The horror of the quantified audience
The cinema is on the deserted outskirts of Pinewood Studios. It's a wet and windy day — not unusual for England in late summer — and I'm here to take a look at the future of audience test screenings.
My friend, Ross, has asked me to take a test drive of his company's newest way of <strike>destroying artistic integrity</strike> helping movie studios connect better with their audiences.
Thursday mornings aren't my preferred time to sit on a cheap polyester chair and tip my head back in wonder at a 200 inch screen but, hey, it's a showing of an unreleased movie. I'll suffer through it. I sign the NDA and stick my phone in the little locker provided. I keep the SIM card, just in case.
The movie studio rep is everything you expect — perky, blonde, super excited, and working hard to keep the superficial smile on her lips.
"Ok guys!" she enthuses, "On your seat you'll find all the kit you need to get set up. How many of you have been to a test screening before?"
Half a dozen hands creep up.
"Fantastic," she continues, without really registering the answer, "Well, this screening is going to be a little different — a little more high tech. If you're right handed, please put the bracelet on your right hand. If you're left handed, please put the bracelet on your left hand."
They had, so Ross told me later, wanted to use a head band. Slightly more accurate but too distracting for the audience.
"Ok guys! Hold up your hands so we can check you've got the wrist bands are on tightly."
Actually, it's so that the wrist band's accelerometer can zero correctly. This latest version we're using can just about tell the difference between you rummaging in your popcorn bag and scratching your balls. That's important, apparently.
"That's super! Now, can you all please pick up your tablets and touch the button on the top?"
She begins to explain how we can register what we feel about the movie using the sliders. We can also type notes saying what we do or don't like.
This, nominally, is why I'm here. Ross knows that I can break just about any app. Give me the world's most polished software and within five minutes I'll have it crashing. We all have our talents.
The tablet is a generic Android tablet. Mass produced, plasticky, crappy screen, and a battery which just about lasts the length of the film. The real reason for the tablet isn't because they're interested in our opinions. It communicates with the wristband using Bluetooth and then shunts the results over WiFi.
Perky-Blonde takes us through the tablet tutorial and then dims the lights. The floating faces partially illuminated by glowing tablets lend a rather spooky air to proceedings.
I won't spoil the film for you (car chase, explosion, soft-core sex scene, betrayal, witty banter from the sidekick, then the good guy defeats the bad guy) — to be honest, the film-makers have already spoiled the film. It's no different from anything else that's come out in the last few years.
Before Perky-Blonde lets us leave, we have to fill in a questionnaire on our tablets. Age, gender, income level, sexuality, favourite brand of lager, weight. All completely anonymised, so we're told.
I leave and meet Ross in his office to look at the data.
"I love you!" — "I know."
It's clear Ross and his boss are unhappy with how the test screening went.
"The sex scene really didn't work. Pulse rate data across the audience didn't rise nearly enough. The car chase at the start got more of a reaction."
"Did women like it?"
"Ha! They all pressed the ‘hate it' button — but looking at the galvanic skin response they fucking loved it!"
That's why they use the bands — people lie. They lie to each other and they lie to themselves.
"We also noticed a lot of fidgeting during [rather dull exposition scene]. We're going to need to edit it down somehow. People were reaching for their drinks and popcorn all the way through. Very low engagement."
"And the final act?"
"Good news there, cutis anserina response was good. The peril worked very well. Adrenal levels also had a steady rise."
One thing the backers were worried about was the homosexual sub-plot. Quite a departure for a mainstream action flick. Nothing explicit, but unusual even in today's modern world.
Ross turned to me, "You're gay, tell me what you thought of [the scene]."
"It was fine. But I'm not gay." Which is true. Some of my best friends are gay, but I'm a committed heterosexualist.
Ross brought up my profile on screen — so much for anonymity!
"Right there!" He jabbed at the screen, "vasodilation, and all the other signs of arousal. Right when the dudes are kissing."
Am I secretly gay? No, as Ross let on, just one of those facets of human biology. Even the straightest, gay-hatin', Bible basher's penis will react to sex acts — no matter who they are between.
"And the sidekick?"
The studio was rightly worried about the sidekick. A second-rate comedian whose last few movies bombed. I certainly didn't find him funny but, as became obvious, I was in the minority.
