Beginnings and Endings in public performances

Knowing how to start and stop

If you have to give a public performance of some kind – from a wedding speech, to a presentation at work, all the way up to a play or musical performance – I would suggest that starting and stopping are the very hardest bits to get right.

It sounds simple in theory, but there’s really a lot going on!


The beginning is the most important part of the work.


The ritual of starting a public performance of any kind has two challenges: people turning up late, and drawing attention to the starting moment. This applies to a speech or presentation, but also to a gig or play, a sporting performance, or even a pre-recorded work such as a film at the cinema.


For the ‘turning up late’ problem, the default strategy is to deploy some form of less consequential entertainment, for a duration in proportion to how late people are likely to be.

With scheduled TV programming (which is still apparently a thing), people are likely to be pretty much on time, so a 30s-60s opening sequence works just great. It has a distinctive soundtrack so other people in the house who might want to watch can also hear it starting and rush to the sofa.

For a film at the cinema, people have to travel to get there so are likely to arrive later than planned, and then realise they really need the toilet, or find the queue for the popcorn is longer than they anticipated. So there’s quite a margin of error; a 15-25 minute period of adverts and trailers is about right, and of course has other benefits to the theatre.

For a gig, the calculus gets blurrier still: the venue wants to make money from drinks, and the performers want a full, warmed-up crowd, so a warm-up act of some kind about an hour before the main performance is perfect.

There are two areas where I feel like society has not happened upon an adequate solution. First, attending a theatrical performance has the same challenges as the cinema, but I’ve never seen that liminal time between arrival and performance filled well. Second, and of more direct interest to me, in a work presentation – and especially an online one – there’s an awkward period of 2-5 minutes between the first arrivals and the latecomers.

Back when we still went in to offices, I solved this problem for a particular weekly meeting by using the TV method: I played a ‘theme tune’. This had the nice effect of raising the energy level, and also signalling to those elsewhere in the office that the meeting was about to start so they should get a move on – precisely the same benefits as a TV show, but a 2-3 minute song is about right instead of 30s-60s.

Now that these meetings take place remotely, we still have that same space of time, but a theme tune does not work well at all. There’s no ambient way of reminding those who haven’t joined yet, and the music would stifle any conversation between early attendees. I still don’t know how to solve this. My very partial solution is to set my video background to something I’ve been interested recently as a soft cue for small talk!

Calling attention for the start of the performance

Since we have a gap between the early arrivals and the latecomers, we then have the second problem: signalling the true start of the main performance.

Indoor public venues have a great cue for this: lighting. Dimming the lights is instantly noticeable and clearly signifies the audience to pay attention to the stage. At the Brussels Puppet Theatre I saw a particularly powerful version of this idea: you hear three quick, sharp knocks (which quickly silences almost all of the audience), followed by three slower knocks, each of which is precisely synchronised with some, then most, then all of the house lights going out. Attention is rapidly and tightly focused!

Cinemas have a harder time deploying the lighting trick as they have generally already engaged ‘movie mode’ to show the adverts and trailers. Some will transition from partial house lights to darkness (not a particularly dramatic change), and in some European cinemas an automatic curtain draws closed over the screen and then re-opens to mark the change – kind of pointless, but effective! In other cinemas, more notably, they play a very distinctive and loud descending run of three chimes, reminiscent of the puppet theatre’s three taps.

Without those tricks, the opening moments of the film itself can also do the job. In the UK at least, after the near-continual sound from the trailers, the moment of silence while the age certification is displayed forms part of the ‘things are starting’ signal. The animated logos of production companies do the final part of focusing attention, especially when accompanied by significant sounds – 20th Century Fox’s fanfare being particularly effective.

Occasionally films may deviate from the logo pattern, at their own risk. Dancer in the Dark (2000) was intended to open with the Overture playing while curtains remained drawn across the screen; in the UK and US, where auditoriums generally lack such curtains, it instead plays out over a black screen, to the confusion of the audience.

(Side-note: for Things readers interested in all things ‘meta’, Dancer in the Dark is notable for tacitly inviting the audience to leave before the film ends. Although I consider it an excellent film, I fear everyone will experience some form of regret whichever choice they make.)

More recently Dune (2021) forces audiences to sit up and pay attention by opening immediately with a (subtitled) quote in an alien tongue; very effective in the screening I attended because the volume was so loud!

Returning to the office setting, cueing the start is another problematic area. There is at least a well-defined approach for a more formal presentation, in which a host of some sort introduces the speaker, but there is one crucial rule such hosts must know: audiences are primed to clap after hearing a name. A successful introduction therefore follows the pattern: “Please welcome serial victim, unknown assailant, and universal stand-in, the enigmatic… John Doe” <applause>.

I frequently see hosts destroy this moment simply by reversing the order: “Please welcome John Doe” <scattered uncertain claps>, “who needs no introduction” <disconcerted silence>.

For a single presenter, there does not seem to be a universal method of kicking things off, which is unfortunate. They simply have to try interrupting everyone with a generic opening remark like “May I have your attention please”.

In less formal and more raucous settings, I was introduced to a fantastically effective alternative by – I think – Tim Sheppard: you say “If you can hear me, clap” – and then you clap. You repeat this a second time, and now roughly half the audience should join you in the clap. This gets the attention of the rest of the audience, so on the third repetition, everyone joins in, and is instantly giving you their full attention. I note this also echoes the rule-of-three seen earlier with the puppet theatre’s knocks, or the cinema’s chimes.


