Paradoxic Fandom

Q: What’s the difference between Star Wars fans and Star Trek fans?
A: Star Trek fans don’t hate 90% of their movies!

It’s funny because it’s true!

As a life-long Star Wars fan, this joke resonated with me: my position of finding a lot to enjoy in literally every Star Wars film now seems highly unusual. It also raises an interesting question. What does it mean to be a fan of something you are mostly angry about? How does that happen?

I’ve been thinking about this for a few years now.

Here’s what I’ve figured out.

That’s No Moon

This phenomenon is much bigger than Star Wars, if such a thing is possible.

As an area of work and personal curiosity, free-to-play mobile games show a particularly stark example of what I’m calling ‘Paradoxic Fandom’. Studying the games that have found greatest long-term financial success, I noticed that almost all their player communities tended to have the same repeating refrains:

  • Every update makes the game worse
  • The game is dying
  • The developers are out of touch with players
  • The developers only care about ‘x’ players (x is either new players, or the biggest spenders)

Where I work, it was particularly noticeable that we didn’t get much feedback like that until we came up with our first truly successful long-term game.

I’ve seen Paradoxic Fandom elsewhere too.

  • The dominant feedback on most social media platforms (Facebook, Twitter etc) is that all changes to that platform are for the worse… but people keep using them.
  • The ‘Bad webcomics wiki’ seems like the go-to place for fans of a webcomic to complain about how much they hate it.
  • In Future Shock (2014), a documentary about the long-running UK anthology comic 2000AD, one of the historic editors draws a distinction between ‘readers’ and ‘fans’; he implies that the fans were the most difficult to deal with.
  • I keep in touch with developments in LEGO through the blog From Bricks To Bothans; it became apparent that main writer Ace Kim (since 2002!) has a similar love-hate relationship with LEGO (sample: a review of the ‘Ultimate Collector Series’ LEGO Star Destroyer, ending with “Things like this re-affirms my decision to stop collecting LEGO”)

It’s not just statistical regression

There’s two phenomena that look a bit like what I’m talking about, but are meaningfully different: the Sophomore Slump, and its close relative the Sports Illustrated cover jinx.

The Sophomore Slump occurs if someone (or a group of people) perform worse when they have less to prove. Having found success on their initial effort, they may try less hard for the follow-up, be it students in their sophomore year, or bands on their second album.

The Sports Illustrated cover jinx is when an athlete performs exceptionally well, gets featured on the cover of the magazine, then has a disappointing performance immediately after. It’s possible that some athletes find the additional scrutiny difficult to deal with, but this seems much more likely to be a simple case of regression to the mean: in anything where there’s a fairly strong random component to performance, an outlier is most often followed by a more average result.

This can apply to almost anything. For example, I found out the excellent line “I’ve come here to chew bubblegum and kick ass… and I’m all out of bubblegum” was from the film They Live (1988), but when I eventually watched it, literally nothing in it was as good as that line. Regression to the mean!

But Paradoxic Fandom is an ongoing, sustained effect, so it’s not just a statistical regression/slump/jinx.

Paradoxic, not yet toxic

When consumers begin to harass or threaten people they perceive as damaging the thing they love, they have crossed a line into what gets termed ‘Toxic Fandom’. Here, I’m interested in the wider phenomenon, which can easily give rise to Toxic Fandom but is meaningfully distinct. So I’m calling it Paradoxic Fandom, and the key defining features are:

  • The product/service is ongoing over months, years or even decades
  • Fans continue to consume and engage with the product
  • … but they report finding the product/service consistently disappointing, getting worse over time
  • … and they believe this is because the creators are out of touch with the fans

So what is going on here?

I think this actually begs four related questions:

1) Why do companies do things badly?
Why can’t video games harness everything they learned and get better with every iteration? Why, when there is clearly money to be made in satisfying a large audience, does capitalism fail to deliver?

2) Why do consumers misjudge things?
If criticism is unwarranted or short-sighted, why does that happen?

3) Why do people remain fans/consumers of things they seemingly hate?

4) Even when responses are mixed, why does criticism dominate the discourse?

I’ve been thinking about this ever since the borderline-allergic reaction of some ‘fans’ to The Last Jedi (2017). Here’s what I think is behind each of those questions.

1. Why do companies do things badly

What about the money?
If something is a financial success, the budget for follow-ups is likely to be bigger, which in theory should help. But while money helps with execution, in artistic endeavours it’s very clear that money cannot buy ‘quality’ (whatever that is) – if it could, movie and game studios wouldn’t lose the most money on the big-budget failures, but that’s exactly what happens.

The second death star: bigger budget, worse results.

You can’t please all of the people all of the time
The follow-up thing will certainly have similarities to the original thing, but also differences; the people who loved the first thing will have liked different things about it, so inevitably some will be disappointed.

More money, more problems
Companies generally try to make more money – like LEGO looking to grow their business, or a free-to-play game looking to maximise profits. Much as capitalism is built on the lucky fact that in the right environment, competitive self-interest can produce great results for everyone, the end-result is always some kind of compromise between buyer and seller.

One simple factor is that a company will try to make more money out of their customers. Most benevolently this could be by making further content, but there are less benevolent ways too (most simply, releasing a DVD with two different slip-cases to try to get fans to buy two copies).

But that can only go so far. Generally, the best and most scalable way to make more money is to find more customers. It’s quite possible that having attracted all the customers you can with your existing product, you’ll need to make some changes to attract larger numbers, which brings us back to not being able to please all the people all of the time.

2. Why do consumers misjudge things?

I should make one thing very clear: when people complain about things, it is often worth listening, and very productive action can be taken as a result. But sometimes, as consumers, we do misjudge things, and that criticism is less useful. Why does that happen?

One thing I’ve found is that making things – really, almost anything – is always more difficult than you expect.

I see a connection between the Gell-Mann Amnesia effect (read a newspaper article about something you know and see how wrong it is; go on to assume everything else you read is just fine) and the Dunning-Kruger effect (non-experts have a sense of illusory superiority, because they don’t know enough to know better).

In whatever area you work or have expertise, you should be able to identify how wrong most people are about it. Most obviously I find this in any question that begins with “Why don’t they just…?”

A personal example: I want a new mobile phone, and battery life is much more important to me than weight or how thin it is – but I can’t find such a phone on the market. Why don’t they just make a version of a great phone that’s much thicker in order to have a bigger battery?

