in Economics

Movie Mind Games: Does manipulating our expectations make movies better?
Get free updates of new posts here

Illustration by Sean Nyffeler of Popcorn Noises fame

[This was originally a three-part series: Part 1; Part 2; Part 3.]

We spend a lot of time gobbling up media. We want to do fun stuff, and we want to do stuff on the cheap. Such is life in a stagnant economy. One of my go-to, quick and dirty ways to choose one option over the others is to figure out the cost per hour for each of my options, and then choose the one with the lowest cost per hour. 1

Here’s a quick summary of the cost to consume different types of media, shown in ascending worst-case dollars per hour 2 3:

  • Podcast – free – As long as I have iTunes and an internet connection, I can get just about any podcast free of charge.
  • News online – free – Yeah, the NYT has a pay wall now, but they don’t have any news I can’t get for free somewhere else.
  • Video games – $.03 to $1.25 – Angry Birds, Madden, NCAA: they all take so long they end up being really cheap by the time we’re through with them.
  • Books – $.25 to $2 – This one obviously depends how fast you read, but a good old paperback can go a long way on short change.
  • MP3 albums – $.3 to $4 – The trick with MP3s is to find them on sale. The Amazon MP3 store runs sales all the time.
  • Movies – $.50 to $5 – Movies tend to run the gamut because there are so many ways to get them. Prices vary pretty widely from Redbox to IMAX.

Almost any way I slice it, movies are one of the most expensive pieces of the entertainment pie. Looking back at my personal habits over time 4, it’s pretty obvious that I’ve been moving to cheaper and cheaper options over time. This wasn’t a conscious decision, but I have been purposely reducing my spending over the past few years, and I’ve obviously accomplished that by buying cheaper media.

Consuming media isn’t just about being entertained as cheaply as possible 5; I want quality entertainment. It’s not as simple as just consuming some type of media–I also have to figure out which examples of a given type of media to choose. If I’m listening to podcasts, how do I decide which ones? How do I find good books to read? How do I decide which movies to see in the theatre and which ones to rent? How do I know which ones to avoid altogether? The easy answer is recommendations. The trickier answer is expectations.

Recommendations

Over the past decade, recommendations 6 have gone from an informal give and take to a very sophisticated marketing tool, employed by giant companies to boost sales. Amazon, Netflix, Apple’s App Store and many other companies rely on recommendations to keep customers coming back for more. “Recommendation Engines” have become a closely guarded secret and a competitive advantage designed keep customers from switching to a competitor. I’ve bought hundreds of items on Amazon, and I like the recommendations it provides based on my previous purchases. If I start shopping at another online vendor, I’ll have to start over from scratch. That would be a lot of work, so I’m likely to stay with Amazon for quite a while unless a competitor offers something significantly better or Amazon totally drops the ball.

Many of my social interactions revolve around either sharing recommendations or comparing opinions on different media. For as long as I can remember, I’ve frequently asked friends what they’re into: “Seen any good movies lately?” or “Have you heard the new Girl Talk? How is it?” For almost any kind of media, I have at least one friend who’s practically on speed dial in case I need new recommendations.

I also make a lot of recommendations. I love it when a friend tweets, “Looking for some good books to read this summer. Any suggestions?” It takes me a few questions to figure out what kind of stuff they like, but once I zero in on their preferences I can usually recommend several titles that I can almost guarantee they’ll like. The same goes for music, movies, podcasts and TV shows. Part of being a maven 7 is that I’ve always got a solid cache of information ready to share if someone’s careless enough to open the door for me.

Expectations

The flip-side to recommendations is the expectations they create. If a friend of mine, let’s call him Morris, has successfully recommended 10 documentaries to me without any stinkers, then I expect his next doc recommendation to be a good one. If another friend, let’s call him Les, has recommended five documentaries for me, and all of them have been terrible, then I expect his next recommendation to be terrible and I’ll eventually just stop listening to his recommendations altogether. If Morris and Les both make recommendations to me at the same time, I can safely choose Morris’ recommendations because I expect them to be better. With each recommendation Morris and Les make, I can reevaluate their recommendations as a whole to determine how much weight I’ll give to either recommender in the future.

This is also true for recommendation engines like those at Amazon and Netflix. If Amazon starts recommending stuff that I hate, I’ll take that into account in the future and begin lowering my expectations for the stuff they recommend. Eventually I’ll just stop buying stuff they recommend, and that may remove the exit barrier I described earlier so that I’m comfortable going to another company and starting over from scratch.

There’s a feedback loop of recommendations and expectations. With each new good recommendation I get from a friend, the higher my future expectations will be that the stuff he recommends is worth my time and money. With each bad recommendation I get from a friend, the lower my future expectations will be that the stuff he recommends is worth my time and money. Eventually, I will learn to anticipate exactly how accurate my friends’ recommendations will be.