The end is important in all things

Yamamoto Tsunetomo, Hagakure: The Book of the Samurai

Endings are a little easier: there’s only problem to address and that is the ambiguity of whether or not the show is over. But perhaps because it is less problematic, this can end up less decisive in many contexts.

In film and TV, the end-credits can make this very clear – although this is currently opening up into a battleground for new norms, as streaming services lean in to automatic credit-skipping (to the delight of some and horror of others), and as the Marvel Cinematic Universe in particular is normalising the idea of the mid- or post-credits scene.

The theatre can make wonderful use of their curtain, and there’s a delight to be had in the ritual of the actors coming out to take a bow – but after that, the audience and actors both have the difficult problem of judging whether continued applause can or should be deployed to bring on further rounds of bows.

Musical performances such as a gig are very neatly divided into individual songs; the ambiguity then becomes about judging which song is the last. The level of enthusiasm an audience is likely to display is also a little too much for a band to feel comfortable standing there and accepting. Both these problems are resolved with the odd but completely accepted ritual of the band leaving the stage, then after judging the performatively correct length of applause, returning for an encore. Any real ambiguity is ultimately resolved by the raising of the house lights, which signals the ‘true end’.

I have seen this defied just once, in Björk’s 2016 performance at the Royal Albert Hall (written up nicely here). Despite performing the standard encore ritual with a radically re-worked version of Pluto, the crowd was not satisfied, I suspect because unlike other London performances, at no time had Björk addressed the crowd directly.

Despite the universal signal of house lights returning, the applause continued, and from the stalls the crowd began to sing the repeated, unresolving five-note refrain from Pluto, the song they had just heard. The singing spread through the crowd and sustained, as the Uncut article notes, “a demonstration so overwhelming that it forces Björk back out of her dressing room to offer further benedictions and thanks.”

Ending a talk or presentation, like beginning one, is much more difficult. Just as introducing a speaker by name creates a perfect moment for an audience to welcome them with applause, the audience is similarly primed to applaud at the end – but the cues are far less well-defined.

One obvious pattern to avoid: do not smoothly segue into a Q&A section. By explicitly asking the audience for something other than applause, you deny almost any chance of it!

The guiding principle to end satisfyingly is simply to ensure the audience is confident you have reached the end. If you can pull that off, and have performed well enough, a small silence while maintaining eye-contact can cue the applause. My favourite way to signify the end is to make a callback to something you mentioned at the very start; that and other ideas are nicely summarised here.

A much more overt method is to symbolically bringing your own hands together, as you deliver your concluding remark, as if clapping yourself. I saw this used by a guest-speaker in a talk about giving talks (a meta-talk); the second guest-speaker completely destroyed the applause-cueing effect by launching into a question before the applause could begin! (That guest-speaker received their comeuppance at the end of their own talk, when the first speaker vindictively deployed the very same trick).

Finally, in an online presentation, as with beginnings, endings are even more difficult than the real-life version, notably because applause simply does not function. The audience is deprived of that cathartic acknowledgement. I’m not sure there can be any substitute for it, so you must instead figure out how to reasonably wrap up and cue the audience to leave the call. Even this is not trivial to do well.

In a Google Meet call, I had some success with a radical method: since participants can kick one another off the call, I invited everyone to play an impromptu form of battle royale by kicking off as many other participants as they could, until only one remained. Precise rules and their enforcement are difficult in such a game, but it seems to work well enough – the only problem is I fear it’s so intense and interesting in its own right that the participant’s feelings about the talk itself risk being lost entirely.

Conclusions for online presentations

Despite not being my intention when I started to gather these thoughts, this leads me to a very significant conclusion: in an age where remote work seems likely to become the new norm, we really need to solve the problems of beginning and ending in online presentations!

So I’m going to propose some ideas, inspired by what works elsewhere, and will try them out myself when the opportunity next arises. I suspect that over time manual solutions will emerge, and that these will ultimately be implemented by the platforms that host online calls themselves.

Waiting for latecomers in an online presentation

While one could deploy a song or video at this point, the way it eliminates any chance for banter among early arrivals is a big problem. But I’ve not seen group call service that can neatly accommodate multiple people talking at once, so you really have to lean into that.

  • Welcome each new joiner (even if you’ve already started!)
  • Engage arrivals in some light banter. This is of course the default option, but can be hard to pull off, especially for a less confident or outgoing presenter. So…
  • … you could explicitly have someone with a suitable personality act as the host, welcoming everyone in, leading the conversation, and cuing the main presenter when the time comes
  • … or you can use my trick of setting your background to something interesting to deliberately provoke small-talk
  • More radical options: you could tell a story one word at a time; each new arrival must add a new word, and everyone can give a recap of the story so far by saying their own words in turn. Sounds fun, but probably doesn’t work at small or large scales… and possibly nowhere in between either?
  • Challenge arrivals to name something (e.g. an animal that begins with ‘P’) that nobody else has so far said. This tacitly encourages early arrival and penalises latecomers, possibly a good thing. Again, may not work at small or large scales, but a middle-ground where this works is a bit more plausible.
  • Hold an informal vote on something low-stakes and binary; ask each member their vote as they arrive and keep a tally (e.g. “Cats vs dogs” or “Time is endless vs Time is finite”)
  • Sing a song that works in rounds; each new participant can join in at the right moment. Probably requires a very specific group of people for this to work, but perhaps if you established it as a routine…
  • Somehow have people arrive directly into small break-out rooms; then with a timed warning bring them back to the main room and begin!

Signalling the start of an online presentation

Depending on which method was chosen above, this may become easier… or harder. Given the way attention is divided, I think an audio cue is most likely to work.