But if I try to imagine the sorts of things I’d know if I designed phones, I can imagine that there might be complicated cooling issues with a thick battery; or perhaps they make trial-concept versions of potential new designs and have people use them for a few weeks, and discover that even if you think you’d be okay with the weight/thickness trade-off, it’s ultimately too annoying. I can even more easily imagine there’s a reason that I can’t conceive of at all!

Knowing the answer to these questions about one’s own areas of expertise, one should really extend it to other areas. When you start to ask “Why don’t they just…” you should remember Gell-Mann’s newspaper experience and the Dunning-Kruger effect, and consider if perhaps you just don’t have the expertise to spot the problems with your idea.

Another example, from an area of personal expertise: in anything involving code, new updates bring new bugs. The frequent response is “Why don’t they just test it properly, and make sure all the bugs are fixed before they put out the update”. The problem with this is that, on average, it takes longer to find each incremental bug than the last (things that occur one time in ten take longer to find than those that go wrong every time; things that only happen under very particular circumstances only turn up if you test extremely large combinations of actions, etc). As such, to truly guarantee all bugs are fixed would take so long that the customer waiting for their bug-free update will have moved on long it arrives. In the real world, ‘the perfect is the enemy of the good’, and compromises have to be made to get anything done.

Follow ups like “Why don’t they just hire more/better testers”, “Why don’t they let players/customers test it first” run aground in a very similar way. (Don’t forget though, there might actually be reasonable ways to improve things a bit! Customers are great at identifying problems, but it’s on you to figure out the solutions).

So we see why professional efforts aren’t improved as simply as consumers often assume . I think there are four other reasons consumers can be disappointed by updates: expectation failures, familiarity contempt, confirmation bias, and in the longer term, simply ageing. Here’s how those break down:

Expectation failure
A combination of factors tend to make us inordinately sensitive to our expectations not being met. For example, I use the internet on my train commute, and one time the train went into a tunnel, so the internet cut off, and I immediately felt annoyed. I realised this was a ridiculous response: I’ve made the journey hundreds of times, I’ve literally made a spreadsheet to identify where in the journey internet access drops. But in that brief moment, my expectation of continued internet access failed, and therefore in that moment I was annoyed.

In creative media this seems most obvious in movies, where the most significant factor in how much someone enjoyed a film seems to lie less in its objective qualities but more in how it compared to their expectations.

As noted above, a follow-up thing must be at least a little different to the original thing. Expectations are based on the original thing, so some expectations can’t be met – so some disappointment is inevitable.

Familiarity Contempt

‘Familiarity breeds contempt’ sums this aspect up well.

The IMDb ‘goofs’ section for 1977’s Star Wars Episode 4 is one of the longest of any film. But this is not because it was a terribly made film. It’s because so many people love it so much, it has received orders of magnitude more scrutiny than most other films.

I think this applies to the LEGO example, and especially to successful mobile games. These superfans are so close to the material they see every flaw. Something that seems perfectly functional to a casual player could be perceived by the superfan as being riddled with bugs.

So this is another effect: the people that engage the most with a thing will also be the most knowledgeable about its flaws.

In a fan screening, the completely missable moment this stormtrooper bumps his head generates a wave of laughter, as everyone knows where to look for it! It’s great.

Confirmation bias and performative criticism
Reality is complex and nuanced, so I think we make it more digestible by applying confirmation bias. If there’s something/someone we think is good, we are more likely to overlook or discount their flaws; if we judge something to be bad, every fresh piece of evidence is another chance to spot something wrong with it. It takes active, conscious effort to try to maintain a balanced view.

As a thought experiment, when was the last time you changed your mind about something? I find this to be alarmingly rare in myself. How likely is it that my immediate judgement of something was wrong and should have been corrected, once further evidence emerged? If I’m very optimistic, maybe not that often – but certainly much more than actually seems to happen.

I think this applies to many artistic endeavours. If for any reason your judgement on the creator of something has soured, you are likely to apply confirmation bias to their subsequent works and seek out their flaws to confirm your belief.

With art taking many forms, this produces an interesting corollary: the sooner you experience a follow-up work to something, the more likely you are to apply confirmation bias and continue to enjoy it – or hate it. If you are enjoying a TV series, I think a new episode is more likely to benefit from positive confirmation bias than a new series, which in turn is much more likely to maintain positive bias than a revival of the series many years or decades later.

Let the hate flow through you!

Related to this, I noticed that many criticisms of The Last Jedi in particular were bad-faith interpretations of plot, applying a level of scrutiny no screenplay could survive – an example of negative confirmation bias. It seems as if there are incentives to enact a kind of performative criticism once you flip into negative confirmation bias. So as an experiment, I wondered what it would be like to apply this to that sacred text, Star Wars Episode IV: A New Hope (as it was called in the re-release; just ‘Star Wars’ in the original release, which is the one post-release change that interestingly escapes criticism).

Please put on your flame-proof goggles and hold your nose as I have at it.

There’s so much wrong with this film it’s hard to know where to start, but let’s just go from the beginning.

Darth Vader is a terrible villain and an idiot. His “big entrance” is to come in after a bunch of cannon fodder have already done all the fighting, and then he does some kind of useless interrogation on someone that’s already been subdued. More importantly, he’s chased this ship to Tatooine, where the rebels presumably have a contact, and when the Death Star plans are obviously sent down to the planet in an escape ship, what does he do, as the guy with the ability to sense things with the Force? Just send down a bunch of useless stormtroopers to walk around asking if anyone has seen any droids lately?! What exactly is Vader doing while that’s going on – did he have to leave in a hurry to get to the big exposition meeting in the Death Star?!

And then later on, the Falcon, the very ship that escaped Tatooine – obviously with the plans –  gets caught by the Death Star; Vader is right there standing in front of it, he senses ‘something’… and then walks off?! Oh it’s fine, we’ll just send in the guys with the scanning machine instead! But wait, what’s with this scanning machine you have to physically take into a ship to operate? At the start when the droids take the escape pod, the Imperials scan it remotely for life forms, so why can’t they do that in the Death Star?

Okay, how about the heroes. Well, Luke is the most pathetic magical orphan character ever conceived. There’s literally nothing likeable about him. All we know is he apparently wants to get off the planet to… join the academy (what, the Imperial academy? What kind of ideal is this?!), but all he does is slouch around and whine about his chores. His step-parents are then apparently roasted alive (weird tonal imbalance, BTW) and he’s sad for about two seconds. People talk about how he’s a great pilot, but the first time we see him in a vehicle he’s in a land speeder keeping an eye on the scanner – while C-3P0 drives?! Whatever happened to “Show, don’t tell”?