Recommendations and expectations are part of an adaptive framework wherein each future recommendation carries the weight of all previous recommendations. This feedback loop is only useful if I compare my actual experience to my actual expectations. 8

Utility-Hours Per Dollar

Before I can compare outcomes to expectations, I need a way to objectively measure my general satisfaction with any particular piece of media. Dollars per hour is a good metric to figure out the cost of consuming media, especially if my biggest concern is keeping a budget. It helps me measure efficiency. I might say, “Well, I’ve got three bucks left in my entertainment budget this month. I might as well stretch it as far as I can. What’re my options that are three bucks or cheaper and provide the most entertainment time?” But I’m not just looking for any old media–I want the good stuff. I need a way to account for both efficiency and the relative enjoyment offered by something. Enter this new thing I’m creating called “Utility-Hours per Dollar” (UHD) 9. The UHD allows me to normalize things so that I can compare apples to apples. Yes, going to see a movie in the theatre is really expensive ($5 per hour), but what if it’s the most fun thing I could possibly do with five bucks? That has to count for something, right? Sure it does.

I calculate UHD like this:

  1. Find the absolute cost (in dollars) of the media I’m looking to buy.
  2. Estimate how long (in hours) it will take to consume. 10
  3. Subjectively determine its utility 11on a 10-point scale (1 is for awful stuff, 10 is for incredible stuff).
  4. Multiply the utility number by the number of hours.
  5. Divide that number by the cost, rounded to the next highest dollar. For free stuff, use $1 (not $0). 12 13

For those who like a tidy formula, here it is:

  • UHD = (Utility * Hours) / Dollars

That’s it. Here are a couple examples 14:

  • A really bad movie at the theatre would be $10, last 2 hours and provide a utility of 2:
  • 2 utils * 2 hours = 4 util-hours
  • 4 util-hours / $10 = .4 UHD

  • A pretty good album that I buy on Amazon for $8 might give me 20 solid hours of listening at 6 utils:
  • 6 utils * 20 hours = 120 util-hours
  • 120 util-hours / $8 = 15 UHD

A UHD near zero sucks. A UHD that ends up in the double digits is pretty good. Stuff with a UHD in the mid-to-high double digits is pretty great. Using this metric, I can figure out my most cost effective, enjoyable option for entertainment.

Our trusty UHD chart–we’ll see this again later

UHD isn’t as esoteric as it seems

I realize that, at first, UHD just seems like a wonky way to describe something that’s already obvious and intuitive. But it actually has real-world applications, especially when it comes to understanding our intuitive-but-not-easily-explained preferences for stuff.

For example, UHD helps me understand why it took me a little while to move from CDs to downloading MP3s 15. Initially, the cost of an MP3 album (on iTunes, for example) was pretty close to the CD and Apple was using DRM 16. My concern was that I wouldn’t “own” the music if I paid for the MP3s. The result was that the utility of the MP3s was less than that of the CD, even though it was the same music at the same cost. Since the cost was similar, and the hours of entertainment would be the same, the difference in utility made the UHD for CDs higher than MP3s. Eventually, Amazon started offering DRM-free downloads and cheaper prices, shifting the UHD for MP3 downloads ahead of CDs. That’s when I made the switch to MP3 downloads 17. Of course, I didn’t actually do a conscious UHD calculation one day and say, “Ah ha! The UHD for MP3s is finally greater than it is for CDs! Time to make the switch!” But that’s basically what happened. The same process is happening for me with eBooks right now. 18

A brief, anecdotal history of cinema

The shift from CDs to MP3s, or from physical books to eBooks is interesting to me. But what’s really interesting to me is the persistence of movie theaters despite cheaper, very similar movie-watching options. Fifty years ago, the only real option for seeing a movie was to go to the movie theatre. This was great for movie companies because they could charge high prices since they were basically the only game in town. The UHD calculation wasn’t really useful for deciding how to watch a movie because it wasn’t so much a matter of comparing different movie-viewing options as just deciding whether it was worth it to spend the money on a movie or not. If it wasn’t, you just had to find something else to do.

Then technology started changing, opening the door for the home theatre experience. First, VHS started enabling people to watch movies at home en masse. Hi-fi began morphing into fancier surround sound setups whose cost was dropping so that more and more people could buy them. LaserDisc 19 came and went. Then DVD took hold and made the home-viewing experience even better.

A sidebar into UHD for movies at the turn of the century

Ten years ago, we really had two options for watching a movie (without owning it). We could either go to the theatre or rent it at Blockbuster. Let’s run through the UHD calculations real quick, just to get an idea of the difference in UHD for these two options at that time:

  • A good movie as a “New Release” rental was about $4 (-ish), lasted 2 hours and provided a utility of 6:
  • 6 utils * 2 hours = 12 util-hours
  • 12 util-hours / $4 = 3 UHD

  • The same good movie in the theatre would have been about $5, lasted 2 hours and provided a utility of about 7 (slightly higher since it was in the theatre):
  • 7 utils * 2 hours = 14 util-hours
  • 14 util-hours / $5 = 2.8 UHD

So the UHD for renting versus going to the theatre was really close even as recently as 2000. They were close enough that there was a real decision to be made: Spend $5 and go to the theatre or spend $4 and stay home? We would often decide what to do based on the number of people in a group (if there were four of us, we could just split the rental for a buck a piece; if there were two of us, then why not just pay for the movie in the theatre?) and our willingness to sneak snacks into the theatre. 20