"The tablets' microphones recorded significant laughter levels at all the right times. I think we could even use more of him."
I cringed — the guy just didn't appeal to my sense of humour. Yeah, humour with a "u".
Ross saw my scepticism. "Let me show you."
A few clicks later and the screen filled with a dozen faces. Well, up-nose shots of faces. I hadn't reckoned on the tablets' cameras being active. To my shame, there was my nose. To my greater shame, there was my laughter. Judging from the readout below my image the sensors reckoned it was "sympathetic laughter". I was laughing because I heard laughter; not because I found it funny.
"Is it raining? I hadn't noticed."
The end result was a timeline impressively drawn underneath the movie. After correcting for age and weight, Ross's software was able to draw the peaks and troughs of the audience's excitement level. Men and women were separated out. The data the audience entered on the tablet was ignored — they went straight to the physiological source.
The movie makers are convinced that with enough feedback, they can create the perfect movie. One where the average audience feels the right amount of excitement, terror, sexual arousal, and amusement. All with the minimum of boredom and distraction.
The technology is impressive. Given how cheap the tablets and sensor bands are, they are surprisingly effective. The app was solid and, despite my best efforts, I couldn't get it to misbehave significantly. And, assuming the average test audience member doesn't bring in a USB-OTG cable and a pen-computer stuffed full of naughty software, they should be robust enough to distribute to audiences around the country.
But is this what we want movies to be? I can understand that studios see these as investment vehicles and sink millions of dollars into them. They want to see a return on investment and — in their twisted minds — this means pandering to the lowest common denominator. Nothing original can be allowed; originality is a risk.
PLOT OF EVERY BOOK EVER: Someone is looking for something. COMMERCIAL VERSION: They find it. LITERARY VERSION: They don't find it.
Once the "quantified audience" gains traction, it's game over for movies as an interesting medium. They'll become more exciting, funnier, and better paced. Boring sub-plots will be done away with. Sex scenes will become more interesting for men and women. Rob Schneider will never make another movie again.
"Here's Looking At You, Kid."
Ross rings me up a few days later to ask if I'd had any more thoughts about the Android app.
I had been wondering how 2001 would have performed under such cold and calculating conditions. I know it's a cliché, but I think it's the movie which defines what cinema can do. The gratification of that movie comes in the days and years after seeing it — when it pops unbidden into your mind and forces you to question the world around you.
Ross has been experimenting on his kids. He got them to watch classic movies from his childhood and measured their reactions. I think it's a little creepy to watch your teenager's brain being exposed to Phoebe Cates in Fast Times at Ridgemont High, or Christian Slater in Heathers. But what do I know?
"Based on our testing, it would be better if Ugarte's death was more protracted and violent. Ilsa needs to show more skin and — according to the audience — ought to leave with Rick at the end."
"Well… it's not exactly a contemporary…"
"Alien is too slow to kick off. Reservoir Dogs needs a more obvious end. Batman Begins causes people to fidget all the way through. Toy Story doesn't have enough laughs — and they're inconsistent. Before Sunset just doesn't get people's hearts racing."
"Anthony Hopkins in Silence of the Lambs is sexier than any man alive. Can we put him in a movie with Misery-era Cathy Bates?"
"That's unlikely. No."
There is a lull in our conversation. A moment of silence which, in a movie, would test very badly but probably be necessary for dramatic pacing.
"I think our data gathering devices are the problem. Wrist measurement just isn't accurate. We need to put them on the temples. Read their damn minds."
"I'm not sure if cheap Chinese Android tablets can do fMRI. Yet. Give ‘em a year."
Ross tells me that the studios aren't that interested in the technology he's helped develop. They love the concept and don't care if the data isn't accurate, but they're worried that it will be to expensive to reshoot the scenes which the audience don't like.
Video games are their next target. It's easier to measure a player's engagement levels and corrections are cheap to code. They'll be able to measure just how scary the monsters are, how dull the cut scenes are, whether the heroine needs a more revealing costume, and if the player gets too frustrated with a puzzle.
Shouldn't we let creators create? Does every damn thing need to be subject to a focus group? Isn't art about describing how you see the world — not showing people what they want to see?
Somewhere, a group of volunteers are having their vital statistics monitored while a variety of political slogans are shown to them. One day they'll find a phrase and a logo which will test off the charts. All they'll need next is a candidate.