Jessie Burlingame ends up handcuffed to a bed, keys just out of reach, not a single person for miles. Living person, at least. Her husband, the one with the bondage kink and the rape fantasy, is dead on the floor. The heart attack probably did it, but banging his forehead on the tiling didn't help. She's pretty sure he's dead, at least. There's too much blood.

This rekindling of the flame is going worse than she expected.

I'd said while writing about Before I Wake that Mike Flanagan likes his crucibles. The characters in his movies are always trapped in a situation and squirming to get out of it. Jessie is the personification of that narrative device.

On the one side of the bed, Gerald's Game is a sensational example of how to adapt a book. It tweaks details, focuses on the performers, retains the story's strengths while enhancing it with the right cinematic language.

Carla Gugino gives a protean performance as Jessie. There's little artifice she could default to, no gimmicks she can pull. She has to spend most of the movie in a single position, handcuffed in place as Jessie is. She doesn't even get a change of outfit. It's all expressions, intonation, what little body language she gets to show.

Bruce Greenwood does a great job as her husband Gerald, whose eponymous game put her in that position. He has a rakishness to him, with Greenwood looking better at sixty-two than I expect I'll ever look, but it's balanced out with a husband who still wants to care. Gerald is someone used to getting things his way, but who wants to make it work with the wife instead of going out for the hookers and mistresses he could have.

(And I'm sure does)

On the other side, there's the issue that as a faithful adaptation of Stephen King's book, it'll carry the weaknesses of the material.

For instance, what's up with her background?

We need a new term for what's going on with characters, where what used to be considered a tragic backstory is now the baseline. People have to be sadder, had to have had a worse life than you thought. A good tear-jerking story doesn't bring in as much waterworks as it used to. Their drama needs to be more dramatic.

It's so much misery inflation.

Is it not enough for a woman to be in handcuffs, ogled by a ravenous dog gone feral, dying of thirst, losing her mind and talking to both herself and her dead husband, terrified by (one would hope) hallucinatory threats?

No? Not enough?

Does she also need to have been sexually abused as a child as well?

I'd have been freaking out at the handcuffed thing myself. Maybe “being eaten alive” at most. We wouldn't need dredging up any past horrors to compound the present ones.

It's Stephen King's fault, not the movie's. There are too many echoes from his other stories, from the solar eclipse to name-dropping Cujo. He writes these things for himself now, all the stuff he's put out blending into one big gumbo pot in his head.

And then there's the inspiring, talky, tacked-on ending. King is a great tease, but when it comes to fucking, he climbs on top and writhes around for a couple of minutes before declaring he is done.

Before it gets to that duct-taped apocrypha, though, you'll get some masterful scenes. Gugino's arguments with herself are impeccable. Greenwood is the perfect bastard, released from all social constraints by death and her agonizing wife's imagination. There's a scene, as Jessie finally figures out a plan, which felt like it went on forever, me squirming on the couch, calling out “oh fuck no” several times. I exhaled when it was over. There's constant tension before and after that horrifying moment.

That's what we watch horror for, so the movie succeeds. Even if the coda feels like someone stapled it on right before release because they felt we needed to get an uplifting lesson out of the whole thing.

#mikeflanagan #stephenking #geraldsgame #carlagugino #brucegreenwood #horror #miseryinflation

Ugh. It seems that Netflix movies are like software – you need to wait a few weeks after release for ratings to stabilize.

Spectral seemed like it would be fun, cheap entertainment. It had a 4 out of 5 rating, which I thought was optimistic, but hey, if it passed a Saturday night, then great.

It didn't.

It's not even that interesting to dissect as a movie, actually. It's a hair above the quality you'd expect out of Syfy. What is interesting is what it might tell us about Netflix' data-driven production approach.

The story of how Netflix decided to produce an American House of Cards is at this point well know, but just in case: with the massive amount of user ratings they have, and no small amount of machine learning, they noticed that people who like the original British series also liked Kevin Spacey movies (and, I would expect, David Fincher). They decided to put them together to target their viewers.

If you squint, you can see where most bits on Spectral came from. Bruce Greenwood is a good choice of inexpensive “that guy” slash authority figure. Emily Mortimer is popular in secondary parts, even if not recognizable by most since she's never had a hit. The set design is so obviously geared to people who liked Terminator 2's future aesthetic that I'm surprised Michael Biehn isn't in it.

Throw in other look-alikes like Discount Ioan Gruffudd and Dime Store Jackie Earle Haley, and you've got yourself a movie.

I wish they sourced better writers, though. You could make a fun B movie with these elements. The idea of Moldova as “future Iraq” is not bad. But it's written without a shred of finesse, filling in blanks on a screenwriting template. You can bet your entire paycheck on there being plot shifts at 30, 60 and 90 minutes. There's a token tough female character who they still manage to turn into both “audience stand-in who needs things explained to” and “human subtitle machine”. You can be 100% sure Bruce Greenwood is there so he can deliver an attempt at a rousing speech before the last charge. There will be a massive face-off – duh! – where we are supposed to feel sorry for the extras sacrificing themselves as a distraction so that Science Boy and CIA Girl can infiltrate the enemy.

From plot to characterization, the script puts in the least amount of effort to even show up.

Not to mention, its shifting, pidgin language identified on subtitles as [speaking foreign] is not nearly as bad as the dialogue you actually do get to understand. It's one of those movies where, after a while, I just can't help but yell at the screen.

For instance, when they encounter disembodied nervous systems near the movie's end:

“How do you know if they can feel anything?” “I can't...”

How about there being an entire nervous system with a brain in that vat there?

“... there are some things science can't answer”

It's a bloody central nervous system! And a spine! And a brain! Which science school did you go to?

I expect in the near future Netflix movies will be written and produced by a convolutional neural network. Maybe then their genre offshoots will get better than this leaky Markov chain.

You have to wonder what the fitness function will be then, though. Right now it's ratings telling them what to produce. Eventually, it may turn to you being engaged, because as long as you are watching Netflix, you are not watching something else. Every two hours they get from you keep you one step further from going to a different service.

If that becomes the case, when in doubt, the only winning move will be not to watch.

Originally published on my old blog

#spectral #action #brucegreenwood #emilymortiner #sydfielded