Our trusty UHD chart from earlier

Snap back to reality

A lot has happened over the past 10 years or so. Netflix popped up, HD-DVD lost the war to Blu-Ray, streaming video became better and better, Blockbuster got crushed, and DVD rentals have gotten cheaper and cheaper. There are options now, options that just weren’t available when movie theaters first became a big deal. Not only are there options, but there are cheap options that rival the actual movie-going experience. And yet, movie ticket prices have been steadily increasing over time. 21

Let’s look at one more sample UHD calculation:

  • A really bad movie that I waited to watch on Redbox DVD would be $1, last 2 hours and provide a utility of 2 22:
  • 2 utils * 2 hours = 4 util-hours
  • 4 util-hours / $1 = 4 UHD

As we saw earlier, watching the bad movie in the theatre gives .4 (that’s point-four) UHD. Watching the same bad movie on DVD gives 4 UHD. Watching the bad movie on DVD is 10 times “better” than watching it in the theatre, and all of this difference is accounted for by the difference in cost. “But wait!”, you say, “What if I enjoy watching movies more in the theatre?! I really like going to the theatre!” Ok, fine. How much better would the movie have to be in the theatre to make up for the difference in UHD?

Some people will want to go to the most extreme case first, so let’s just go straight there. Let’s say that the bad movie moves from 4 utils to 10 utils just because I enjoy going to the theatre so much. It only jumps to 2 UHD (still half of the 4 UHD if I wait to watch the bad movie on Redbox DVD). “That doesn’t make any sense!” My counter would be, “So you’re saying there’s no way any movie can be better than the bad movie in the theatre? What if you go see a good movie in the theater?” Since the 1-10 scale is a subjective scale, I have to leave room above the bad movie for less-bad movies. Either that or I have to slide my Redbox DVD experience down to a 1 or something. If I move the theatre experience up to a 10 and move the Redbox DVD experience down to a 1, then I get the same result for both options: 2 UHD.

The present, seemingly uncrossable gulf between UHD for going to the movies and watching them at home is due to super high, sticky movie prices and much, much cheaper alternatives for watching movies at home. This has created such a big gap in cost that watching movies in the theatre is just that much more expensive, ruining their UHD relative to the very-similar experience of watching movies at home now.

Going to see movies in the theatre is expensive. I realize some people will say, “But your formula is just wrong. It weights the cost too much.” Using the UHD calculation as-is, it’s hard to see many situations where it would be better to go to the theatre than to watch the flick at home. It’s possible that I’m weighting cost too heavily, but I think the real problem is that movie theaters are just way too expensive now because we have more, better options. There really is that much of a difference in cost between movie theaters and rentals.

Movie Expectations – A bizarre special case?

And yet, new releases continue to set box office records as people go to the theatre in droves. At the same time, the movie theaters are much more expensive than the alternatives, and the quality of the movies released has been consistent (or at least not improving enough to justify the growing gap between theatre prices and the alternatives). What gives? If I’m right that waiting for the movie on DVD is almost always better than going to the theatre, then why do so many people continue going to see movies in the theatre? Why did I go see three movies in the theatre this summer?

I’ve overheard this sentiment several times recently: “That movie wasn’t that bad. I just went into it with no expectations and it turned out ok. I’ve decided I just won’t have expectations for movies because I end up over-hyping them and when they don’t meet my expectations I feel ripped off.” So the idea is that movies are often bad because we expect them to be good, or at least because we expect them to be better than they actually are. To solve this problem, we play mind games with ourselves, intentionally under-hyping a movie so that when we go see it and it’s just an ok movie, it exceeds our deflated expectations.

At first, I saw the wisdom in this tactic. If we get really good at lowering our expectations, almost any movie will be a success, at least inasmuch as it will exceed our expectations. That way, we can pretty much guarantee that when we pony up $10 for a ticket and another $10 for concessions, we won’t be let down.

The more I think about this, the more ridiculous the idea seems. We don’t lower our expectations for music, books or TV shows, do we? So why do we do that with movies? Of all our options, movies are one of the most expensive options we have. Why would we trick ourselves into doing something super expensive that we don’t really enjoy that much?

But this one goes to 11.

What we’re doing when we go see movies with deflated expectations is trying to trick ourselves into accepting the lower utility of the movie and ignoring the greater cost. Let’s say we go see the same bad movie we’ve talked about already, but we have “no expectations”, meaning we essentially expect the movie to be about 1 util of entertainment. That way, when we go to the movie and it’s 4 utils, we exceeded expectations! We practically created three utils out of thin air! Um, ok. But the problem is that the movie itself is still only 4 utils. It has to be because we have to put every other movie we’ve ever seen on the same scale. We can’t artificially inflate the number of utils to, say, 5 because what happens to those other movies that really were a 5? 23 This also reduces the perceived value of a 10 because we’re watering all of our other movie experiences down. So if we inflate the utility from a 4 to a 5, what we’re really doing is inflating the whole scale. Now it goes to 11. We have created UHD inflation.