Oh, but he’s magic! Obi-Wan teaches him to ‘stretch out with his feelings’ and suddenly he can hit a target no computer can hit, in a spaceship he’s never flown before! And also, by the way, despite never having fired a gun, he’s actually a crack shot, able to shoot out door controls from across a room, and out-shoot trained soldiers in multiple encounters! Also: he’s given an incredible weapon in the form of a light sabre… and then literally never uses it. Has Lucas never heard of Chekhov’s Gun? Why didn’t Luke use his light sabre to escape the trash compactor?

Oh, but the film has such a great back-story, right? Obi-Wan Kenobi has been apparently waiting for Luke to grow up so he can give him his father’s light sabre, and train him in the force. But, er, what was he waiting for exactly? Could have started a little earlier maybe? If Darth Vader had actually come down to sort things out Luke would have been killed before they even met! As it is, Obi-Wan literally gets in about 2 minutes of training before dying! (Another tragic death which Luke shrugs off in less than a minute, before running off to man the ship’s gun-turrets, which, by the way, is yet another thing he’s never done before that he’s apparently great at).

Alright, how about the Death Star. This whole thing is one giant plot hole. So apparently you can fly around the galaxy at faster-than-light speed in a station the size of a small moon? Why even build a hugely expensive space station then? Just stick a light-speed engine on a moon and fly it straight into any planet you don’t like! Oh, but I guess if that’s possible maybe the planet could just fly out of the way with their own giant engine!?

And how about that Death Star security? Literally any random droid can unlock doors and operate machinery from any random access port? Except tractor beams, which can only be disabled from a lever on a vertiginous ledge for some reason? And doing any of this doesn’t notify anyone anywhere apparently? And there’s so little CCTV that a bunch of random idiots can run around and you don’t even know where they are? They literally escape a dead end by jumping through a waste pipe and nobody can figure out to just throw a grenade down there?!

So probably the most sensible character is Princess Leia. She spends the whole time being rude to everyone she meets, but she’s at least smart enough to figure out they are being tracked when they leave the Death Star – but what does she do about it? Try to find and remove the tracker, or maybe go to another system and switch ships? No, let’s just go straight to home base, it’s fine, we’ll have at least half an hour for our techs to figure out a weakness in this giant planet-destroying space station that can be exploited by, er, about 30 small ships! I’m sure we can do that before it blows us all up. Wow! Good thing the plot wants this to succeed or this rebellion would be over!

And then finally, the big climax: the trench run. First, what even is this trench? And the secret weakness – literally one torpedo in the wrong place on the outside of the station blows the whole thing up? What kind of design is this exactly?

Given this lucky gift, what do the rebels do? Take it in turns to fly really far away from the exhaust port, and then spend ages flying along this trench to reach it, so the enemy can take them out one by one?! If you have the element of surprise, why not go straight to the exhaust port? Or why not all go at once? And even if there was some reason for the long run-up, the day is apparent saved when the Millennium Falcon comes in at the last second and shoots the bad guys at the end of the trench – er, literally any of the other X-Wings could have done that on any of the previous runs?

And…. breathe.

So, what just happened there? Wasn’t that far too long? Why yes, yes it was. Writing it was easy, fun, and made me feel smart, so I wanted to keep going. I can actually now understand why someone might read/write things like this over a prolonged period of time, rather than positively engaging with something they enjoyed. If it seems hard to imagine, I recommend giving it a go with something you know well yourself!

(Edit: I’ve seen cases of people reading this post and then arguing against the points raised in the screed above. This is the opposite of the point! Almost none of those arguments ever occurred to me before – I only thought of them when I considered the film specifically with an intent to pick it apart. I could argue strongly against them myself! Rather, the point is instead to consider a film you love, and then try to find the problems with it. Going through this exercise was very revealing for me as I noted above. It’s also interesting to examine how you’d argue against your criticisms; if you find yourself extrapolating beyond what is explicitly shown in the film (eg. the mechanics of different life-form scanners), or drawing on other material not in the film itself (e.g. Anakin’s history with Tatooine), do consider that these approaches could also benefit films/games/whatever that you are less inclined to extend the benefit of the doubt – T.M. 13th Jan 2022).

It’s not them, it’s you: fandom vs. ageing

On a time scale of multiple years or even decades, a newly significant effect enters: the viewer/consumer/player themselves has significantly changed. I suspect this is where Star Wars suffers the most. The main saga films are (I think) most enjoyed by children; when a new trilogy comes out 16 years after the last, those children have grown up!

I remember many Star Wars fans who were disappointed by the prequel trilogy nonetheless really enjoying the 2003 animated TV series. This had a lot of over-the-top action that would never stand scrutiny in live-action form, but I suspect the animated format bypasses a lot of the adult reality-check apparatus, and allowed these folk to be childishly delighted in the way they originally were.

I particularly remember Mace Windu taking down an army of battle droids in a way that I am confident would not have been received positively in live-action form.

This became particularly evident with Disney’s more recent Star Wars sequel trilogy. I’ve seen many young fans citing the prequel-trilogy as a superior era (echoing – yet reversing – the response to the prequels at the time, when they were reviled by fans of the original trilogy). This is also evident from the comments endorsing the Scene 38 Reimagined video – a fan video replacing the somewhat feeble 1977 Obi-Wan vs Vader lightsaber battle with something more like the prequel or sequel trilogies (and which would be utterly out of place in Episode 4).

So there are many reasons that account for strong criticism from fans, but don’t get the wrong idea: that should never be taken as an excuse to disregard all criticism! Criticism from consumers is often valid and useful for companies to heed (especially that familiarity/mistakes one) – you just have to take these effects into account.

3. Why do people remain fans of things they seemingly hate?

See what I did there

So, companies disappoint people because money can’t buy quality in artistic areas; and also because you can’t please all the people all of the time but companies want to continue and grow. Fans will be disappointed by new works due to expectation failure, confirmation bias (earlier disappointment drives fresh disappointment), familiarity contempt, and ageing; their analysis of the flaws may well be completely valid and useful feedback, or may be flawed due to the Dunning-Kruger effect.

The obvious conclusion is that people would simply stop consuming things they hate and find something new. But evidently, in many cases, that doesn’t happen. Why?

First, I think it’s a clue that everything affected by Paradoxic Fandom is in the arts or services; products don’t seem to have the same issue.