  • Deploy the rule of three. Clapping or knocking seems too aggressive; playing a recorded three-note chime could work (like this). Judging the volume level correctly may be a challenge, but perhaps you test that out with the very first arrivals?
  • A more elaborate version of this would be to deploy an actual (brief) musical intro, probably more in the style of a Radio Jingle. This again has the volume challenge, and also requires more manual preparation – but perhaps for a meeting that repeats every week it could work well
  • You could deploy a ‘manual jingle’ – by singing something sweet and short! Sure to gain attention, but not an approach that would work for everyone.
  • Try to replicate the ‘if you can hear me, clap’ trick… replicating it exactly may or may not work; perhaps something more like ‘if you can hear me, touch your nose’ would be better? It depends on the default visual layout of the call, I suppose.
  • Failing all these, one could perhaps develop a very specific introductory sentence, long enough that any excess banter could be shut down, but banal enough that it doesn’t matter if you miss the first half of it. I’m not sure there’s a way to do this that wouldn’t come across as rude though. Perhaps something like “I’m going to begin the presentation in 8 seconds, in 7 seconds, in 6 seconds, in 5 seconds I’m going to begin the presentation, 3, 2, 1, now. Hello!”

Signalling the end of an online presentation

Other than notes on how to end a talk in general, the key problems here are to replace applause, and to cue the final end of the call.

  • If you started with a short musical cue, perhaps you could also signal the end the same way
  • “So, if you learned something from this presentation, please show that by waving at the camera – but if you’re still wondering about something I said, instead please stroke your chin and say ‘hmmmm'” (this gives a ritualised group activity and also segues nicely into Q&A)
  • Lean into the format and just say “mike drop!” and leave the call abruptly – either by walking off, or hanging up. Maybe re-join 5 seconds later for Q&A.
  • “That’s the end! Please show your appreciation by rapidly turning your video feed on and off”
  • Not repeatable, but for fun you could look off to one side and say “that looks like smoke…” under your breath, then deliver your final line in a slightly hurried way with an unconvinced smile before running off camera.

Closing notes

Ed pointed out to me that streamers address these problems online in a few interesting ways – running other videos as warm-up entertainment, or conducting text-based games for the waiting/arriving audience. However, these are quite specific to the streaming set-up, and I don’t think they can easily be used for the online work meetings I’m trying to solve for.

A new and much harder the challenge is also becoming more common: presentations where the audience a mix of those physically present and others joining online. In the absence of good online solutions, I don’t know where to start with solving the start and end of these meetings.

Finally, a side observation: arriving late to online meetings doesn’t feel as bad as arriving later to a physical one. However, the irritation at someone arriving late to your meeting is just as bad either way.

Consider arriving for online meetings early!

Transmission ends


One Thing at a Time

This blog began as an email newsletter, collecting a few links to things I found interesting that week. As my commentary on these links lengthened, it made more sense as a blog (as well as a newsletter).

Like a fantasy author in the late stages of a series of books, updates are getting both longer and further apart, and now that 2 years have passed since the last update, it’s clearly time for a change.

So, a simple rule: from now on, One Thing at a Time.

After a (currently undefined) period of time or number of Things, I’ll aggregate them in a newsletter for the usual recipients and possibly a post. Let’s see how that goes!


Things July 2019: Bias, Co-operation, Location, Video Games

Extremely Generic Questions

In earlier iterations of Things I would often set readers a puzzle or ask a tricky question, with answers reviewed in the next edition. I’d like to start a new series of these that I’m calling Extremely Generic Questions: questions often asked, but not very specific or even necessarily well-defined. The puzzle is not only to try to find the best answer to a question, but also to understand why it is so often asked.

So, the first question: what is wrong with young people today?

Subconscious bias

If you want to make optimal decisions, and also just be a decent person, I think subconscious bias is a very important factor to be aware of. Many people believe their judgement of aptitude is not influenced by gender (or any other unrelated characteristic), but data suggests this is not be the case.

Anecdotally but compellingly, there’s the email signature swap story, in which a male and female colleague swap their email signatures for a week and observe radical differences in how clients interact with them. Pleasingly, you can read the accounts of this from each side.

Auditions for an orchestra have the advantage that they can be conducted in a thoroughly meritocratic manner without ever actually seeing the candidates. It turns out that blind orchestra auditions improved women’s chances of success by 50%.

Similarly, scientific proposals for time using the Hubble Space telescope tried going partially blind; the results again showed women benefited from a more meritocratic process.

As I am sometimes involved with hiring people for work, I tried a version of this by making sure names were removed from CV’s before I reviewed them. Of course, I don’t have any large scale data to compare results, but the feeling of trying to assess a nameless CV was alarmingly transformative! It became very clear that as soon as I saw a name, I would start to construct a mental image based on (irrelevant) associations I had with people similarly named, and would then build on that image as I read the rest of the CV. Without a name as a starting point, the process of evaluation immediately felt like harder work, but also a lot more objective. Based on this and the above findings, I highly recommend it.

Thinking, Fast And Slow

After many years of seeing it recommended, I finally read Daniel Kahneman’s “Thinking, Fast and Slow”, which covers a lot of the biases I’ve been fascinated by for so long from essentially the very coal-face of that research.

I found it fascinating throughout and can see why so many recommended it, although I didn’t always agree with the interpretations. Given that I’m just some guy who studied maths/physics and read a few things, and the author is a Nobel prize winner with decades in the field, I have to recognise that my position on this is likely tenuous, but at least as far as maths go I feel I can comment.

Here’s my brief highlights.