A Delightful Food-Poisoning Analogy

It’s as though we have a friend who likes to cook. He says, “Hey everyone! Come over to my place and bring $5. I’ll cook something for all of us to eat. It’ll be delicious!” So we all go over there and bring five bucks. We eat the meal and it’s really freaking terrible. Half of us are disappointed and the other half are left with Oregon Trail flashbacks. A few weeks later, the same friend makes the same offer. We all decide to give it one more shot–maybe he just had a bad night, right?–and we head back over there with our five bucks. Same thing happens. Half of us are disappointed and the other half end up battling the dysentery. A couple weeks later, the same friend makes the same offer again. Would I go? Of course not. But what if I said, “You know what? I’m going back! I’ve learned that I just have to lower my expectations so I can really enjoy the food poisoning! I’m going to just assume it will kill me this time, or at least that it’ll ruin my digestive system for the next few days. It’s going to be terrible! I can’t wait!” I’m tricking myself into paying $5 for the privilege of being food-poisoned by my friend.

Ridiculous, right? How’s that any different than saying, “The trick to movies is that I just lower my expectations as much as possible. If I go in without expectations, I can’t be disappointed!” Well, kind of. But you’ll trick yourself out of $10 and you’ll waste time that you could’ve spent doing something cheaper and better.

What’s going on here?

So why do we trick ourselves into going to see bad movies in the theatre at 10 times the cost of the Redbox DVD? What’s really strange is we don’t do this with other stuff. If anything we tend to inflate our expectations, especially when it comes to music. I’ve heard friends totally pan a new album just because it didn’t live up to their expectations, which were far higher than they should’ve been. This happened with She & Him’s “Volume 2”. Many of my friends said they didn’t like it, and it was just more of the same from She & Him. Well, duh. They’re still She & Him and their first album was awesome. I’d say “Volume 1” was like 8 utils to me. I expected “Volume 2” to be about 8 utils. I think some of my friends expected it to somehow magically be 10 utils. Why? I have no idea. Turns out it was right about 8 utils (maybe slightly less, but it was really close). I ended up enjoying it a lot (and still listen to it regularly), whereas they ended up all disappointed and annoyed. Of course, they just did that to themselves.

So, with music, a lot of my friends do the opposite of this movie theatre trick–they inflate their expectations so they’re artificially disappointed when they hear a new album. 24 This is silly, but at least it makes some kind of sense: If we’re going to mess with expectations, we should manipulate the numbers so we tend to be less satisfied. That way, we’re basically tricking our future selves into spending less money. But this leaves us in an awkward place: we trick our future selves into spending more money on expensive stuff (movies at the theatre), and less money on cheap stuff (music). If we’re going to trick ourselves, we should be tricking ourselves so that we’re less satisfied with expensive stuff and more satisfied with cheaper stuff. At least that way we end up tricking our future selves avoiding the more expensive purchases on stuff we don’t like anyway. The UHD for music is almost always higher than it is for movies so, if anything, we should tend to “trick” ourselves into consuming more music and fewer movies in order to use our time and money more efficiently by consuming better stuff.

Why do we have it backwards? There could be a few explanations for this. Rather than trying to make sure we spend our money as efficiently as possible, we’re more focused on justifying expensive movies because they’re a cultural staple:

“Did you see the new X-Men movie?”
“No, I’m waiting for it on DVD.”
“Lame-o! Hey everybody! This guy’s super lame!”

It’s not cool to wait to watch movies on DVD. If I show up at the water cooler on Monday morning after a big release 25, nobody wants to hear anything I say if it starts with, “Well, I’m waiting for that one to be available on Netflix streaming.”

Sometimes there are benefits to seeing a movie in the theatre, especially action flicks. But we can account for that by bumping the utils for the movie just a little bit. Maybe X-Men on DVD is 6 utils, but X-Men in the theatre is 8. Ok, but does that justify the 10-times-higher price tag? The UHD sure doesn’t think so.

Conspiracy Theory!

There’s also some clever marketing going on by the movie studios. Most of the movies made are crap. We tend to remember good things more than bad, but most movies are really, really bad. This is easily confirmed by just browsing Netflix for movies 26. Some friends and I recently spent about two hours scrolling and scanning through Netflix to find a movie to watch. We ended up just re-watching Arrested Development Season 1. The issue wasn’t a lack of options, it was a lack of good options. There just weren’t any. We looked at hundreds of movies and all of them were terrible. If we used movie studios’ previous product as an indicator of future quality, we’d almost always opt not to go see movies in the theatre because there’s just too much risk that the movie will be crap and we’ll waste $10. Waiting for the DVD gives us more time to get recommendations from other people who have seen it so we can decide whether it’s even worth seeing on DVD.

But movie studios have cleverly convinced us that we should set our expectations aside so we can enjoy the movie-going experience itself. 27 Of course, as discussed earlier, the movie theatre experience isn’t really much different than just watching the movie at home (there are a few exceptions). If we all had the appropriate level of expectations for movies, we would rarely go to the theatre because we would mostly be disappointed. But instead of just saving that cash and doing something with higher UHD, we buy into this idea that we should essentially forget all the crap they’ve fed us previously so that we can enjoy this current experience more.