My friend John Broughton referred me to the 1970 book “Exit, Voice and Loyalty” by Albert Hirschman. To paraphrase significantly, “exit” means a customer stops using the service, and “voice” is what happens when “exit” is not possible: the customer voices their feedback on what they don’t like in the hope it will improve. I think this explains why products get off the hook – exit is fairly easy, you just buy something else.

In creative products, there is no easy exit. If you don’t like the sequel to a game or film you love, there is no alternative version you can switch to. In services, switching is either impossible (you can’t take everyone you interact with on a social network with you) or has a high enough cost that it’s better to stay. In the games-as-a-service model, it’s amplified further: when an online game updates, there’s no way to carry on playing the old version you liked better. To keep playing you have to accept the changes.

A fascinating exception that proves the rule: Runescape found a way to work around this. They released ‘Old School’ Runescape, recreating the older version that most fans first fell in love with – while continuing to develop the newer version. They now continue to develop the ‘Old School’ version, but all changes in it must pass a majority vote by the players.

So if you love a thing, and it changes in a way you don’t like, and you can’t switch… well of course you’ll be vocal about it! This is perfectly reasonable!

However, as “Why don’t they just…” is almost always unfounded, and as cultural products must appeal to more than just one person, very often that feedback won’t (or even can’t) be heeded. At this point, things can get pretty rancorous.

4. Why does criticism dominate the discourse?

From Paradoxic Fandom to Toxic Fandom

SamSykesSwears summed up the stages of toxic fandom (I think referencing the pattern of abusive relationships) in this tweet:

  1. I love this
  2. I own this
  3. I can control this
  4. I can’t control this
  5. I hate this
  6. I must destroy this

For example, the changes George Lucas has made with each edition of the original Star Wars trilogy particularly provoke “I can’t control this”. The clue to the irrationality here is how in many reviews, literally every single change is reviled. It seems improbable that literally every change the original creator would make to their creation before it came out was good, and every one after is bad. This looks far more like a near-religious adherence to some sort of holy text.

I think this was insightfully extended in a response from EricVBailey:

  • I can’t destroy this
  • I am even more mad now
  • I can harass those who are part of it though
  • I found a whole community of people doing this
  • I have found my validation
  • I love this

Self-selection and feedback loops

Even given all the above, it seems odd that a casual glance at much online discourse tends towards negativity, and especially the more toxic end of it. I think two things are at work here.

One is self-selection. This is easily seen in Amazon reviews of products: the majority of reviews seem to be people who only just got it (so have nothing useful to add), or who have had some terrible problem. This is because both of those moments are cues to leave a review. Using something and having it work just fine does not prompt you to go write a review. Similarly, playing a mobile game or watching a movie and simply enjoying it does not motivate you to review / talk about it online as much as hating it does.

My armored walker was destroyed by primitive weapons… would give 0 stars if I could

So self-selection skews what people tend to write about things.

The other factor is Feedback loops.

There are a few feedback loops online that end up fomenting more toxic discourse. One that has been well-covered (I thought particularly well by Tom Scott’s Royal Institution lecture) is that algorithms optimising for people to spend time on a platform will tend to find success by showing people more extreme and click-provoking content, which is often negative. So YouTube will naturally take you from “10 Things You Missed In The Last Jedi” to “27 Last Jedi moments that made no sense” to “237 reasons I hate The Last Jedi and You Should Too”.

There’s also a very natural social feedback loop that can work in tandem with this. I think for many people, it’s quite scary to make comments you think others will disagree with. If you found problems with something everyone loves, you’re less likely to shout about it; but if drama-optimising algorithms are showing more people that agree with you, that will embolden you to speak out more (see also: politicians making racist remarks emboldening racists).

My colleague Chris Hohbein saw this play out dramatically in the No Man’s Sky community. After that game’s launch, the community was incredibly toxic (mostly due to the game failing to meet their expectations); as updates to the game improved things, the hate diminished, and positive discussion flipped over to dominate the discourse instead.

Perhaps this doesn’t sound like that strong an effect. As a person who watches the first midnight showings of Star Wars films and feels moved to make comment on them in public right after, I can testify that the feeling of not knowing which way everyone else is going to go does make it feel a bit scary! That said, it’s Star Wars, I really should have figured out the pattern by now…

What I told you was true… from a certain point of view
Paradoxic Fandom certainly doesn’t apply to everything, and some of the above noted effects operate in reverse – for example, as in No Man’s Sky, confirmation bias can be positive; positive online conversation can beget more of the same.

But there are some particularly interesting counter-examples. For example, from that joke at the start, why is Star Trek different? And what about the ever-expanding Marvel Cinematic Universe?Are they doing things right in a way that others don’t?

In the case of Marvel, I’ve seen the argument that they play it “too safe”. In most Star Wars films, someone significant dies or loses a limb; Marvel films are frequently a battle to retain the status quo, and good guys and bad guys alike usually survive. So they entertain in the moment, but don’t do anything that might upset anyone. Over the last 10 years and 23 films, there have been a handful of notable exceptions from that, which seemed acceptable. Do they just have the right frequency of ‘not rocking the boat’**?

For Star Trek, each TV series will also tend to avoid disrupting its own status quo, but it’s unclear to me why subsequent series in different settings don’t seem to get as much hate as other franchise follow-ups*. This one is a mystery to me. Maybe I should become a Star Trek fan.

*Edit: I am reliably informed that I’m guilty of my own familiarity / contempt bias, and that actually Paradoxic Fandom is alive and well among Star Trek fandom. Prior to writing this, I had taken a brief look over various YouTube trailer comments, Reddits and forums to compare the Star Trek vs. Star Wars communities, and this seemed to confirm my hunch (or bias?), but this was hardly a rigorous study, and multiple people have now informed me of my error!

**Edit: This was written in July 2020. Since then the MCU launched Phase 4 which has been much less successful across various metrics, and is a whole other story. Still, it’s impressive they avoided negative feedback loops for so long and across so many films.

Conclusion: What do we do about it?

As a company / content producer

The biggest impacts on us as consumers are when things are better or worse than we expected. To the extent that you can, you want to exceed expectations. In practice this is tricky – if you hold your best bits or biggest surprises back from the marketing, perhaps fewer people will want the product. If you announce updates to an online game in advance and always under-play things, fans will spot the pattern and then be disappointed any time you fail to overdeliver. If you hold everything back, fans will worry nothing is happening at all.