Probability Problems
There is a strange and fuzzy line between not understanding what a question means, and not getting the answer right. I could ask people what 3 χ 4 is, and if they think it’s 12, that doesn’t mean they’re mysteriously misguided, it more likely means they just don’t know chi-notation*.  In many of the studies, participants were shown to give incorrect answers to statements involving probability, but one could just as well argue that participants didn’t really understand the statement and so were guessing. To be fair, the book goes on to show how phrasing probability questions differently (to my mind, more clearly) helps people reach more accurate results.

This is what I talked about in Things 122 on the topic of the Linda Problem / Conjunction Fallacy.

Forecasting and regression to the mean
I have to do quite a bit of forecasting at work, and I was surprised I had never come across this excellent rubric for anticipating a certain amount of regression to the mean.

Briefly: if you evaluate, say, fifty people on a task that involves some luck as well as skill, like accurately throwing something, then the people who did the very best (or the very worst) on their first attempt are unlikely to do as well (or badly) on a second attempt; their results were probably mostly flukes, and they will tend to ‘regress to the mean’.

If I am evaluating 12 different marketing campaigns and trying to forecast how well they do in future, the same kind of rule applies. The one that did very best was at least partially ‘lucky’, so will not necessarily be the best in future.

The rubric is as follows:
a) If the measure you want to predict has zero correlation with their future values (which you can figure out by viewing historical data), then you should predict that regardless of how they did, they will all perform averagely in future.

b) If the measure you want to predict perfectly correlates with the future, so whatever is the best now will be the best in future, then obviously you should predict that.

c) If the correlation between the present and the future is x%, then you should forecast any present deviation from average performance will decrease by (1-x)%!

That’s the terse version, you can read more about it here.

Evaluation of experiences
How much you liked or disliked an experience would intuitively be based on how long it was, and how much you were liking or disliking it at the time. Something that was unpleasant for 10 minutes should surely be ranked as worse than something that was unpleasant for only 5 minutes.

In practice, this isn’t how we evaluate things at all. We very highly weight our peak enjoyment (or discomfort), and how happy (or unhappy) we were at the very end of the experience, and a little bit the beginning; the absolute duration plays only a small part.

This probably means you should take fewer, shorter holidays, but it also depends on how you weight the importance of what Kahneman calls the “experiencing self” and the “remembering self”, which is quite a tricky philosophical problem.

Life is like an Iterated Prisoner’s Dilemma, except different

As a student I was very interested in the Prisoner’s Dilemma, and the iterated version, in which one must choose to Co-operate with (C) or Defect against (D) another player making the same choice, knowing that you will come out best (and they worst) if you D while they C, but you will both do terribly if you both choose D.

One critical element: you can’t communicate with the other player. In the iterated version, the choice of C and D is effectively the method by which you communicate.

Meanwhile in real life, most of the time, the thing most likely to help you is another human, and the thing most likely to do you harm is also another human, which means interacting with other humans is a pretty crucial business. In particular it’s good to be able to figure out – and influence – who is likely to co-operate (C) with you and who is likely to try to take advantage of (D) you.

I studied maths and physics as a student, but struggled to understand human behaviour. By studying subjects where exam answers were simply right or wrong, and doing quite well at those, I (and I suspect many others in the same situation) thought that I must be quite clever, and the reason I can’t understand human behaviour is because other people are just acting irrationally.

Now, it is true that people act irrationally a lot of the time (see the last Things), but that also includes me, and a lot of the things I couldn’t understand eventually made more sense when I realised that life was like an iterated Prisoner’s Dilemma, except that rounds aren’t discrete or simultaneous and there are multiple and varying pay-off matrices in play all the time.

For example, I noticed people asked “How are you?” but didn’t actually want to know, which seemed irrational. This was laid bare for me in a dentist’s waiting room when one elderly person entered and recognised another, and the following exchange took place:

A: Oh, hello there! How are you?
B: I’m fine, how are you?
A: I’m fine thanks. So [short pause] how are you then?
B: Well, I’ve been having this awful pain in my side, so I went to the doctor last week …

Similarly, as a marketing grad I was sent out with a cameraman to stop people on the street and get their opinions on climate change for a vox pop montage. I would walk up to people and ask them right away, and nobody stopped to answer. The cameraman, who had done this before, told me I should ask them how they were first. This seemed ridiculous, as a person approaching you with a microphone and film camera obviously doesn’t care how you are, they just want to film you saying something. But I tried asking anyway, and suddenly just about everyone was then happy to give their opinion on climate change for the camera.

I realised the whole “How are you” bit is like a tiny move in an iterated Prisoner’s Dilemma in which you are really communicating “I will co-operate with you”, and the other person can demonstrate a reciprocal intention by asking you the same thing back. This then sets the scene for further and deeper co-operation.

Moves aren’t just made in speech either. I thought buttoned shirts were ridiculous in comparison with t-shirts: uncomfortable, more time-consuming to put on and remove, and harder to clean. Why would anyone choose to wear one? But it turns out clothing is a widely understood opening move in our co-operation dialogues. We learn that we can estimate by someone’s clothes how likely they are to co-operate with or benefit us in certain ways; uniforms do this in an overt way, but even a slight deviation from your company’s dress code sends a signal.

Cat and girl covered this, of course.

More generally, these kinds of behaviours make a society cohesive – by doing what everyone else does, you tacitly signal that you are a good co-operator in your society. At the same time it can make society conservative, as anyone deviating from locally normal behaviour (even for rational reasons) might be read as less co-operative, and so they will encounter more friction.