Positive Reinforcement of Terrible Moviemaking

But it’s actually worse than that. By tricking ourselves into “liking” (and paying for) bad movies, we’re encouraging movie studios to make more bad movies. If I trick myself into paying $10 to see Zookeeper, then I just gave the movie studios $10 and a green light to start production on Zookeeper 2: Flying Poo. Now I have to trick myself into going to see that dud, too?

Not only am I tricking myself into spending a lot of money on a bad movie, but I’m feeding the movie making machine so that it continues to churn out garbage that I have to trick myself into overpaying for in the theatre next time. I’m essentially a recommendation engine whose recommendations are made in dollars. I’m saying, “Movie studios. I recommend that you make Zookeeper 2: Flying Poo! Here’s ten bucks to get you started!” When does it end?

It is ending

In many ways, this phenomenon is ending, if only subtly. There are myriad modern sources of recommendations that are almost forcibly removing our self-imposed myopia. For example, Rotten Tomatoes is a crowd-sourced recommendation engine that is almost impossible to ignore once you know it’s out there.

“Want to go see the new Zookeeper movie?”
“What’d it get on Rotten Tomatoes?”
“Eleven percent.”
“And how did the first one do on Rotten Tomatoes?”
Fourteen percent.”
“This one is worse than the first one? That’s pretty bad. I’ll pass.”

That’s what saving $10 and two hours sounds like. I have conversations like this one a couple times a month. On the flip-side, I often decide to go see a new movie specifically because Rotten Tomatoes rates it highly. For example, when I was in Vancouver last year, my friends and I went to see Drive because I saw it got something like 90% on Rotten Tomatoes. They had never heard of the movie, and I had only heard a little about it, but the Rotten Tomatoes score pushed me to recommend it to them. We also went to see The Help and Moneyball because of high scores on Rotten Tomatoes. All three movies ended up getting Oscar nominations 28 There were a few other movies that we skipped because the Rotten Tomatoes score was so low.

Recommendations Engines are enabling us to make better decisions and making it more difficult to declare that we’re lowering our expectations to justify a trip to the theatre. It’s a lot easier to lower expectations when there’s some chance that the movie will be good despite the trailer or word-of-mouth buzz it’s been getting. But when thousands of people have already said it’s a bad movie and we know that, then it’s much harder to pretend it might be good.

The trick is that we, as consumers, have to listen to what so many other people are telling us through all the recommendations vehicles that are out there. When we start listening to others’ recommendations, we can set our expectations appropriately to maximize two of our scarcest resources: time and money.

SPECIAL THANKS

I put a lot of work into this piece, but I also got a lot of help from other people. Jason Killingsworth offered his editorial insight and sage advice to help make it better and more readable. Jason was also one of the first bloggers I read, so there’s a nice historical symmetry here. Danny Anderson did the final review before I published, and helped me figure out how to wrap it all up. Sean Nyffeler illustrated the piece (twice, actually: he did a draft, took some notes and re-did the illustration for the published version). Several other people were sounding boards who helped me refine the basic ideas over the past several months. Thanks to everyone who helped make this a better piece.

  1. I see you rolling your eyes. Don’t judge me. You do this too.
  2. I tend to break down costs into dollars per hour of entertainment (I also do this with stuff like shoes and clothes, only in dollars per day {a}). There’s a lot of free entertainment–podcasts, most news, blogs–for $0 an hour. Books are a pretty good value: I paid $12 for my copy of Infinite Jest, and probably got 40 or 50 hours of entertainment at like $.25 an hour (it helps that I’m a slow reader). To be fair, I should include the normal case where I’d pay $12 for a paperback and get six hours of entertainment at $2 an hour. MP3 albums are a decent value: if I pay $8 for an album on Amazon and it’s a good album, I might listen to it 20 or 30 times at about $.30 an hour. If the album’s no good (which rarely happens thanks to previews and friends’ recommendations), I could end up paying $4 an hour if I only listen to it twice. Video games used to be a pretty decent value: $50 for a console game that would yield 30 to 50 hours is somewhere around $1.25 an hour. Of course nowadays I can get Angry Birds for $.99, giving me 35 hours (and counting) of entertainment at about $.03 an hour {b}. Redbox DVDs are a really good value: $1 for two hours of entertainment, or $.50 an hour. Watching a movie in the theatre is about the worst possible dollar-per-hour value: $10 for two hours of entertainment works out to five bucks an hour, and that’s assuming I don’t buy concessions.

    {a} If you wonder why I’m wearing the same old shoes I’ve been wearing every day for a couple years, it’s because I’m trying to get down to a nickel a day on those suckers.

    {b} This isn’t where I’m going here, but I’m impressed at how consistently innovation and disruptive technologies make things cheaper and cheaper. Nintendo is feeling a lot of pressure from smartphones, especially in the portable gaming business. Why pay $249 for a Nintendo 3DS, plus $30 or so for each game when I can just pay $230 for an iPod touch and get excellent games for a buck? Never mind that iPod touch isn’t just a portable game system.