To any extent possible, you can also be honest about things – educate consumers about the challenges you face. If you’re sure you’re right and they’re wrong, can you explain why clearly? Doing that – without making it an attack – can help both producer and consumer get closer to the more useful aspects of feedback.

More generally, I think it’s helpful to be aware of the above effects. Feedback from fans is extremely useful to help learn and improve; discounting it entirely is unwise. Knowing what drives it in different forms can help you unpick what’s most useful.

As a consumer / fan

I think a couple of Star Wars analogies help a lot here. As an engaged fan, do you choose the light side, or the dark side?

Luke: Is the dark side stronger?

Yoda: No. Quicker, easier, more seductive. […] A Jedi uses the Force for knowledge and defense. Never for attack.

You can use your deep knowledge of the material to engage in performative criticism (attack), or for knowledge and defence: what good-faith analysis can explain plot-holes? What practical aspects of production led to these compromises being made? What can we learn more generally about the craft from these decisions?

Luke: What’s in there?

Yoda: Only what you take with you

If you go in to a new experience ready to attack, you will find material to attack. Going in to anything new with an open mind, searching for elements to admire as well as those that justify criticism, is a more enlightening and fun way to go about things.

As an individual trying to find entertainment, I do think this is how you win – not by focussing on what you hate, but what you love. I distinctly recall spending the first 15 minutes watching The Force Awakens alternating between thinking “that’s not very Star Wars” and “that’s just a rip off of this earlier bit of Star Wars”, before realising no film could ever walk that line. With that in mind, I was able to enjoy the rest (of that film and the sequel trilogy as a whole) – while still finding a lot to criticise too! I find it highly rewarding to go in looking to enjoy everything I can, while also gaining enjoyment from engaging my more critical faculties later on.

There was just one more thing…

There’s a very important generalisation of Paradoxic Fandom. Consider these more generic forms of the regular mobile game complaints:

  • Every change makes things worse
  • This group/activity is failing and will die out
  • Those making the changes are out of touch with participants
  • Those making the changes don’t care about ordinary people

Does that sound familiar?

This is almost a textbook description of Populism in politics! The incumbent government is portrayed as out of touch with the people, that the country is failing, the government only care about themselves/the elite/the rich.

The country you live in has that same crucial quality we saw from “Exit, Voice and Loyalty”: it’s something you can’t easily change, and all the above listed effects can apply to politics just as they can to a game or multimedia franchise.

This means something! I just need a few more years to think about that.

Tim Mannveille tweets as @metatim, and sometimes writes about Star Wars on NothingAboutPotatoes


It Takes Two to Tango

Not a real headline: may contain traces of Photoshop. The cat is called Wadsworth.

[This post was originally published in March 2013 on the blog of Stubble & Glasses, which sadly no longer exists.]

“Average [English] man has 9 sexual partners in lifetime, women have 4”
The Telegraph, 15th December 2011

I see a headline like the above every so often, and when confronted with such a significant finding (or any finding, or really just any number) the first thing I do is apply a sense-check. In this case, the result clearly makes no sense.

It takes two to tango

Imagine all men are on one team, all women on another, and the two teams are playing a very long, inadvisable, and poorly-thought-through game to see who can get the highest average number of sexual partners from the other team. The big problem is this: any time one team ‘scores a point’, shall we say, the other team necessarily gets a point too.

Since we also know there are roughly the same number of women as men, the game is always going to result in a tie: with each team fielding the same number of players and each team getting the same number of points, the average score has to come out the same. Cracked helpfully have some diagrams demonstrating this just in case you’re not sure.

So, the sense-check failed. Let’s understand this discrepancy, while keeping things completely Safe For Work.

Fix 1: Confirm figures at source

The Telegraph provided a citation. There, we find that the methodology is fairly standard, and that more specifically the figures are actually Men 9.3, Women 4.7. This means the male team is claiming a 98% higher score than the female (rather than the more dramatic original figure of 9 to 4, or a 125% more), but this is still a long way beyond the expected 0% difference.

Fix 2: Check assumptions

There’s one major assumption in the sense-check which you probably noticed already: that the figures quoted are for heterosexual pairings only. Fortunately (for the purposes of this analysis) according to the original study, they are.

Fix 3: Estimate loopholes

There’s a major loophole with the argument that the game should result in a tie: this is technically only true if we’re considering the entire viable population of the world. This particular survey only applies to England, so we should consider what we might call ‘away games’.

In the absence of data, let’s just make a rough guess and see where it gets us. What if 1 in 10 of the men’s scores came from away games, and only 1 in 20 for the women? (Of course, this is going to bias the figures the other way for the rest of the world, and the real figure might even be reversed – we’re just trying to get a feel for how much of the discrepancy this might explain).

Well, applying those adjustments would leave the men with an 87% higher score, at 8.4 to 4.5.

Fix 4: Allow for subjectivity

There is no referee in this game, which means there’s a certain amount of… interpretation as to what really counts. This could prove a highly diverting side-track if you were discussing this in the pub, but for now let’s just imagine men might overstate the case 5% of the time, and women might understate 5% of the time. This takes the score to 8.0 : 4.7, with the men down to a 70% claimed lead.

Fix 5: Misreporting due to societal pressure

Having corrected for any major discrepancies arising from the methodology, we’re left with one major problem:people lie. So, perhaps men are tempted to exaggerate and women to down-play their actual scores when it comes to number of sexual partners.

They say you should divide a man’s claim in this area by three, and multiply a woman’s by two – this overcorrects for the data we’re considering, but perhaps that’s because people are slightly more honest in a survey than in the casual conversation this rule of thumb is probably meant to apply to.

Fortunately, a widely reported study was carried out on exactly this effect. Men and women were asked this same question, but in different settings. Some were told a researcher may look at their answers, raising fear of social judgement. Others were hooked up to a (fake) polygraph machine, creating pressure to be more honest. (If you’re interested, check out the original paper here).

Women reported 2.6 partners when worried someone would look at the answer, but 4.4 under a fake polygraph. Men reported 3.7, but this went up to 4.0 under the fake polygraph. Ah-ha!

This is interesting because it suggests both men and women down-play their actual figure, but most of the discrepancy is coming from the women. If we apply these corrections to our estimated figures so far, we have men at 8.4, women at 8.0 – much closer.

Unfortunately the study was small (with just 200 participants answering this question), so while these particular results are suggestive, the researchers concede that they are not statistically significant. (As with any emotionally charged research subject, this didn’t stop the media reporting on the result as an established fact).