Location Encoding

What3Words (W3W) assigns each 3m x 3m square on Earth a three word designation (e.g. Each.Useful.Shark). This makes it fantastic for real-world treasure hunts, so long as the participants can use the mobile app, and I’ve made a couple of events that leveraged it to (I thought) rather fun effect.

However, Richard brought to my attention that among people interested in the general problem of addressing, W3W is viewed very negatively. Why is that?

Reading up on the subject (this post was particularly useful), it seems like W3W lacks some attributes a truly general Location Encoding system should really have. But what really annoys people who understand this area well is that W3W tends to put out PR that claims to be strong in the areas it is weak. In brief:

  • W3W is a private company (probably hoping to be acquired by another one). Location/address is something that works best when it’s a standard, and having a private company own a standard leads to conflicts of interest. (See the Microsoft ‘Embrace, Extend, and Extinguish’ strategy for an example of how private companies fight public standards).
  • W3W is not a good solution for emergency situations (calling an ambulance to your location; calling a Fire Engine to a location you see on fire).
  • W3W is not error tolerant and has no hierarchy (e.g. one mis-spelt/mis-remembered character has very little chance of being corrected, in contrast with traditional addresses, where post accurately addressed apart from saying “Brighton” instead of “Hove” still successfully gets to Hove with the word Brighton angrily crossed out and corrected).

Still, I do think W3W has some value, and it would be unreasonable to discard it entirely because it can’t do everything – indeed, no address system can meet all the requirements we might ask of it.

Google Maps’ location-sharing functionality covers many options, and has the benefit of being already available in many people’s pockets, but I recently had a situation where both intuitive addressing and Google failed: meeting at the “Joe’s Café in Soho” does not specify a unique location, and the inaccuracy of GPS meant a shared Google location didn’t resolve the matter either. W3W is actually pretty excellent for this sort of spontaneous meeting. All things considered though, the best thing about it really does seem to be the opportunities for Treasure Hunts.

Video Games

I’ve played some games since the last Things, some of which I recommend, and some of which I don’t!

Baba Is You (Steam, Humble store,, Switch)

The mostly instantly-gettable trailer seems to be in this tweet:

For Things readers partial to self-referentiality and all things meta (and I know there are a bunch of you), this is certainly worth a look. As the above video shows, the game is played by pushing things around, including words that define the rules of the game.

In practice it’s even more mind-boggling than I expected, but not actually as much fun as I had hoped.

Celeste (Steam,, Switch, PS4, Xbox One, Pico-8 prototype)
If you want a platform game with puzzle elements and enjoy dying repeatedly while you slowly get better at doing difficult things, this is extremely the thing for you. The soundtrack is quite lovely too.

My save file, 44 hours, 10,000+ deaths, is a review in itself (implicit spoilers split-by-level version is here):

Lovers in a Dangerous Space Time (Steam, Switch, PS4, Xbox One)
Looks exactly like what it is: a rather nice local co-op shooter in which you and some friends control characters running around a ship manning the helm/gun/shield/panic-button and rescuing animals in space.

Thomas Was Alone (Direct for Mac or PC, iOS, Android, Steam)
A kind of “self-aware” puzzle-platformer that everyone was going on about a few years ago; I finally tried it and found it dull and not at all as funny as it seemed to think it was, with frustratingly vague platforming physics.

The Legend of Zelda: Breath of the Wild (Switch and Wii U only)
An open world epic about saving the world which I tried – and failed – to enjoy for about 12 hours before giving up. Almost everything about it felt like a chore and I couldn’t understand how it gained such universal acclaim.

After a weirdly long time adventuring in my underwear, I finally found someone who would sell me clothes.

To pick one example, a complaint I had heard from some was that weapons could only be used a certain amount of time before they fell apart and had to be replaced. I thought this was just misguided resentment of a feature clearly designed to add strategy to battles, but then I spent the first hour of the game picking up and discarding 37 differently ineffectual sticks (you can only hold a few at a time) and fought about 5 monsters. It felt like I spent substantially more time managing my ineffectual stick inventory than having battles, and so the whole weapons feature then felt like busywork.

Loading screen tip: read loading screen tips. By definition, this is useful to nobody.

Perhaps the worst part is I still feel like I “should” give it more of a chance, or at least get more entertainment for my money, when rationally I know there must be more games out there like Celeste which were an order of magnitude more fun for a fraction of the price.

Horizon Zero Dawn (PS4 only)
Snakes on a Plane is a great title for a film because it perfectly sells the premise. Horizon Zero Dawn is a terrible name in that regard, but the promotional art sells the premise perfectly:

Tribal humans hunting robot dinosaurs! Which immediately looks like something I want to try, and also very quickly raises the question of just how such a scenario could even come about. It turns out the game is exactly about those two things: hunting robot dinosaurs while figuring out how this happened!

It certainly stands on the shoulders of giants in terms of the use of ‘Open World’ game conventions, but it adds a few interesting ideas and does just about everything you would hope to extremely well.

*I made up “chi-notation”, because I couldn’t think of a clearer example, it seemed funny, and illustrates the point just as well.

– Transmission finally ends


Things September 2018: Disney, Star Wars, Optimisation and Motivation

Disney Live Action

With the live-action version of Beauty and the Beast surpassing expectations, and the trend of Disney live-action remakes continuing ever-onwards, I wondered how well these films were performing at the box office compared to the animated originals.

I think the fairest comparison is the US box office (as global distribution can vary massively over decades), to exclude re-issues (which made substantial money for the old animations in the days before home video), and of course to adjust for inflation.