  3. Yeah, yeah, yeah. I know there are costs associated with these things (like in order to browse online news, I have to have a laptop or something), but I’m just talking about the cost of the actual media to keep things simple.
  4. I go through phases where some things are more interesting than others. This year, I’ve listened to podcasts a few hours a day, and I’ve been reading more and more blogs. The last couple years, I listened to ridiculous amounts of music. Before that, I went through a phase where I was watching movies like it was my job. I’m always reading at least three books.
  5. At least not under normal circumstances. There are times when simply passing time as cheaply as possible is the most important factor. While I’m unemployed, I skew toward the cheaper stuff with less concern about how good that stuff is. Of course, I’d like to get good stuff cheap, but since I can’t afford to pay more, I end up going cheap and hoping it also happens to be good. Normally, I’d look for a combination of good and cheap, allowing myself to spend a little more to get more “good” for less “cheap”.
  6. Recommendations are really a signal (the term is used a lot in economics and marketing, but I first heard it when I was boning up on my game theory). A signal is just a way for one party to convey information about a thing to another party. Luxury car makers signal quality by showing rich-looking older dudes riding around in really sleek cars in isolated settings, implying (or explicitly stating) that the car is of a quality consummate with rich people who want to be comfortable and feel separate from (better than?) the rest of society. Lite beer commercials signal fun or carefree-ness.

    So recommendations qua signals have been around forever. What’s unique about recommendations in this context is that they’re essentially third-party signals, not signals being sent directly from the provider of the thing. Amazon recommends stuff to me based on what other people like me have bought {a}, so the manufacturer isn’t sending me signals, Amazon is sending me signals based on other Amazon shoppers’ experiences. This form of signaling isn’t per se new, but its ubiquity is a new thing. {b}

    {a} There are actually a couple kinds of recommendations happening at Amazon. There are recommendations that Amazon makes directly (“Others who bought this book also bought these other books.”), but I also get implicit recommendations on specific items by looking at the aggregated user ratings for the item. “That coffee maker has an average of 4.5 stars, so I’ll get that one since I don’t know anything about any of them.” So Amazon signals based on other users’ purchase history. And Amazon users signal based on their own experiences with individual products. These two things combined are what I’m referring to as Amazon’s “recommendation engine”. {i}

    {i} Since you’re still following… This is an example of market efficiencies improving as more information becomes available. Ten years ago, if I needed a coffee maker, I would’ve gone to a brick and mortar store, stood in the “Kitchen Appliances” aisle and stared at the coffee makers for like 20 minutes before finally just grabbing a familiar name brand in my price range. I would be wondering, “Which of these is the best one? Where can I get the most bang for my buck?”  If I was really diligent, I could go get a copy of Consumer Reports and dig around for information about different brands and models and all that. But that would take a lot of effort that probably wouldn’t be justified for a $20 purchase. This is pretty inefficient since it’s a time-intensive process, I don’t actually know how good the coffee maker is, and I don’t know whether it’s priced appropriately. Now, I’d just log onto Amazon, search “Coffee Maker”, and probably buy the highest-rated one in my price range. The whole process would take about five minutes and I’d know I’m probably getting the most bang for my buck. That’s a much more efficient market, directing my “investment” to the item that is most likely to meet my needs.

    {b} In a meta twist to typical third-party signaling (like word of mouth), some companies are starting to simulate third-party signals in their ads. The most recent version of this I’ve seen is the Ford Focus commercials where they ambush “real customers” by telling them they’re going to be interviewed about their experience with the car, and then dump them in a mocked-up press conference that is recorded for TV. Manufacturers are realizing that third-party signals are valuable and have a lot of pull, so they’re trying to simulate that same type of signal in their own commercials. That way, the company isn’t telling us how awesome the product is, but “real customers” are telling us about it.