If you want to get technical about it

As well as being small, that study was only conducted on college students aged 18-25 in the US, who one would frankly hope behave somewhat differently to the general population of England.

Even in the original more robust English data set, there are some fascinatingly subtle problems. Sexual behaviour will change over decades (some of which is covered in the study), and the extent to which people lie about it will also presumably vary significantly by age. In combination with the fact that men and women have different life expectancies, and that cohorts by age group are not actually equal, this introduces some additional distortions to our assumption that we should see a tie – although a few quick calculations suggest these effects are likely to be smaller in magnitude than those we estimated above.

Okay but what’s the actual answer

This excellent paper goes beyond aggregated data to study the distribution of responses, and convincingly finds an explanation. It turns out that the discrepancy is driven primarily by those claiming over 20 sexual partners, because these rare-but-average-biasing individuals evidently round their score (which they may have difficulty remembering accurately) in the direction they deem to be more in keeping with society’s expectations – so men round up and women round down.

In Conclusion

If you skipped straight here, you should know that you missed some fun stuff where we talked about some subtle issues with the parity assumption, and you probably didn’t notice that that compelling-looking chart was actually not statistically significant. But if you just want the two-line answer, it’s this:

The overall average number of heterosexual partners has to be almost identical for men and women. The discrepancy found in studies like this arises primarily due to people with >20 partners rounding the figure they report, which they probably can’t remember exactly, up (for men) or down (for women), in response to perceived societal pressure.

More practically, always sense-check your data, especially if it’s self-reported and on a sensitive subject. And if you can’t make sense of the data, ask an analyst.

–        Tim Mannveille

analysis game

Learning to Cheat Without Breaking the Rules, Part 2: Frameworks for Cheating

For the same reason as in Part 1, these posts are illustrated using the Hand of Fate: Comic Strip Playing Cards by Karen Rubins. This is Part 2. Part 3 is here.

In the first part I explained how a sequence of games taught me to become comfortable with bluffing and even lying outright, from a position of not being able to do either. That’s simple. The harder question is this: what kind of games can teach you to bend or break rules?

To tackle that, we should probably step back and ask how games teach us anything. And to tackle that, we should probably make sure we have a working definition for a game.

Here’s one provided by Greg Costikyan (emphasis added):

A game is a form of art in which participants, termed players, make decisions in order to manage resources through game tokens in the pursuit of a goal.

Yes, something like that. From there it seems clear that playing with decisions in a game is how we learn. So the question becomes instead: what kind of game could let us play with the decision of whether or not to break rules?

Form or Content?

I don’t know of one, but I imagine you could make a game about cheating (in the rule-breaking sense, not just bluffing or lying). Perhaps players take the role of political factions within a Kafkaesque bureaucracy, with in-game costs and benefits for following or breaking in-game rules. Or it’s Moon colonists with competing sets of societal rules, struggling for domination while following or breaking the rule sets of the competing factions. Or, I wonder, do football simulations these days give you the option of going for a risky tackle, knowing that you risk a red card if it goes badly? I don’t know, but they certainly could.

But in the process of abstracting away the costs and benefits of cheating, it feels as if something gets lost. Just as making decisions in poker has a very different feel when real money is on the table, a game with a cheating theme seems far removed from the potential power of a game that is about cheating the very rules of the game itself.

So we want a game whose rules somehow give players interesting decisions to make regarding whether or not to follow those rules, which sounds like an almost Gödelian paradox. So let’s take a step to one side and look at exactly what it is that drives whether or not rules are followed in a game.

Game frameworks for cheating

There are two key dimensions to consider here, I think:

1) Do people play the game to win, or do they play it for fun? Perhaps more accurately, does the pleasure come from winning – or striving to do so – or just from taking part? There’s a spectrum here, of course, and players of the same game could be at different points along it. For the sake of simplicity we’ll assume it’s binary.

2) How are the rules enforced in the game?

  • By design. In video games this is simple: the game should only let you do what you are allowed to do. Other kinds of games can also facilitate enforcement by their design – for example, playing a piece in Connect 4 makes a distinctive noise, so it’s hard to secretly take an extra turn if your opponent is briefly distracted.
  • By the other players. In most well-defined tabletop or party games, it’s understood that the players are watching each other to ensure they all play by the rules.
  • By a moderator. Generally speaking, if a game can’t work with either of the above approaches, it resorts to some kind of moderator who enforces the rules.

Let’s take a look at how cheating works across the 6 kinds of game this framework implies, and in particular whether choosing to bend or break rules is an interesting* choice for the players to make.

I should probably define ‘interesting’ here. Let’s say that an interesting choice is somewhat balanced (one choice is not obviously better than another), not damaging to longer-term goals, and is fun. And then let’s not try to define ‘fun’ or we’ll be here all day.

Games Played to Win

Play to Win, Design enforced (e.g. PvP video games)
The rules are enforced by the design of the game. As a corollary, anything you can do ought to be allowed, since cheating is theoretically impossible. In reality, it’s not that simple.

A reasonably notorious example is ‘snaking’ in Mario Kart DS. By performing a particular manoeuvre when cornering you get a speed boost. For certain karts, you could perform this manoeuvre almost continuously, alternating left and right – ‘snaking’ – and you could gain more from those speed boosts than you lost by taking a wiggling route. This takes some skill and practice, which makes the benefit seem fair. And since it’s possible in the game, it should be permissible, even if it’s not clear it was the designers’ original intent.

In practice, however, snaking gives such a big advantage that a player that does it will almost always beat a player that doesn’t, no matter how skilled. People are playing to win – that’s what usually makes a race fun – but if even one player is snaking, anyone that can’t snake can’t win, and so, implicitly, can’t have fun. For that reason, the London DS meetup group I played with had a no-snaking rule, and I see on the internet that other groups did as well. Snaking was considered a kind of cheating, and nobody would want to play with you. So no-snaking becomes an informal rule, and this becomes effectively the same as Player Enforced Play to Win.

(You can dive much further into this topic with Sirlin’s “Playing to win” archives)

Play to Win, Player enforced (e.g. Chess or Go)
As noted by Kirk Battle in this Kill Screen article, players can apply a certain level of flexibility to how rules are enforced in a game like Magic: The Gathering, and get a better game as a result. But this is not a situation in which such ‘cheating’ represents an interesting choice.