Well, if you do all of that, here’s what you get:

So Beauty and the Beast did arguably out-perform expectations, although not by as much as Alice in Wonderland. I wondered if perhaps the original Alice was released at a time where Disney’s reputation had dipped, but funnily enough it actually came out in 1951, one year after the original animated Cinderella. Perhaps people just really love Tim Burton.

Why do people do the things they do?

Many believe themselves to be rational beings who do things for logical reasons. At first I thought this must be approximately correct, but I’ve gradually come to relegate it to the bottom of the list of reasons people do things.

Here’s my hierarchy of why anyone does anything in approximate priority order:

  1. They have done it before
  2. Other people are doing that thing right now
  3. It will grant them short-term satisfaction
  4. They have seen other people do that thing in the past
  5. It makes sense to do it

There’s a lot more nuance of course, but this seems a generally useful guide. examines motivation

There are some famous experiments that dig into ideas relating to this. If you’re interested in this sort of thing and know about the so-called Stanford Prison Experiment, you should read this excellent long-read on how the findings were misrepresented, and how that misrepresentation then persisted for decades. (Short version: the experiment was purported to show that people ‘slip’ into roles that society places on them, even doing abhorrent things for no other reason than it was consistent with their assigned role. In fact, people do abhorrent things if they are told by someone in authority that it is for a greater good, and this is what the results actually confirmed).

There are some interesting experiments on the bystander effect. These in general show that if there’s something it makes sense to do, but someone sees other people not doing that thing, they will default to not doing the thing. Failure to evacuate on hearing a smoke alarm is a classic example we experience regularly, reinforced by the poor ratio of signal:noise for those sorts of alarms. Ever since I spent some time in the Royal Holloway Founder’s Building, which I was told could burn to the ground faster than any fire drill had evacuated people, I’ve generally been the first person out of my chair, and I also like the idea of treating fire-alarms as a life-long “game” to try to be the first person out of the building whenever one goes off, but even with these ideas in mind, I still feel myself significantly held back by everyone else’s inaction.

(There are some interesting solutions to the fire alarm problem here, although they wouldn’t solve the Founder’s Building problem).

Structured Debate with Kialo

Back in the days of debating issues with fellow students at university, I got frustrated by how poorly dialogue worked as a method of reaching a conclusion, and visualised (very vaguely) some sort of system that could show the arguments all at once, and allow someone to explore an issue in a more considered way.

There have been a few attempts at that online since, the latest I know of being Kialo. For example, here’s a sub-argument about Universal Basic Income that I was interested to know more about.

Optimising for one thing makes everything else bad, including (sometimes) that thing

As someone that has spent many years analysing data to deduce what organisations should do, I’ve become ever-more wary of any efforts to improve a single metric. In general, the easiest ways to improve one metric will ruin other metrics, and a myopic focus on one thing for a long period of time is usually a path to disaster.

(Mathematically: A/B testing for something leads to local maxima; when the environment you operate in changes over time, local maxima can become very suboptimal)


– Taking the fastest route on a journey saves time, but may cost a lot of money. If you ask TFL for a route from Oxford Circus to Heathrow it will recommend the Gatwick Express, without revealing that it is disproportionately expensive for the relatively small amount of time saved. (Incidentally, the Citymapper app, unlike the site, makes this very clear, as it shows both the times and prices alongside one another)

– Companies that try to hit quarterly revenue targets are tempted to run massive promotions at the end of the quarter to hit targets, sacrificing longer-term profitability

– A social network might focus on improving daily engagement in order to drive more ad impressions. They will end up doing so in ways that reduce long-term engagement on timescales that don’t show up in short A/B tests. For example, a notification system that highlights when a user is mentioned or someone has interacted with their content is powerful; it can drive even more engagement if additional notifications are added to it for random other things, because this increases the number of notifications and people are trained to check them. But long-term this reduces the signal:noise ratio of a notification and is likely to ultimately reduce engagement. I’ve seen this exact example on Facebook and Twitter; I think the general problem is one reason Facebook is losing active users in mature markets.

Still, if you’re smart about it, you probably can optimise one thing without eventually making that one thing worse. When I first got a job in marketing (back in 2007) I became aware that internet services wanted you to spend more time on them to make more ad money, and that they would iterate and A/B test to get better and better at it, until we would find ourselves effectively addicted to online content. Then I learned how incredibly difficult it is to actually change behaviour, so I got less worried. Then, very belatedly, I realised that if you’re a vast venture-funded monopolistic internet behemoth with years and years to keep trying, and you’re smart about it, you gradually will discover ever-more effective methods, and sure enough by this point just about everyone I know (including me) uses monopolistic internet services for more time than they want to.

(Youtube and Netflix autoplaying more content unbidden and Spotify making it hard to queue up a finite amount of music being the most obvious examples; a personal favourite is the way the ‘search’ button on the Twitter app doesn’t actually initiate a search, but rather shows you new content, with the option to actually search available if you tap again on the least convenient part of the screen.)

A further challenge is that if you do manage to improve a metric long-term, most likely some aspect of quality in that metric will suffer.

For example,  a government focussed on reducing unemployment will be tempted to support anything that improves that measure, even if the forms of employment are less-secure or leave people underemployed or unsafe, decreasing the relevance of the original metric and potentially making the core problem worse.

Optimising for people spending time consuming your content long-term is likely to make the quality of that time go down. I think this sort of thing created the collective abomination that is children’s “content” on Youtube, which if you’ve never seen it is summarised in this article by James Bridle on the topic:

 “Someone or something or some combination of people and things is using YouTube to systematically frighten, traumatise, and abuse children, automatically and at scale”

And in this article on how ‘fiction outperforms reality” on YouTube, a quote from Zeynep Tufekci gives an apt analogy with food:

  “This is a bit like an autopilot cafeteria in a school that has figured out children have sweet teeth, and also like fatty and salty foods […]So you make a line offering such food, automatically loading the next plate as soon as the bag of chips or candy in front of the young person has been consumed.”

