but a meditation on what the quantum nature of reality could come to mean in the daily life of people who had internalized it, from the point of view of someone almost ready to.

    Second, being paid to go to school is great. But why actually go, if you get paid either way? Get the best of both worlds by attending class only for tests. If that’s not possible, bag it: the opportunity costs make it unaffordable.

    About REAMDE: If you treat it as a serious novel, it will be almost as irritating as the first part of Anathem. But don’t. It’s a romp, a send-up of every action / suspense trope ever, all shoehorned into one story that actually hangs together despite everything he seems to do to try to make it fall apart. (But a hint: skip every paragraph about simulated geology. It goes nowhere.)

    I read the whole Baroque Cycle, and it got better and better, but Quicksilver could have benefitted from stricter editing. I had to drop Mongoliad. It seemed to be trying too hard to be the basis for an RPG. I have doubts, though.

  • Skud Post author

    Nathan: you raise an interesting point about “why actually go”. I have actually arranged for one of my prospective subjects (“Mix Live Audio”) to do just that: the teacher said I can just work some gigs throughout the semester, and he’ll give me a practice exam and handouts so I know what’s in the end-of-semester exam. The other subjects, though, are group project based. There *may* be a possibility that I could do the projects independently (not being in a group, not using school facilities, rarely showing up on campus) and I’m looking into that, but the logistics may not work out.

    *If* I can arrange that, then those subjects become a matter of “work on my own stuff, make sure at least some of the stuff I do matches the criteria of the class projects, and do the necessary paperwork that goes with the class projects.” That’s definitely not as bad/timewasting as actually having to attend classes, so it shifts the balance a lot. However, that’s still a pretty big “if” as I have a lot of stuff to sort out before I would know I could do that, and my best chance for studio facilities won’t be available til some time after semester 2 has started.

  • Big

    First – no to reading the book. If it’s not enjoyable there’s _zero_ real payoff in finishing it (unless perhaps you value being able to discuss it later from the point of view of someone who read it all cf someone who read 80% of it).

    The course is different. As you say, in “the industry”, it’s all about what you’ve done – wh you’ve worked with – that matters, not some vocational training certificate. Having said that, consider a few things the course _might_ provide that dropping out of the course might mean you don’t even get tafe-course-level exposure to. If all you want to do is live sound at pub gigs, it might not matter, but exposure to recording/radio/exhibition/conference/studio/cinema/film/whatever-else-the-course-might-offer could be useful and much harder to get if you’re “just some pubrock sound engineer”. The other possible reason to staying is industry networking – you can work out which of your classmates has any talent at all – and befriend and stay in touch with the few you’d like to work with in future (and make sure that any of _them_ looking for in-class-talent know you’ve got it).

    In response t your comment in the replys that you’ve been offered real industry work but couldn’t fit it around class schedule – is it worth letting the school know and seeing if they’ll give you credit for actual work, or work out some way for you to fit class around the job?


  • Lindsey Kuper

    Y’know, I’ve read a lot of Stephenson and really enjoyed it, but man, the a few weeks ago I picked up Reamde and read the back cover blurb, and all I could think was Haven’t I already read this? I didn’t think I could stomach yet another long-winded book about an International Geeky Man of Mystery. I picked up something by Octavia Butler instead — the first time I’d read anything of hers! — and I am so, so glad I did.

  • Mary

    Right, the Bachelors thing is just an area where I’ve had this style of discussion before (and I wouldn’t blanket recommend finishing one, but I’d probably recommend it more strongly in the general case).

    Anyway, the case for staying in TAFE doesn’t seem awfully compelling to me when there are active cons. The “six more months to the piece of paper” argument is more compelling when:
    (a) the person has no other obviously competitive path into the career
    (b) the person had some reasonable investment in the course for its own sake

    That doesn’t really seem to scream “stay”.

  • David Gerard

    TAFE course: dunno, but if there’s some way to extract sufficient value from being there another semester …

    Anathem: put it this way, you’ve just saved me ever bothering.

Comments are closed.