  7. Speaking of book recommendations, if you haven’t read Malcom Gladwell’s The Tipping Point, you probably should. To cut to the chase, a “maven” is someone who gathers and distills large amounts of information, and is also able to share that information in a way that people just “get it”. I don’t know much, but the stuff I do know is ready to be shared at the drop of a hat.
  8. If you know me well, you know that I usually give sort of pessimistic reviews of stuff. Last night, I watched a movie with friends. When it was over, they asked what I thought. “It was pretty entertaining. It started off a little cheesy, but it hit its stride in the middle. I enjoyed it.” This is pretty typical for me–I rarely give a glowing review of a movie. But this is a conscious decision to try to be as realistic as possible and to leave myself room to give good reviews to stuff that’s really good. I don’t want to be that guy that always says, “It was great! What a great movie!” What good is that for anyone else who’s deciding whether to see a movie? My input is almost totally useless if it’s always very positive. If anything, I bias my reviews so that they’re low, giving my good reviews more weight. I’m playing it safe so that if I say something’s good, I know others will perceive it as good. Of course, people who know me already know this, so they temper their expectations accordingly when accounting for my recommendations.
  9. Yes, I know this isn’t the most scientific (it may not be scientific at all, actually) way to do this, but it’s easy and allows for a quick and dirty way to compare stuff in a pinch. It’s also convenient for me because it helps set up the real point of this piece.
  10. I’m using this as a proxy for how much entertainment it will provide. This is another quick and dirty estimate since there’s no guarantee that a book that takes 10 hours to read will actually provide 10 hours of entertainment. I don’t want to count “amount of enjoyment” twice, so I’m ignoring the actual amount of enjoyment here so that I can account for it in the more subjective part of the UHD, the Utility.
  11. Ok, so I’m just sort of cavalierly assuming everyone is comfortable with the concept of “utility”. In a nutshell, utility is the amount of enjoyment we get from consuming one unit of something. The unit of measuring utility is the “util” (a fictitious, arbitrary unit), so my scale of 1-10 is really a scale of 1-10 utils, where 10 is maximum enjoyment. I’d say that eating an apple is about four utils for me, and eating ice cream is between six and 10 utils depending on what kind of ice cream and how long I’ve been eating it {a}. Here is the Wikipedia page for utility. {a} I’m alluding to the so-called “Law of Diminishing Returns” here. If I haven’t had ice cream in a week (a long time for me), and I finally get my fix, that first bite will be like nine utils. The second, third and fourth bites will be like 8.5 utils. Each successive bite will likely be slightly lower on the scale. If I eat enough ice cream in a sitting, I could eventually get to zero utils and then creep into negative territory, where each bite makes me feel worse and worse until I run out or puke. {i} Since you’re reading the footnotes, I might as well go on a sub-sub-tangent, right? So the question is, “When should I stop eating the ice cream?” An easy deciding point would be to keep eating until I’m out of the positive until territory–I would stop eating when each successive bite provides zero utils (or fewer) of satisfaction. But there are other factors to consider: Should I be doing something else right now? How much exercise will I have to do to get rid of all these calories? How expensive is this ice cream? The more ice cream I eat, the less satisfaction I get from each successive bite, but I also set myself up for sacrificing utils in other areas of my life. This can get really, really complicated, but I’ll just go with a really simple possibility. Let’s say the next bite of ice cream would be 4 utils, but the amount of exercise necessary to work off the next bite would be -5 utils (that’s negative utility meaning I really don’t like exercising just to burn calories). So, eating the next bite of ice cream is really a -1 util decision. If the only two factors in each bite of ice cream I eat are (1) the satisfaction of actually eating the ice cream and (2) the negative satisfaction I’ll get when I have to work it off, then I should stop eating ice cream when the combined utility of the next bite and its correlated exercise is zero or less.
  12. The UHD starts to break down when the cost is less than a dollar because we start dividing by decimals, inflating the UHD to pretty large numbers as we get close to zero. Rounding to the next highest dollar means we’re always working in relatively round numbers and we avoid dividing by zero. We’re trying to figure out how good a movie is, not putting a man on the moon here.  I’m also basically assuming that the stuff we’re talking about–movies, music, books, etc.–is somewhere between free and $10-15. Hence the utility scale of 1-10. If we start including stuff like concerts, theme parks and higher-cost activities, it might be better to use a 100-point utility scale. But this is supposed to be simple, right?
  13. This is way beyond the scope of this post, but I can’t help but wonder what’s on the other side of “zero cost” in this equation. What if the cost is negative (we’re getting paid to consume the media)? What if there’s negative utility (consuming the media actually makes me worse off)? Or what if we’re not even talking about media anymore, but just anything. Some form of this equation could be useful for evaluating job prospects while accounting for both pay and personal satisfaction (utility) in doing a job. It would take some tweaking, but I think it might be useful. Of course, that’s beyond the scope of this post.
  14. I’m always assuming we’re talking about one person watching a movie. If there are multiple people, then the additional value in renting the movie increases proportionately to the number of people involved. This is because multiple people can chip in to split a rental, but everyone pays full price if the group goes to the theatre. For every person added to the group, the movie theatre gets linearly more expensive according to the formula C*n where “C” is the cost of the movie and “n” is the number of people in the group {a}. While, for every person added to the group, the rental experience gets geometrically cheaper according to the formula C/n {b}. So the examples I’m using are the best case examples if we’re trying to find some reason to go to the theatre.

    {a} There may be some economies of scale vis-a-vis snacks with larger groups, though.

    {b} “n” is really the number of paying people in the group. There are always some jokers who say, “Yeah, I’ll kick in!”, but then they don’t.

  15. Ok, ok. It took me a while to purchase MP3s, but we’ve all been downloading MP3s since the Napster days (or earlier via AOL chatrooms and such, right?). The fact that I can still download all the free music I want, and that I don’t do that is a tangent I just don’t feel like taking right now. That said, I do think it’s pretty interesting that I (and a lot of other people) pay for music now when I used to (and still can) just get it all for free.
  16. DRM is basically a way for content providers to help protect their copyrighted work by dictating how many (and what kind of) devices the purchased stuff can be played on. It’s basically dead (or almost dead) for music, but is alive and well for video.
  17. The last CD I bought was Cool Kids’ “Bake Sale“.
  18. I’d say it’s more or less inevitable that we’ll all eventually be streaming MP3s (or maybe I should drop the “MP3” label and just say “music”) with some sort of subscription (think Spotify) instead of purchasing via download. The problem is almost the same as my problem with switching from CDs to MP3s, only switching from “buying” an MP3 to “streaming services” (essentially leasing access to a catalogue) is a bigger paradigm shift that people are very, very slowly embracing. This shift has already happened with streaming video. I can’t really think of any reason to buy a DVD anymore. It just seems like a waste when I can probably stream it online whenever I want. The irony is that we’re making the buy-to-stream shift faster with video than with music. I’m still working out why these two seem to be happening out of order.
  19. LOL.
  20. When I was a kid, my dad rented movies almost every weekend, but if we went to the theatre, we were absolutely sneaking in some Twizzlers and sodas {a}.