Once that system is up and running, however, Tufekci suggests that anything fractionally more edgy or bizarre becomes novel and interesting, and a single-metric-focussed content-recommendation algorithm will steer things in that direction.

“So the food gets higher and higher in sugar, fat and salt – natural human cravings – while the videos recommended and auto-played by YouTube get more and more bizarre or hateful.”

So what does that mean for someone like YouTube in the long run? It means people prepared to produce the highest-volume,  most compellingly-terrible content rise to the top. Penny Arcade’s Tycho, who does articulate rage pretty well, sums it up:

 “They made a kind of monster machine, with every possible lever thrown towards a caustic narcissism, and then they pretend to be fucking surprised when an unbroken stream of monsters emerge.”

Watch out for online review scores

I’m sure for Things readers it’s obvious that compiling a rating for something out of the people who choose to go to a website and give that thing a rating is not going to give the most objective results. But because of the Streetlight effect, we might be tempted to assume it’s at least “directional”, in that something with a higher score is probably better than something with a lower score.

Well, here’s two reasons to be a lot more cautious.

The most obvious issue is the self-selection bias. This was truly laid bare by the ratings of The Last Jedi. Here’s a comparison of the ratings from:

  • RottenTomatoes (a mostly-consistent pool of critics, comparisons are useful)
  • ComScore (a true random-sample poll, comparisons are useful),
  • Rotten Tomatoes viewer score (a hive of self-section and ballot stuffing)
  • IMDb rating by Male and Female (policed for ballot stuffing but vulnerable to self-selection):

That ComScore data and this general comparison comes from this BirthMoviesDeath article by James Shapiro which is well worth a read. Clearly some self-selection and ballot-stuffing can skew a metric.

A less obvious issue is a mix effect, in which the aggregate result is affected by the composition of the voters. Walt Hickey at FiveThirtyEight has some brilliant analysis of this effect when it comes to ratings from men and women on movies and TV shows. To reduce the article to just two charts, Men are overrepresented among IMDb voters:

… and men are more likely to rate female-targetted shows badly (using share of vote as a proxy for the target) than women are to rate male-targetted shows badly:

The end result is the global average rating for female-targetted shows will tend to be worse than male-targetted shows that are equally enjoyed by their target audience.

Finally, we have to remember to consider the Streetlight effect one more time. We can look at the data for male vs female rating because IMDb share that – but it seems very likely that the skew will be just as bad (or worse) for other groups.

Bitcoin follow-up

Last time I cited an article by Charlie Stross on Bitcoin, which built to a political conclusion from the assumption that Bitcoin didn’t make sense in the long run because (briefly) the incentive for people to supply the necessary computation to run it will disappear, and the energy requirements don’t scale (although to be fair he also drew out a conclusion on what might happen if it does work in the long run).

Thomas, one of the cryptography experts that reads Things, replied pointing out some issues with this assumption.

First, Stross assumed the processing incentive derives purely from the remaining (finite) amount of Bitcoin that is mined, but as Thomas quotes from the original Bitcoin paper, this problem was anticipated and planned for: “Once a predetermined number of coins have entered circulation, the incentive can transition entirely to transaction fees and be completely inflation free.”

As for the energy requirements, Thomas makes an argument I realised I had already made myself in other areas: Status Quo bias means we take the disadvantages of existing technology for granted, so when a new technology has different disadvantages it can seem much worse (Things June 2015 – Tesla owners review petrol car; Things 130 – review p2p games from the perspective of f2p, films from the perspective of games). So we tend to implicitly assume the new problems are significant, net-negative, and insurmountable- all of which should be carefully questioned for an otherwise promising new technology.

Reading around a bit more, it looks like the estimates of energy use may not have been reliable, especially regarding how it may scale in future.

Looking for a more informed, long-term view I found this by Daniel Jeffries. This has a good reminder of that Status Quo bias by pointing out that it’s easy to look at something future-like (e.g. an Encarta CD-ROM in the 90’s), identify problems with it, and then rashly conclude “computers will never replace encyclopedias”. So with projections of cryptocurrencies, not only are many unaware of the intended steady-state, they also tacitly assume no further advances will be made.

Quite excitingly, my opinion has now changed in light of all this (I’m always looking for moments when my opinion changes on something, because if it never happens you have to wonder if you’re really thinking about anything). I didn’t think cryptocurrencies would scale and be significant in the global economy; now I think they might.

I’ll leave the last word to Thomas:

“Bitcoin will either die or it won’t. Anyone who tells you which one with certainty is selling you something. The world has been given a taste of the benefits of cryptocurrency and there is no going back. Whether in the end it’s Bitcoin or a competitor that takes the throne, the advance of progress is inevitable.”

Star Wars update

As I remain the biggest Star Wars fan you know, you probably want to know my opinion about the latest films in the franchise! Alright, you probably don’t, but I want to tell you about them, so I’ll keep it brief, or you can just move on because this is the last Thing in this edition.

First, let’s have a recap in the form of data. Using Rotten Tomatoes as the most reasonable long-term comparator (although it’s hard to say how well the panel-based approach holds up over multiple decades as the panel composition changes), we get the following:

  • The Disney-era films (blue) outperform the entire prequel trilogy (red)… until Solo (2018)!
  • Episode III (2005) rated close to Return of the Jedi (1983), woah!