In it’s more blatant form, cheating in these games will always be considered bad form, and if discovered can have bad social and future gaming consequences. Chess or Go (when played without an audience) fall into this category. A player would only choose to cheat if they thought the short-term advantage would outweigh the risk of long-term negative consequences if they were found out. That’s not a very interesting choice either, according to the criteria above.

Play to Win, Moderator enforced (e.g. Football, Tennis)
Cheating outside of the moderator’s sight gives an advantage, and can’t be stopped in this system. You’re playing to win, and if you don’t exploit this fact, maybe the other guys will.

We see this in competitive sports like football or tennis, and it seems to be (I say this as an outsider) a key part of the entertainment: arguing about whether something counts as a foul, which side of the line the ball bounced on, and in general whether or not the moderator’s decisions are accurate.

Does this mean there’s some scope here for a game that could teach us about cheating? It seems close, but if people are playing to win, it gets dangerous – some degree of violence attempted outside of the moderator’s views seems likely, so I’m going to disqualify this on the grounds of encouraging non-fun behaviour.

Games Played for Fun

Play for Fun, Design enforced (e.g. Endless MMOs)
If you (and everyone else) is playing for fun, things change. If something is possible, and makes for more fun, few can blame you. If you were (somehow) playing Mario Kart just for fun, and through a streak of bad luck found yourself in dead last by some margin, perhaps it would be okay to snake your way to 7th place – that would be more fun for all concerned.

In Kingdom of Loathing, a kind-of MMO in which fun is primarily derived from exploration, an ‘exploit’ was discovered to generate more meat (the game’s currency). Who could resist such an exploit? Why resist it? Could you really begrudge those that used it?

In the case of online games like this, the rules are enforced by design, but there’s also some moderation in the form of game updates and code changes in response to things like this. In this particular case, the exploit effectively crashed the game’s economy, which impacts everyone’s ability to have fun. (Brilliantly, this was fixed with the addition of some entertaining currency sinks, rather than some kind of hard rollback).

So cheating is kind of interesting here, but again becomes more about player- and moderator-enforcement, so strictly speaking this category is ruled out.

Play for Fun, Player enforced (e.g. Mornington Crescent, DDR)
At this point, what constitutes ‘cheating’ is massively dependent on the players: having fun is more important than the rules. We’re right on the edge of what constitutes a game here and it’s an area I think most adults struggle to give themselves permission to enter. The most well-known example I can think of in this category is Mornington Crescent.

Less directly, this arises in Dance Dance Revolution (aka Dancing Stage), in which (for the benefit of the one person reading this that doesn’t know) players must step on specific directional arrows in time to the music. On the arcade machines, a raised bar is supplied behind the player, ostensibly to prevent anyone from falling off the dancing platform backwards. At a high level of gameplay, working out how to shift your bodyweight between feet while meeting a high-speed series of instructions is part of the challenge. However, some players realised they could lean their weight back on the bar and tap away with their feet without worrying about this issue; this also uses less energy. Is this cheating? As could be expected, that depends on who is playing.

In both of these cases, the decision of whether or not to cheat tends to have very little riding on it, so is unlikely to be interesting.

Play for Fun, Moderator enforced (e.g. Schooner or Later)
The Moderator’s role is to ensure people have fun, arguably as a higher priority than ensuring that people follow the rules. In this context, you might ‘cheat’ but do so with the moderator’s implicit or explicit blessing; or you might try to cheat by hiding your action from or misleading the moderator, and since people are only playing for fun this shouldn’t lead to anything particularly harmful. With the right set of incentives and approach by the moderator(s), this could well provide a framework in which a decision to cheat is actually interesting.

In conclusion, returning to our grid of possible games, here’s what the options for learning to cheat look like in each:

This final category is where The Haberdashery’s Schooner or Later comes in, the game that led me to cheat in a manner that could only be described as brazen, and shocked me into this whole line of thinking. I’ll describe how exactly that came about in the third and final part of this series.

Tim Mannveille tweets as @metatim, and has previously not cheated in order to earn stickers he made up at the Edinburgh International Film Festival.


Inception Diagram and Explanation (spoilers, obviously)

If you’ve somehow found yourself looking at this page but actually want to avoid Spoilers for Inception, you should leave now.


People are posting Inception Theories all over the internet, but even the ones that agree with me aren’t explaining it properly. So naturally I’m here to try to put that right.

Of the major categories of theories out there, my preferred interpretation of the film is this: Mal Was Right. More specifically, at Mal’s external direction, Saito incepts Cobb to believe he must wake up, and Ariadne purges him of his demons. Here’s a diagram that shows what I think the underlying setup actually is (click for full size):

I’ll first explain what’s happening in that diagram, and then explain as briefly as possible why I think that’s the case, closely referencing lines in the film at each point.

The Diagram

Prior to the moment shown here, the flashbacks seen in the film took place. Mal (Marion Cotillard) and Cobb (Leonardo DiCaprio) went into limbo, Cobb incepted Mal so she would want to wake up, which she did first by joining Cobb on the train tracks in limbo and then by jumping from the building after failing to convince Cobb they were still dreaming. She thus reaches the top level of the diagram.

Cobb remains in a nested dream. In that level, he believes Mal is dead, blames himself, and although he wants to get back to his kids he’s punishing himself for what happened. At the same time, he is haunted by his subconscious projection of the pre-incepted Mal, who just wants him to hang out with her in limbo forever.

At the top level, Mal realises that Cobb is effectively in danger, as he could potentially be drawn back into limbo and lose his mind. She calls Miles (Michael Caine) for help; Miles brings Saito (Ken Watanabe) and also realises the character we know as Ariadne (Ellen Page) might be able to help. They formulate a plan to incept Cobb so that he will wake up.

The Reasoning

When interpreting a film like Inception, a good guideline is to try to take as much of the film as possible at face value. The more of it it you treat as being a misrepresentation, the more interpretations become possible, and things quickly get out of hand.

The following is based on notes I took while watching the film for a second time, so while the dialogue may not be word-perfect the sequence of events and key lines are accurate.

The Setup

The first scene to note is in the helicopter, when Saito requests that Cobb perform an inception. Cobb asks for a guarantee, but Saito has no way to prove he can and will arrange for Cobb to go home. His final plea to Cobb is interesting:

Saito: Do you want to take a leap of faith, or become an old man, filled with regret, waiting to die alone?

Cobb seems to be moved by this, and ultimately agrees.