    {a} Probably Winn-Dixie brand stuff like Strawberry Check {i}.

    {i} I really wanted to include like a Wikipedia link or something with that Strawberry Check reference, but I couldn’t find anything on Google. Am I making that up?

  21. Movie ticket price increases have more or less tracked with inflation since the 60s {a}. Before that, they increased really quickly, which makes sense since it was a new technology where demand was increasing quickly and the supply (movie theaters) was growing slowly. Eventually, there would be enough movie theaters that anyone who wants to go see a movie can go see it. Then, barring any substantial changes in technology, the prices should flatten out. Eventually, prices would be expected to decrease if there were competitors (direct or indirect) that reduce demand. This is happening now with the proliferation of the home theatre, DVDs, Redbox, Netflix, etc. It makes sense that movie prices should plateau pretty soon because of all the competitors to “movie theatre movies” now.

    If you want more data, you can go here to see US Inflation numbers and use a calculator. You can see average ticket prices over time here. Because of really high inflation in the 70s, movie ticket prices actually lag by a few years, but they catch up to inflation in the 80s. There seems to be a larger-than-inflation increase in ticket prices in the 90s, but that doesn’t have a huge effect on the general trend that movie ticket prices track with inflation from the 60s forward.

    {a} At least that’s what my few hours of analysis tell me. As far as I can tell, there’s about a 95% correlation {i} between inflation and movie prices since the 60s. It’s entirely possible I screwed this up.

    {i} I used the two sources above and the Excel “CORREL” function to get this number.

  22. I’m using the same utility for theatre and Redbox, basically assuming there’s no difference between theatre and home theatre for most movies now – this isn’t IMAX. Also, this makes the next discussion a little easier.
  23. This could apply to anything where we artificially inflate our subjective evaluation of something on a bounded scale. University grade inflation is a good example of this. Ok, so everyone gets all As and Bs… how do we find the really good students now? And where did all the mediocre and bad students go? We’re just watering down the accomplishments of the really good students through illusory grade inflation, and we’re hurting ourselves because when it comes time to find a really good student, we can’t find them. They’re all buried in a pile of mediocre and bad students, who are all reveling in their fake good grades. Put differently, we’re taking what should be a sort of “normal” curve centered around “C” (which is supposed to mean “average”, right?) and shifting the center of the curve upward. {a}

    {a} Shifting the center upward isn’t the only issue. The main issue is that the center of the curve is being shifted upward to a fixed boundary {i} at “A”. So we end up with a glut of students with grades around A/B and nobody below them. That’s not helpful to students, professors, or potential employers.

    {i} I guess we can sort of cheat here since “A+” exists now. But how far can we go? “A++”? “AA”?

  24. This obviously seems a lot like a general hipster thing, but I’m already like 10,000 words into this sucker and I don’t feel like trying to explain how hipsters trick themselves into being ambivalent about everything they experience. “They silently overhype everything so they can publicly pan everything.” That seems like a sufficient explanation. {a}

    {a} I know I’m pretty much safe here because hipsters don’t care enough to read long articles, and they especially don’t care enough to read footnotes. Even if they did care that much, they definitely don’t care enough to bother responding in any way. They’re totally above that nonsense. Fight the establishment!, etc.

  25. This would be doubly awkward because I don’t have a job {a}, so I would be just showing up at a random water cooler in someone’s office, looking to talk about a new movie that I don’t plan to see until it drops on Netflix streaming. This seems like a great way to mess with people. Challenge accepted!

    {a} HA! Now you know how long ago I wrote this piece! Still, I’m not cutting this footnote. So there.

  26. Yeah, I guess it’s possible this could be a result of sampling bias, viz. Netflix’ streaming selection isn’t representative of the set of all movies as a whole.  But we weren’t looking for new movies, just movies. Since we were looking for any movie to watch (unconstrained by release date), we should have been browsing a set representative of all (or most) previous movies that have been released, including good movies. I guess it’s possible that somehow Netflix’ streaming selection is mostly bad movies (meaning that the good movies aren’t available there), but that seems unlikely.
  27. Eg, the Oscars, Golden Globes, AFI Top 100. These are all marketing as much as anything else. Much of the marketing is designed to just get us to watch movies, but much of it is focused on the “unique” experience of going to the theatre.
  28. Granted, Oscar nominations aren’t a guarantee that a movie is good, but they’re a pretty good indicator.

Write a Comment

Comment