Now let’s check financial performance, looking just at the US domestic gross, and of course adjusting for inflation:

  • As I wrote before, the very first film was a crazy break-out hit, and no film since has gotten close
  • Each major series sees diminishing returns after a strong start
  • Solo looks like a pretty big disappointment…

Since we just learned about IMDb voting patterns of men vs women, let’s check those:

Note, we learned that men and women rate differently, but we can perhaps interpret directionally:

  • Since Empire Strikes Back (1980), just about all the Star Wars films show relatively more appeal to females
  • The Last Jedi (2017) looks like a particularly big outlier… but see the above bit about self-selection and mix effects

That’s all well and good, you’re thinking, but what did Tim think of these new films, as a Star Wars “Fan”?

First, I note that arguably 7 of the 9 post-New-Hope Star Wars films disappointed a notable portion of those who considered themselves “fans” of the films that came before. As such I feel like an increasingly rare sub-group of fans that has found a lot to enjoy in every single Star Wars film released to date. So, here’s my terse opinions.

The Last Jedi (2017)

  • Overall, extraordinarily refreshing after the uncomfortable familiarity of The Force Awakens (2015)
  • It subverted some tropes that were long overdue questioning
  • Finn’s story arc didn’t ‘read’ to most people I’ve spoken to (including me) on a first viewing, partially due to a couple of scenes that were deleted
  • As a Hero’s-Journey graduate by the end of RotJ, Luke risked being a narrative-ruining character that could just come in and solve everything; Force Awakens did a neat/cheap trick by keeping him out of it entirely; Last Jedi tackled this narrative problem head on and, I thought, to brilliant effect

Solo (2018)

  • I suspect this film’s box office was most undermined by a weak elevator pitch: compare Rogue One’s “How the rebels got the plans to the first Death Star” with “Han Solo got up to hjinks with Woody Harrelson when he was young.” Perhaps most damningly, I wasn’t even sure if I would go see it in the cinema – me, the person who queued up for the Star Wars marathon that culminated with Episode III!
  • I’m really glad I did because it turned out to be an exuberant ride with an almost perfect balance of reverence/irreverance for Star Wars lore (Teräs Käsi!), and introduced my new favourite droid
  • Some implied sexual abuse/exploitation rather undermines the overall light tone, making it a tougher movie to embrace overall
  • The Auralnaut’s review of Solo from the perspective of Kylo Ren is pretty great – harsh, but fair

Plot points and spoilers

As before, I can’t let this topic go without chiming in on some of the debate around plot holes!

In The Last Jedi, Holdo won’t tell Poe her plan for the Resistance to escape. Poe is frustrated with this, and as an audience rooting for him it’s easy to feel frustrated too. In context, Holdo’s decision is rational: being traced by unknown means makes one worry about a double-agent onboard, and letting too many people know of the escape plan only increases the risk it will be found out. Indeed, such a fear is immediately validated when Poe eventually learns of the plan: he passes it on insecurely, where it is overheard and later used against them.

So to be fair to Holdo, she made the right choice, but to be fair to Poe, if she had given a better explanation of why she couldn’t tell him, perhaps everything would have been fine (although knowing Poe I doubt it). And to be fair to the audience, giving Holdo a line to explain the reasoning more clearly (rather than poetically) would make the experience less frustrating.

The other big plot criticism is that the whole casino-planet sequence (Canto Bight) “feels” redundant; for me my first response was it felt “prequelly”, and was definitely when I stopped feeling on the edge of my seat. As noted above, I think this is down to poor communication of Finn’s arc.

I think the audience implicitly assumes Finn is now a 100% committed member of the Resistance, but all of Force Awakens showed pretty clearly that that was not his concern, and there is no reason that should have changed –  as he spent the intervening time unconscious (!). The setup as given in Last Jedi is that he wants to get away from the fleet so Rey will be safe on her return, and this is in conflict with Rose, who wants to save the Resistance by any means possible.

When the Canto Bight opportunity comes up it’s a chance for both of them to get what they want, with the big question being whether or not Finn chooses to come back. The events on that planet lead to Finn making a decision to commit to the Resistance, demonstrated most clearly in the final act when he tries to sacrifice himself to save everyone else.

So on paper, Canto Bight sounds like a great, engaging story arc in that Finn must overcome obstacles to achieve something but also learn or change along the way, ultimately not getting what he wants but what he needs. Two scenes that were deleted for time would have made this clearer (Finn showing his motivation at the start, and Finn explicitly choosing to return afterwards), and I suspect the writing generally needed to be clearer on it. Not that it wasn’t clear – with this in mind it’s all fairly obvious on a re-watch – but purely because the film had to work uphill against the audience assumption that Finn’s loyalty wasn’t in question.

[And now, a late addition! – T.M. 18th September 2018]

The Holdo Maneuver

Admiral Holdo takes out a significant number of ships by – it appears – simply jumping to hyperspace straight into them. This raises the question of how we should square this with all other Star Wars combat as it seems like the kind of thing people should be doing all the time (not least against a Death Star).

These days I’m much more comfortable accepting that there’s a good reason for these sorts of things in movies, and a film isn’t improved by wangling in some characters giving some exposition on the matter.

Still, if we have to justify it, I think we have to assume that the outcome wasn’t expected and the intention was just to cause a distraction and/or some minor damage. Things presumably turned a lot more catastrophic because of some other unique circumstance. As it happens there’s already a novel and related technology in play – hyperspace tracking. This is evidently something that hadn’t been seen before, and sounds like exactly the kind of thing that might cause unusual results if the target jumped to hyperspace into the tracker.

– Transmission finally ends