The next scene to note is when Cobb visits Miles, where the following exchange takes place:

Miles: Come back to reality Cobb… please.
Cobb: This last job, that’s how I get there.

That’s top-level Miles, creating the subconsious idea that after this job, Cobb can come back to ‘reality’.

Miles then arranges for Ariadne to join the team. As they begin to train her, Ariadne investigates Cobb. At one point he is connected to the dream machine alone and she joins him, to discover that he is attempting to lock away the projection of Mal, but evidently can’t resist going back to her. She presses upon him the idea that this won’t be enough:

Ariadne: You’re trying to keep her alive, aren’t you? You can’t just create a prison of memories to hold her in. Do you really think that could contain her?

Once the inception begins, Saito is shot, and it is explained that under their heavy sedation death will put you into limbo, where time passes much faster and you can effectively lose your mind. At this point there is a reprise of the earlier dialogue as Cobb expresses concern that Saito will fall into limbo and forget their arrangement, but Saito reassures him:

Saito: I will still honour our arrangement
Cobb: No, you will be an old man
Saito: Filled with regret…
Cobb (seemingly unsure of why he is saying this): … waiting to die alone.
Saito: No, I will come back, and we will be young men again.

The Missing Scene

Later, Ariadne pushes Cobb for more detail on what happened in the past, and we are given a more detailed flashback. He explains how he planted the idea that they were not awake, and this idea remained with her even after they both escaped limbo by lying on the train tracks.

Mal is seen at a chopping board, toying idly with the spinning top in her left hand. At this point, one particular scene is conspicuous by its absence. We know Mal uses the top to determine if she is still in a dream. The natural thing to do would be to spin the top, but we are not shown this happening.

We can infer what must have happened. If the top had kept spinning, Cobb would have been forced to admit she was right; therefore, it must have stopped. Why was this not enough to convince her? The clue is in this piece of dialogue:

Cobb: If this is a dream, why can’t I change anything?
Mal: Because you don’t know you’re dreaming.

Mal’s hypothesis would appear to be that as the dreamer, Cobb shapes the world to his own assumptions – he doesn’t think he can change anything, so nothing changes; he doesn’t believe they’re dreaming, so the top falls. (This idea is also endorsed by Saito’s line from near the beginning: “In my dream we play by my rules.”)

While this might seem a bit of a reach, we can also infer that the use of totems, invented by Mal, changed after this point. It instead became about the heft of the object, and rules about others touching the object were introduced. Cobb learned from the mistake of using the top.

(We might also ask why such a scene wasn’t shown. My guess is that showing Mal make this argument in any more detail would lend too much weight to the “It’s a dream” interpretation at the end; it would also risk alienating the audience seeing it for the first time, for whom the top is a key navigational aid).

The Final Setup

Finally in the flashback we see Mal’s apparent suicide, where she asks Cobb to make a leap of faith. Now we understand something strange is happening – Saito echoed this line in the helicopter, then associated with it the risk of becoming an old man, filled with regret, waiting to die alone.

If Mal was right, and her suicide woke her up, then the situation depicted in the diagram arises. The best way for her to get Cobb to wake up is to incept him, just as he did to her. But this would be difficult, because he knows how it is done. We’re also told the mark for an inception must believe they came up with the idea themself. As we’ll see, this is exactly what happens.

At the close of this interlude, Ariadne lays out exactly what Cobb has to do in order to move on:

Ariadne: Your guilt is what powers her. If we are to succeed you have to forgive yourself, and confront her

After the apparent failure at the snow fortress, it’s Ariadne who proposes they enter limbo to bring back Fischer. There they confront Cobb’s projection of Mal. Guided by what Ariadne has said earlier, and helped by Ariadne in the moment, Cobb finally rejects her:

Mal: We can still be together… right here.
Ariadne: You can’t stay here to be with her!
Cobb: I’m not. […] I can’t stay with her any more because she doesn’t exist.

The Inception

Ariadne and Fischer kick out of limbo, leaving Cobb to bring back Saito in a climactic scene given special emphasis by being introduced at the very start of the movie. Here, the inception takes place: Cobb echoes the lines he has been primed with, convinced that it is his own idea that he and Saito are not really awake, while actually delivering exactly the message the real Mal wants him to receive (full dialogue taken from IMDb):

Saito: So have you come to kill me? I’ve been waiting for someone to come for me…
Cobb: Someone from a half remembered dream…
Saito: Cobb? Impossible – He and I were young men together, now I’m an old man.
Cobb: Filled with regret…
Saito: Waiting to die alone…
Cobb: I’ve come back to remind you of something… something you once knew…
[the camera dwells on the still spinning top]
Cobb: that this world isn’t real…
Saito: To convince me to honor the arrangement.
Cobb: To take a leap of faith, yes. Come back, and we’ll be young men together again. Come back to me…
[Saito reaches for the gun]
Cobb: Come back…

As Cobb delivers these lines, he looks confused, as if he’s not sure where they are coming from. He has been incepted.

The plan as originally stated seems to be complete, and Cobb returns to his kids. He spins the top, but does not wait to see what happens to it. It’s apparent that at least the first result of his experience is that he is no longer concerned with which world he is in; having forgiven himself and let go of his regrets, he allows himself this moment of happiness.

If the inception worked, it’s possible over time he will come to believe that Mal was right after all, ultimately committing suicide. On the other hand, thanks to Ariadne guiding him to reject the temptation of his projection of Mal, he’s no longer in danger of falling prey to limbo. As such, even if the inception doesn’t work, eventually enough time may pass that he will simply wake up when the dream comes to an end naturally.

Bonus Points

The idea built up in Cobb’s mind is that if he does not escape the dream, he will become an old man filled with regret. It’s interesting that the song chosen to signal the end of a dream is “Je ne regrette rien” / “I regret nothing”, associating waking up with having no regrets.

Even under a more straightforward interpretation of the film, Ariadne is very mysterious. Aside from having an obviously fitting name for someone that will lead Cobb out of the labyrinth, Cobb observes that he has “Never seen anyone pick [dream architecture] up so quickly”; she’s also aware of the plan while in the dream, despite the fact that Cobb later says to Fischer (possibly as misdirection) that to do so takes “years of training”. Her active part in investigating and then purging Cobb’s hangups is also particularly obvious on a second viewing.

Finally, a nice line which is by no means definitive, but still gives you pause if you’re thinking along these lines at the time you hear it:

Cobb (on the phone to his kids): Mommy’s not here any more.
Child: Where?